Grid Sage Forums
- April 19, 2024, 03:25:49 PM
- Welcome, Guest
|
Show Posts
This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.
Pages: [1]
1
General Discussion / Question about the new branch
« on: February 27, 2019, 04:42:37 PM »
Hi! I just finished a whole buttload of Tangledeep runs and figured hot damn is it time for some Cogmind again. I looked over the changelogs since I last played and the new content looks great - balancing, RIF, Exiles, oh my! Congrats K on making an amazing roguelike even better.
[some minor spoilers follow]
I love the new Exiles map since it really gives you a reason to go into Mines which wasn't there before. Some great roleplaying happens down there too. My first time trying to find the Exiles, I (obviously) hit an infestation, have to dodge them and the demolisher teams sent in response, have to dodge a bunch of ARCs carrying pests, and only survive cause some derelicts appear from nowhere and do me a solid blocking a choke point. Your robomartyrdom is appreciated.
Here's my Q. In caves, it's very easy to find the special branches - just go left. Is there a similar mechanic to know which exit is the exiles exit, or is the only way to be sure carrying around a signal interpreter?
[some minor spoilers follow]
I love the new Exiles map since it really gives you a reason to go into Mines which wasn't there before. Some great roleplaying happens down there too. My first time trying to find the Exiles, I (obviously) hit an infestation, have to dodge them and the demolisher teams sent in response, have to dodge a bunch of ARCs carrying pests, and only survive cause some derelicts appear from nowhere and do me a solid blocking a choke point. Your robomartyrdom is appreciated.
Here's my Q. In caves, it's very easy to find the special branches - just go left. Is there a similar mechanic to know which exit is the exiles exit, or is the only way to be sure carrying around a signal interpreter?
2
Fixed Bugs & Non-Bugs / Delay in game registering loss of signal interpreter
« on: January 10, 2018, 06:38:13 AM »
So this is a tiny one. I was in caves with a flight build, sensors, and a signal interpreter. I was moving fast, about 374% or so. I was doing some digging and a cave-in destroyed my signal interpreter. On my screen, I could still see nearby robots that were out of sight represented by their tile as opposed to the "?" mark. I moved about 2 more spaces and could still see their tile. After a third move or so, they became question marks.
What I suspect happened is that since I was moving so fast, the game still showed their tile on the turn that signal interpreter was lost, and only switched to show the "?" after a game turn passed.
Like I said, very minor issue (if I'm in fact right about what I observed, I wasn't in a position to reproduce)
What I suspect happened is that since I was moving so fast, the game still showed their tile on the turn that signal interpreter was lost, and only switched to show the "?" after a game turn passed.
Like I said, very minor issue (if I'm in fact right about what I observed, I wasn't in a position to reproduce)
3
Fixed Bugs & Non-Bugs / [beta 4] Enemy corruption FCS glitch affects your targetting
« on: January 05, 2018, 07:48:25 PM »
Pretty easy to reproduce.
1) Attack an enemy with an EM weapon.
2) Each time you hit f thereafter, the target will be right over the enemy, per usual.
3) Corrupt the enemy a bit.
4) Have the enemy do a misfire via corruption, one of the new beta 4 mechanics
5) hit f again. Instead of being right over the enemy, your target square will be where the enemy's misfire attack went.
1) Attack an enemy with an EM weapon.
2) Each time you hit f thereafter, the target will be right over the enemy, per usual.
3) Corrupt the enemy a bit.
4) Have the enemy do a misfire via corruption, one of the new beta 4 mechanics
5) hit f again. Instead of being right over the enemy, your target square will be where the enemy's misfire attack went.
4
General Discussion / Should there be a corruption threshold before data loss occurs?
« on: November 30, 2017, 12:19:46 PM »
I think corruption is a smart mechanic that's handled well in the game. Once you become moderately corrupted you begin to suffer heat spikes, misdirections, parts fall off, and energy drains. Mostly harmless fluff, but in the right (or wrong) circumstances this can cause a fight to go south. You also have weapons failing to cycle or even worse firing randomly. You can't play cogmind without quickly learning to never carry a launcher around equipped while corrupted.
The one aspect of corruption that I think has some problems is data loss. The worst offender, maprot, has already thankfully been removed. Data loss as it remains essentially takes two basics forms. One, you forget what parts are. As a mechanic I think it works. After you play long enough, you know what parts enemies are dropping after a fight. What this mechanic really does is it forces you to make a decision when you encounter stockpiles. Maybe those unrecognized devices are advanced force fields, maybe they're field recycling units. You can strap one on to find out and restore the loss data but that will cost matter and, more importantly, time. It forces the player to make a strategic decision with food clock ramifications, so I think it works.
My big problem with data loss is that it can destroy data that's effectively irreplaceable. Some of the best rewards in the cave branches are the derelict logs. A log that gives you a branch entrance, zone layouts, or index(machine) for a future floor is huge. The problem is, it's so easy to lose this data, even in the face of extremely trivial corruption.
An anecdotal example: I was in the proximity caves on -3, looking for the exit to research. I found a log giving me the entrance to Testing on -3. That's a huge bit of data! I bumped into a programmer who gave me 2% corruption. I slapped on my system restoration module and made for the exit. In that interval I still managed to suffer two "data loss" corruption events. On -3, I did not get the intel reveal for testing, and had to find it the hard way.
Maybe that was exceptionally bad luck, but I feel like it's way too easy to lose data from derelict logs due to corruption. The whole point of branches is that they're high risk / high reward scenarios, and if the rewards can be that easily lost, they're much less desirable. I know that the derelict logs aren't the main rewards in each cave branch, but they're certainly big reasons why I go in in the first place.
While there are anti corruption devices in the game, I don't think that solves this particular problem this given how easily data loss can occur. A system restoration module can take hundreds of turns to get rid of corruption, and walking around with a corruption screen to block corruption effects 1) is not guaranteed to work, and 2) handicaps your combat abilities. They're certainly useful devices, but they're not going to prevent you from suffering the occasional data loss effect. As it stands, even with anti corruption utilities, as little as 1% corruption in caves can result in you forfeiting extremely valuable intel.
tldr: People are inherently loss adverse. They'd rather not lose something they already have than gain something of equivalent value. (See https://en.wikipedia.org/wiki/Loss_aversion). It's a bummer to lose a branch reward, even if a minor reward, due to an RNG mechanic. My suggestion: Put a corruption threshold before data loss occurs. Even 5% would do it. If you're running around caves super corrupted that's one thing, but if it's extremely minor corruption it'd be nice to avoid corruption effects that are in essence irreparable.
The one aspect of corruption that I think has some problems is data loss. The worst offender, maprot, has already thankfully been removed. Data loss as it remains essentially takes two basics forms. One, you forget what parts are. As a mechanic I think it works. After you play long enough, you know what parts enemies are dropping after a fight. What this mechanic really does is it forces you to make a decision when you encounter stockpiles. Maybe those unrecognized devices are advanced force fields, maybe they're field recycling units. You can strap one on to find out and restore the loss data but that will cost matter and, more importantly, time. It forces the player to make a strategic decision with food clock ramifications, so I think it works.
My big problem with data loss is that it can destroy data that's effectively irreplaceable. Some of the best rewards in the cave branches are the derelict logs. A log that gives you a branch entrance, zone layouts, or index(machine) for a future floor is huge. The problem is, it's so easy to lose this data, even in the face of extremely trivial corruption.
An anecdotal example: I was in the proximity caves on -3, looking for the exit to research. I found a log giving me the entrance to Testing on -3. That's a huge bit of data! I bumped into a programmer who gave me 2% corruption. I slapped on my system restoration module and made for the exit. In that interval I still managed to suffer two "data loss" corruption events. On -3, I did not get the intel reveal for testing, and had to find it the hard way.
Maybe that was exceptionally bad luck, but I feel like it's way too easy to lose data from derelict logs due to corruption. The whole point of branches is that they're high risk / high reward scenarios, and if the rewards can be that easily lost, they're much less desirable. I know that the derelict logs aren't the main rewards in each cave branch, but they're certainly big reasons why I go in in the first place.
While there are anti corruption devices in the game, I don't think that solves this particular problem this given how easily data loss can occur. A system restoration module can take hundreds of turns to get rid of corruption, and walking around with a corruption screen to block corruption effects 1) is not guaranteed to work, and 2) handicaps your combat abilities. They're certainly useful devices, but they're not going to prevent you from suffering the occasional data loss effect. As it stands, even with anti corruption utilities, as little as 1% corruption in caves can result in you forfeiting extremely valuable intel.
tldr: People are inherently loss adverse. They'd rather not lose something they already have than gain something of equivalent value. (See https://en.wikipedia.org/wiki/Loss_aversion). It's a bummer to lose a branch reward, even if a minor reward, due to an RNG mechanic. My suggestion: Put a corruption threshold before data loss occurs. Even 5% would do it. If you're running around caves super corrupted that's one thing, but if it's extremely minor corruption it'd be nice to avoid corruption effects that are in essence irreparable.
5
Ideas / Add a chance to have a derelict fabricator in Zion
« on: November 08, 2017, 08:14:47 AM »
(some spoilers follow)
I love the random mini rewards in Zion, and the more variety the better for immersion and replayability. One possible reward I would like to see is a machine that is a derelict take on the fabricator, allowing you to load schematics for fabrication but without any fail chance or risk of trace/lock out. The balancing factors would be 1) it would necessarily be on -6 or -7, so you won't have the best schematics; 2) give the machine a finite amount of matter so it can only fabricate so much; and 3) you wouldn't get it every game since it would not always spawn.
This unique reward would allow for the one place in the game where you could have risk-free successful fabbing, a nice perk for combat builds without hacking gear. In the event that Data Miner is -7 and Zion is -6, this reward would allow combat builds to put some of DM's rewards to use and would create further synergy between the branches (which again which be an infrequent turn of events, since it would depend upon the ordering of the branches, successfully reaching both, and having the fabricator randomly spawn in Zion).
There is already some precedent for risk free fabbing as a branch reward. Adding the chance to see this machine in Zion would give you a different take on that and could alter how you plan runs and how runs turn out if you get this reward. Finally there's little risk of this being too good of a reward, since you can just tweak how much matter the machine has and can therefore fab.
(Also if the title is too spoilery let me know and I'll change it).
I love the random mini rewards in Zion, and the more variety the better for immersion and replayability. One possible reward I would like to see is a machine that is a derelict take on the fabricator, allowing you to load schematics for fabrication but without any fail chance or risk of trace/lock out. The balancing factors would be 1) it would necessarily be on -6 or -7, so you won't have the best schematics; 2) give the machine a finite amount of matter so it can only fabricate so much; and 3) you wouldn't get it every game since it would not always spawn.
This unique reward would allow for the one place in the game where you could have risk-free successful fabbing, a nice perk for combat builds without hacking gear. In the event that Data Miner is -7 and Zion is -6, this reward would allow combat builds to put some of DM's rewards to use and would create further synergy between the branches (which again which be an infrequent turn of events, since it would depend upon the ordering of the branches, successfully reaching both, and having the fabricator randomly spawn in Zion).
There is already some precedent for risk free fabbing as a branch reward. Adding the chance to see this machine in Zion would give you a different take on that and could alter how you plan runs and how runs turn out if you get this reward. Finally there's little risk of this being too good of a reward, since you can just tweak how much matter the machine has and can therefore fab.
(Also if the title is too spoilery let me know and I'll change it).
6
General Discussion / Cogmind beta 3 [SPOILER] thread
« on: November 02, 2017, 06:42:16 PM »
HERE THERE BE SPOILERS
This is the thread for all questions and discussions about spoilery things in the Steam release. It goes without saying that this thread will be filled with spoilers so read at your own risk.
You've been warned!
Here's my spoilery question. In caves I found today an alien artifact called the Dimensional Node Initializer. There was a terminal nearby but it's log didn't give me any ideas as to what this thing does. I just decided to attach it, which resulted in it implanting in the ground and creating an orange tile. I couldn't pick the tile up and stepping on it had no visible effect.
My theory is that this artifact creates a little dimensional pocket, so that when you attach it and stand on the square you're invisible to enemies. Just a guess though. It'll be awhile before I find this guy again, making experimentation hard, so if anyone has insight into what this little guy does i'd love to know!
This is the thread for all questions and discussions about spoilery things in the Steam release. It goes without saying that this thread will be filled with spoilers so read at your own risk.
You've been warned!
Here's my spoilery question. In caves I found today an alien artifact called the Dimensional Node Initializer. There was a terminal nearby but it's log didn't give me any ideas as to what this thing does. I just decided to attach it, which resulted in it implanting in the ground and creating an orange tile. I couldn't pick the tile up and stepping on it had no visible effect.
My theory is that this artifact creates a little dimensional pocket, so that when you attach it and stand on the square you're invisible to enemies. Just a guess though. It'll be awhile before I find this guy again, making experimentation hard, so if anyone has insight into what this little guy does i'd love to know!
7
General Discussion / Congrats on the (imminent) Steam release!
« on: October 15, 2017, 11:17:50 AM »
Hi Kyzrati (and everyone else)! Congrats on being 1 day and 7 hours away from steam launch. I'm staring at the countdown on steam and wishing tomorrow was a Sunday instead of today so I could play all day.
I took a break from Cogmind because I didn't want to burn out. I played so much DCSS back in the day I can't go back to it and I didn't want that happening with this wonderful game. The Steam release seems like a perfect time to jump back in!
Any chance we'll see an official steam release tournament??
The alpha tournament was ludicrously fun.
I took a break from Cogmind because I didn't want to burn out. I played so much DCSS back in the day I can't go back to it and I didn't want that happening with this wonderful game. The Steam release seems like a perfect time to jump back in!
Any chance we'll see an official steam release tournament??
The alpha tournament was ludicrously fun.
8
General Discussion / Charting the NPCs and their factions [MAJOR SPOILERS]
« on: November 02, 2016, 10:44:57 AM »
Major spoilers ahead, proceed with caution.
Kyzrati has been posting a blog series about how to weave a plot into a roguelike/procedurally generated game. It's definitely worth a read.
Part 1: http://www.gridsagegames.com/blog/2016/10/weaving-narratives-procedural-worlds-part-1-value/
Part 2: http://www.gridsagegames.com/blog/2016/11/weaving-narratives-into-procedural-worlds-part-2-characteristics/
In part two, he posted a fascinating graph:
This graph depicts every major NPC in the game; where you can find them; their faction; and the impact they have on future events/other NPCs. Or as Kyrzrati put it, the "major NPC encounter visualization ... showing those with a direct effect on some later encounter (arrows), as well as those with a relatively significant long-term impact on gameplay (bracketed length)."
Let's try to decipher it! I haven't thoroughly explored Testing, Armory, or Quarantine so I'm not well versed in the plot events in there. People who have seen more of the end game will probably have a much better idea than me about who some of these dots could be.
First off, assuming I am not colorblind, we have 8 majors factions. Interesting and more than I expected.
The first one is the easiest. The red dot on the left is revision 17, since that's the only early game NPC. What about the other two red dots? I believe the middle is O7 and the third is revision 17 again in Archives. I'm not the biggest lore/branch junkie though so correct me if I'm wrong!
The lime green one is probably data miner. You encounter him early and he can potentially change the rest of the game. Fits the bill. Interestingly, if I'm right that means Data Miner is not partied with the other factions. No other lime green dots.
I believe orange is Zhirov's faction. The orange dot on the far left is probably O8, who can be found in Zion and gives you that hack to summon help when needed. Two of the central orange dots are Zhirov himself, representing that he can be found in archives or the caves. The far right orange is Zhirov again, since he can show up in access under certain conditions. If I'm right that Zhirov = orange, there's one orange dot I can't account for.
Note that one of the oranges links to black. If I recall correctly, Zhirov gives you a Main.C related hack... perhaps black is Main.C itself?
There's a free floating yellow dude. No idea who that is. Could be O7 if some of my assumptions are wrong.
Warlord is either light yellow or purple. I suspect purple, since that would allow one of the other purples to be God Mode. I have no idea who the other purples are and no idea who the light yellow is.
No idea who light blue is.
Finally, it's worth noting that the branch off of Zion is locked. I bet there's a major NPC in there that no one has seen yet.
Lots of robots still to meet (and blow up).
Kyzrati has been posting a blog series about how to weave a plot into a roguelike/procedurally generated game. It's definitely worth a read.
Part 1: http://www.gridsagegames.com/blog/2016/10/weaving-narratives-procedural-worlds-part-1-value/
Part 2: http://www.gridsagegames.com/blog/2016/11/weaving-narratives-into-procedural-worlds-part-2-characteristics/
In part two, he posted a fascinating graph:
This graph depicts every major NPC in the game; where you can find them; their faction; and the impact they have on future events/other NPCs. Or as Kyrzrati put it, the "major NPC encounter visualization ... showing those with a direct effect on some later encounter (arrows), as well as those with a relatively significant long-term impact on gameplay (bracketed length)."
Let's try to decipher it! I haven't thoroughly explored Testing, Armory, or Quarantine so I'm not well versed in the plot events in there. People who have seen more of the end game will probably have a much better idea than me about who some of these dots could be.
First off, assuming I am not colorblind, we have 8 majors factions. Interesting and more than I expected.
The first one is the easiest. The red dot on the left is revision 17, since that's the only early game NPC. What about the other two red dots? I believe the middle is O7 and the third is revision 17 again in Archives. I'm not the biggest lore/branch junkie though so correct me if I'm wrong!
The lime green one is probably data miner. You encounter him early and he can potentially change the rest of the game. Fits the bill. Interestingly, if I'm right that means Data Miner is not partied with the other factions. No other lime green dots.
I believe orange is Zhirov's faction. The orange dot on the far left is probably O8, who can be found in Zion and gives you that hack to summon help when needed. Two of the central orange dots are Zhirov himself, representing that he can be found in archives or the caves. The far right orange is Zhirov again, since he can show up in access under certain conditions. If I'm right that Zhirov = orange, there's one orange dot I can't account for.
Note that one of the oranges links to black. If I recall correctly, Zhirov gives you a Main.C related hack... perhaps black is Main.C itself?
There's a free floating yellow dude. No idea who that is. Could be O7 if some of my assumptions are wrong.
Warlord is either light yellow or purple. I suspect purple, since that would allow one of the other purples to be God Mode. I have no idea who the other purples are and no idea who the light yellow is.
No idea who light blue is.
Finally, it's worth noting that the branch off of Zion is locked. I bet there's a major NPC in there that no one has seen yet.
Lots of robots still to meet (and blow up).
9
Fixed Bugs & Non-Bugs / Core remaining value in scoresheet incorrect
« on: June 29, 2016, 05:45:09 PM »
Here's an excerpt from a recent score log of mine:
Average Core Remaining (%) 87
Depth 10 Exit 58
Depth 9 Exit 100
Depth 8 Exit 97
Depth 7 Exit 92
Depth 6 Exit 86
Depth 5 Exit 93
Depth 3 Exit 168
Depth 2 Exit 91
You'll notice there's no Depth 4 exit, and it says I had 168% core on Depth 3. I went into upper caves on -4 and got to -3 that way. I suspect that's the cause of the problem.
e: Yep, it's definitely caused by going into branches:
Average Core Remaining (%) 71
Depth 10 Exit 68
Depth 9 Exit 24
Depth 8 Exit 96
Depth 6 Exit 164
Depth 4 Exit 75
Average Core Remaining (%) 87
Depth 10 Exit 58
Depth 9 Exit 100
Depth 8 Exit 97
Depth 7 Exit 92
Depth 6 Exit 86
Depth 5 Exit 93
Depth 3 Exit 168
Depth 2 Exit 91
You'll notice there's no Depth 4 exit, and it says I had 168% core on Depth 3. I went into upper caves on -4 and got to -3 that way. I suspect that's the cause of the problem.
e: Yep, it's definitely caused by going into branches:
Average Core Remaining (%) 71
Depth 10 Exit 68
Depth 9 Exit 24
Depth 8 Exit 96
Depth 6 Exit 164
Depth 4 Exit 75
10
Stories / Near miss on -1
« on: June 29, 2016, 05:41:12 PM »
Had my most successful Alpha 9 run to date (24k score). It was extremely close to a win but what can you do. Some spoilers follow, read at your own risk.
My plan was to skip the lower caves and enter the upper caves on -4. I was hoping to get Warlord for the backup in Research. Instead I got Zhirov, but those benefits certainly are appreciated. The reason I skipped lower caves is that I typically come out of there pretty beaten up and I rarely find the trade offs worth it. I think it's not a coincidence that I did fairly well after avoiding lower caves for once. Until now I've been going into them every game.
I also changed up my build a bit by going with 6 propulsion slots instead of my usual 2. This let me use treads as armor and it was very effective. I attribute getting through Research to this.
Admittedly, I got lucky in Research and found the stairs quickly. By the time I got to Access, I had 3 Ragnarok launchers, but none of the really good defensive items (improved powered armor, advanced force fields, etc). With the launchers (plus a matter compressor) I was able to squish all enemies and really rip through Access. The security level spiked but it didn't matter since I could 2 shot whatever came out of the ARCs.
I went left when I should've gone right and found myself at a locked branch. Sigh. Very deflating I attribute this loss to the unfinished state of the game
After a ton of backtracking attrition finally won out and I ate it. What didn't help is that I saw a B-99, thought I had found the real exit and duked it out, only to find it was just a rando behemoth guarding a hall. GG Access, you are truly a meat grinder.
The gory details:
My plan was to skip the lower caves and enter the upper caves on -4. I was hoping to get Warlord for the backup in Research. Instead I got Zhirov, but those benefits certainly are appreciated. The reason I skipped lower caves is that I typically come out of there pretty beaten up and I rarely find the trade offs worth it. I think it's not a coincidence that I did fairly well after avoiding lower caves for once. Until now I've been going into them every game.
I also changed up my build a bit by going with 6 propulsion slots instead of my usual 2. This let me use treads as armor and it was very effective. I attribute getting through Research to this.
Admittedly, I got lucky in Research and found the stairs quickly. By the time I got to Access, I had 3 Ragnarok launchers, but none of the really good defensive items (improved powered armor, advanced force fields, etc). With the launchers (plus a matter compressor) I was able to squish all enemies and really rip through Access. The security level spiked but it didn't matter since I could 2 shot whatever came out of the ARCs.
I went left when I should've gone right and found myself at a locked branch. Sigh. Very deflating I attribute this loss to the unfinished state of the game
After a ton of backtracking attrition finally won out and I ate it. What didn't help is that I saw a B-99, thought I had found the real exit and duked it out, only to find it was just a rando behemoth guarding a hall. GG Access, you are truly a meat grinder. The gory details:
Spoiler (click to show/hide)
11
Ideas / Add a terminal to Waste
« on: December 07, 2015, 11:01:10 AM »
Spoilers ahead, beware.
There are numerous reasons to go to Waste. There are also numerous ways to survive Waste (speed, luck with certain pre-fabbed events, serious firepower). A combat cogmind who goes into waste without enough speed, without any of the more favorable events occurring, and without big enough guns is very likely to end up naked, or at the very least severely mauled.
I suggest adding one more way: a terminal close to where you land. Give it a single command to turn off the trash compactors, and give the hack a low success chance and the terminal a high lockout chance.
As precedent for this addition, I submit the following:
There are numerous reasons to go to Waste. There are also numerous ways to survive Waste (speed, luck with certain pre-fabbed events, serious firepower). A combat cogmind who goes into waste without enough speed, without any of the more favorable events occurring, and without big enough guns is very likely to end up naked, or at the very least severely mauled.
I suggest adding one more way: a terminal close to where you land. Give it a single command to turn off the trash compactors, and give the hack a low success chance and the terminal a high lockout chance.
As precedent for this addition, I submit the following:
Spoiler (click to show/hide)
12
Strategies / Combat Cogmind Strategy Series #2: The Alert Level
« on: November 22, 2015, 06:41:30 PM »
This is the second chapter in my series on strategy advice for combat-oriented cogminds. Here we’ll cover the Alert Level, which is probably the most important mechanic in the game. In fact, the first topic in my series is closely intertwined with this one. To recap, our main goal on each level is to find the stairs and get off the level as fast as possible. Why? Because we’re in a race against time. Over time, the threat level will inevitably rise. If it gets too high, you will die. Horribly. Learning how to avoid raising the alert level (and learning how to lower it once it does begin to climb) are vital skills.
1. For Every Action There Is A Reaction - The Global A.I.
It’s worthwhile comparing Cogmind to other roguelikes. In the vast majority of other RLs, all of your encounters and actions essentially exist in a vacuum. Consider DCSS. I can go into the Elven Halls and slaughter every single elf in the entire game. What effect does this have on the rest of the game? The answer is basically nothing.
Of course, by virtue of clearing this branch I’ll have more loot and more experience, although perhaps I had to burn some comestibles in the process. But my (genocidal) actions in Elf do not alter how the rest of the game looks like or reacts to me. I could have leveled up by clearing Elf; I could have leveled up by scumming the Abyss. The end result from either course of action is indistinguishable. My local actions do not have global consequences.
Cogmind does things very differently. There’s a Global AI that tracts your actions as you explore the world. Your actions will determine the game’s reactions to you. Kyzrati has touched upon this mechanic in a blog post here:
http://www.gridsagegames.com/blog/2015/04/living-dungeon/
This portion of the above post is worth quoting at length:
“We have … an actual overarching AI controlling the community’s reaction to your presence and actions on a larger scale. You not only have to think about your interactions (combat or otherwise) on an individual robot-to-robot level, but in many cases must also consider the repercussions of your decisions further down the road.
Depending on the circumstances, your unauthorized or hostile actions will be reported, and you will be hunted, or cause enough mayhem and invite a robot army to converge on your position.”
How does this work in practice?
2. The Alert Level.
As you play the game, it keeps track of something called the Alert Level. There’s 6 different possible Alert Levels: low security, and Alert Levels 1-5. The game starts at low security. As you do certain actions (usually of the “blowing things up” variety), the Alert level rises. Here is an example from one of my winning runs of what the Alert Level looked like throughout the game:
Maximum Alert Level 2
Low Security (%) 71
Level 1 27
Level 2 1
Level 3 0
Level 4 0
Level 5 0
As you can see, I spent almost three quarters of the game at low security level; never got higher than Level 2; and only spent a mere 1% of the game at Level 2. Keeping the Alert Level low like this is vital to winning.
Easier said than done. As a combat Cogmind, you’re going to be blowing a lot of stuff up and raising the threat level. What does an increase in the threat level result in?
As the alert level rises, enemy squads get sent against you. These are teams of grunts supported by melee bots, protectors, or saboteurs. They also tend to be more advanced bots than what you’d normally find on the level. They’re packed into an ARC, hone in on you wherever you are, and deploy when they spot you. If you’re unprepared, these squads can wreck you.
Around level 4 and higher the global AI will start deploying even worse squads against you. These include powerful and rare robots that you otherwise will not see, including bots armed with explosive weapons (think missile launchers). Once these things begin dropping on your end, it’s effectively game over. You want to avoid things getting to this point at all costs.
What makes it hard to avoid getting to this point is the Alert Level death spiral. Let’s say you’re so buff you’re capable of swatting down an extermination squad. Well, that just raises the threat level even more! You kill stuff, the threat level goes up, more stuff gets sent out against you, you kill more stuff, etc. It’s unsustainable. To avoid this death spiral you need to know what raises the alert level; how to avoid needlessly raising it; and how to lower it once it becomes to climb up.
3. Things That Directly Raise the Alert Level
Killing Stuff
This one is pretty obvious and it’s the most common way you’re going to raise the Alert Level. Every robot that you (or your allies) kill raises the threat level. The global AI understandably gets concerned when its little minions start exploding.
However, not all robots are created equal. All robots in the game (including you) have a rating. (Your scoresheet will let you know what your peak state was and what your rating was at this state). The higher the robot’s rating, the greater the alert level goes up when you kill it. In practice, this means that you can kill the occasional non-hostile robots like serfs and recyclers without much repercussion. Kill a Behemoth and you’ll really get their attention.
This mechanic is how Cogmind gets its food clock. Every so many turns Programmers are dispatched against you. Even if you can kill them easily and manage electromagnetic corruption, you can’t stay on a Factory or Research level indefinitely since eventually the threat level will get too high.
In the past, causing a robot to meltdown or corrupting it to death would not increase the threat level. As of Alpha 5, this is no longer the case – you’ll now get charged for those deaths security-wise.
Blowing Parts of the Complex Up
As you explore the complex, you’ll find machines that you can hack (e.g. terminals) and machines that you can’t interact with (e.g. neutrino reactors). All of these can be blown up. Some of the non-interactive machines will destabilize and explode when shot at. These large explosions are more than capable of killing enemy robots and setting off chain reactions. Unsurprisingly, nuking large swathes of machinery will get you into trouble real quick.
It’s worth noting that blowing up walls and doors does not increase the Alert level. Terraform to your robotic heart’s content.
Allies
Allies raise the alert level in two different ways. First, you’re on the hook for any enemy robot they kill. Since large groups of allies are very effective in taking out enemies (for a while at least), they’ll be jacking the alert level up as they go along. Also, having lots of guys around you increases your footprint and makes it more likely than enemy patrol will spot you and your motley crew. You’ll be getting into more fights with lots of bots around you than you would if you going solo.
More importantly, using allies will raise the alert level higher than what it would be without them. The alert level goes up by simply having allies around you. This is understandable: the global AI is going to react differently to a killbot leading a large army as opposed to a single solitary killbot. If you keep lots of allies around, you’ll quickly see the alert level shoot up. Because allies can effectively take out enemy squads (at first), it is very easy to enter an Alert Level death spiral through overuse of allies.
Garrisons (The Map)
Entering a Garrison will almost definitely raise the alert level. They’re full of enemies, so lots of killing is inevitable. Even worse, every now and then you’ll get a message along the lines of “Garrison compromised.” Every time you hear this, the alert level has gone up. On the bright side, you won’t face any programmer dispatches as long as you’re in the Garrison.
Garrisons also contain various Garrison-only machines (relays and phase generators). Blowing them up will raise the threat level even more than blowing up normal machinery.
Trojan Hacks
With Alpha 4, there are new hacks called Trojan that you can pull off. Even though this malicious little hacks wreak havoc on the enemy, they do not raise the security level. Trojan away to your hearts content!
One Trojan (Trojan(Detonate)) causes machines to become proximity mines and blow up when enemies are near. It is currently unclear whether you're on the hook for machines and enemies blown up by these detonations.
It is currently unclear whether Brute Force hacks raise the threat level.
4. Things That Indirectly Raise The Threat Level
Machine Lockouts
If you repeatedly try to use any machine for too long, you’ll eventually be detected and locked out of the machine. This will cause an investigation squad to be dispatched to see what’s causing the trouble. Although I don’t believe lock outs by themselves raise the threat level, more enemies for you to blow up means a rising security level.
Brute Force Hacks
Alpha 4 introduced a new hack type called Brute Force Hacks. These hacks get a machine to do something at the cost of overloading the machine and rendering it otherwise inoperable. I don’t know if this raises the security level, but like a lockout it will definitely cause an investigation squad to get dispatched.
Watchers, Haulers, Alarm traps and Operators
Watchers will emit a distress signal if they see you and a combat enemy is in range. Haulers will emit distress signals when attacked, and higher level ones will call in reinforcements if attacked. Operators will use terminals to call in reinforcements if any enemies are nearby. Alarm traps will dispatch an investigation squad to your position. None of these actions will raise the threat level by itself, but again the more you kill the higher the level goes.
Garrisons
If you fight near a Garrison machine, enemies will periodically pop out of it as combat enemies report your presence. Under no circumstances should you fight near a Garrison if avoidable.
5. How to Reduce and Manage The Threat Level
Hack The Security Level Down
There are several hacks that can help you manage the threat level. The most important one (and arguably the most important hack in the game) is Alert(Purge). Every time you pull this hack off, the alert level goes down. I try to do this at least once a level if possible. You cannot do this hack indefinitely on a level. Each time you pull this hack off, the chances of pulling it off successfully decreases. This makes camping forever on a Factory or Research level impossible.
Alert(Check) is a fairly easy hack to pull off that lets you know what the current alert level is. Once you’ve played long enough you’ll get a fairly good feeling where things stand at any given point, but this hack can be useful if you’re not sure how pissed they are.
Finally, there’s Recall(Assault), Recall(Extermination) and Recall(Reinforcements). These hacks are hard to pull off, but if it works it’ll cause a dispatched squad to leave the map and spare you the hassle of fighting them.
Get Off The Level
Going to the next level causes the alert level to drop. This, combined with programmer dispatches and assault squads, is the reason why our primary goal is to get off the level ASAP.
Don’t Get Locked Out Of Every Terminal You See
When I first started playing Cogmind I would hack away at every terminal I came across until lock out. Don’t do this. As noted, this will cause an investigation squad to be dispatched. Unless you’re a dedicated hacker, use level one terminals for your indirect hacking needs, use terminal 2 primarily for direct hacks, and don’t bother with security level 3 (this will probably change around Alpha 5).
Use Transmission Jammers
Watchers and enemy combat unites can send enemies to your position; watchers have a greater range then enemy combat bots. If you have a transmission jammer active, no one can hear them scream. Limiting how many enemies attack you will necessarily limit how high the alert level goes. This method will also stop Garrisons from squirting out enemies if you're fighting near them. If you have to fight near a Garrison, attach a jammer if at all possible
Jammers can also be used to jam alarm traps, which means one less squad for you to fight.
Jammers do not stop Haulers or programmers from calling in reinforcements.
Use Sensors and Optic Arrays
One of the best things you can do in Cogmind is avoid a fight. Since we don’t get experience for enemies, almost always take damage in combat, and typically find our best items from caches, fighting is usually a net loss experience (without even taking the increase to the alert level into account). Sensors and optic utilities can help you see enemies before you see them. Every needless fight you skip indirectly keeps the threat level down.
Don’t Kill The Survivors
If you blow the last weapon of a combat bot, it’ll usually flee (and look for a non combat bot to graft a weapon on). I usually let survivors go. They pose little risk to you at this point, and sparing them keeps the alert level low.
Garrisons
Somewhat counterintuitively, Garrisons can play a role in threat level management. If you go through a Garrison and end up on the same map as where you started (e.g. enter one at -6, and after leaving the Garrison find yourself back at -6), all enemy squads on the level are reduced by one member. Facing smaller squads in late Factory and Research not only means lower increases in the threat level, it also means you’re less likely to die.
Let Them Blow You Up A Bit
This is also counter intuitive, but the game keeps track of how many parts you lose from combat. Getting parts blown off can actually reduce the threat level. This represents the AI concluding it’s doing a decent job taking you out and adjusting its response accordingly.
To illustrate, if you have some armor and some spare armor in your inventory, consider letting the 1st armor get entirely destroyed rather than swapping it out for a fresh one before it reaches that point. Item attrition can make the enemy go easier on you.
Allies Getting Blown Up
The security level goes down when your allies die. This helps mitigate some of the downsides of using allies, but probably can’t be realistically used as an independent strategy keep the threat level down.
Super Secret Tech
Finally, there is apparently a super secret way to set the security level to zero (at a cost). As far as I know, no player has figured it out yet. I suspect the answer is in Waste but that's just a hunch. I’ll add the method to this list with the appropriate spoiler tags once we know.
As we learn new things that cause the alert level to rise and fall, I'll update the guide.
1. For Every Action There Is A Reaction - The Global A.I.
It’s worthwhile comparing Cogmind to other roguelikes. In the vast majority of other RLs, all of your encounters and actions essentially exist in a vacuum. Consider DCSS. I can go into the Elven Halls and slaughter every single elf in the entire game. What effect does this have on the rest of the game? The answer is basically nothing.
Of course, by virtue of clearing this branch I’ll have more loot and more experience, although perhaps I had to burn some comestibles in the process. But my (genocidal) actions in Elf do not alter how the rest of the game looks like or reacts to me. I could have leveled up by clearing Elf; I could have leveled up by scumming the Abyss. The end result from either course of action is indistinguishable. My local actions do not have global consequences.
Cogmind does things very differently. There’s a Global AI that tracts your actions as you explore the world. Your actions will determine the game’s reactions to you. Kyzrati has touched upon this mechanic in a blog post here:
http://www.gridsagegames.com/blog/2015/04/living-dungeon/
This portion of the above post is worth quoting at length:
“We have … an actual overarching AI controlling the community’s reaction to your presence and actions on a larger scale. You not only have to think about your interactions (combat or otherwise) on an individual robot-to-robot level, but in many cases must also consider the repercussions of your decisions further down the road.
Depending on the circumstances, your unauthorized or hostile actions will be reported, and you will be hunted, or cause enough mayhem and invite a robot army to converge on your position.”
How does this work in practice?
2. The Alert Level.
As you play the game, it keeps track of something called the Alert Level. There’s 6 different possible Alert Levels: low security, and Alert Levels 1-5. The game starts at low security. As you do certain actions (usually of the “blowing things up” variety), the Alert level rises. Here is an example from one of my winning runs of what the Alert Level looked like throughout the game:
Maximum Alert Level 2
Low Security (%) 71
Level 1 27
Level 2 1
Level 3 0
Level 4 0
Level 5 0
As you can see, I spent almost three quarters of the game at low security level; never got higher than Level 2; and only spent a mere 1% of the game at Level 2. Keeping the Alert Level low like this is vital to winning.
Easier said than done. As a combat Cogmind, you’re going to be blowing a lot of stuff up and raising the threat level. What does an increase in the threat level result in?
As the alert level rises, enemy squads get sent against you. These are teams of grunts supported by melee bots, protectors, or saboteurs. They also tend to be more advanced bots than what you’d normally find on the level. They’re packed into an ARC, hone in on you wherever you are, and deploy when they spot you. If you’re unprepared, these squads can wreck you.
Around level 4 and higher the global AI will start deploying even worse squads against you. These include powerful and rare robots that you otherwise will not see, including bots armed with explosive weapons (think missile launchers). Once these things begin dropping on your end, it’s effectively game over. You want to avoid things getting to this point at all costs.
What makes it hard to avoid getting to this point is the Alert Level death spiral. Let’s say you’re so buff you’re capable of swatting down an extermination squad. Well, that just raises the threat level even more! You kill stuff, the threat level goes up, more stuff gets sent out against you, you kill more stuff, etc. It’s unsustainable. To avoid this death spiral you need to know what raises the alert level; how to avoid needlessly raising it; and how to lower it once it becomes to climb up.
3. Things That Directly Raise the Alert Level
Killing Stuff
This one is pretty obvious and it’s the most common way you’re going to raise the Alert Level. Every robot that you (or your allies) kill raises the threat level. The global AI understandably gets concerned when its little minions start exploding.
However, not all robots are created equal. All robots in the game (including you) have a rating. (Your scoresheet will let you know what your peak state was and what your rating was at this state). The higher the robot’s rating, the greater the alert level goes up when you kill it. In practice, this means that you can kill the occasional non-hostile robots like serfs and recyclers without much repercussion. Kill a Behemoth and you’ll really get their attention.
This mechanic is how Cogmind gets its food clock. Every so many turns Programmers are dispatched against you. Even if you can kill them easily and manage electromagnetic corruption, you can’t stay on a Factory or Research level indefinitely since eventually the threat level will get too high.
In the past, causing a robot to meltdown or corrupting it to death would not increase the threat level. As of Alpha 5, this is no longer the case – you’ll now get charged for those deaths security-wise.
Blowing Parts of the Complex Up
As you explore the complex, you’ll find machines that you can hack (e.g. terminals) and machines that you can’t interact with (e.g. neutrino reactors). All of these can be blown up. Some of the non-interactive machines will destabilize and explode when shot at. These large explosions are more than capable of killing enemy robots and setting off chain reactions. Unsurprisingly, nuking large swathes of machinery will get you into trouble real quick.
It’s worth noting that blowing up walls and doors does not increase the Alert level. Terraform to your robotic heart’s content.
Allies
Allies raise the alert level in two different ways. First, you’re on the hook for any enemy robot they kill. Since large groups of allies are very effective in taking out enemies (for a while at least), they’ll be jacking the alert level up as they go along. Also, having lots of guys around you increases your footprint and makes it more likely than enemy patrol will spot you and your motley crew. You’ll be getting into more fights with lots of bots around you than you would if you going solo.
More importantly, using allies will raise the alert level higher than what it would be without them. The alert level goes up by simply having allies around you. This is understandable: the global AI is going to react differently to a killbot leading a large army as opposed to a single solitary killbot. If you keep lots of allies around, you’ll quickly see the alert level shoot up. Because allies can effectively take out enemy squads (at first), it is very easy to enter an Alert Level death spiral through overuse of allies.
Garrisons (The Map)
Entering a Garrison will almost definitely raise the alert level. They’re full of enemies, so lots of killing is inevitable. Even worse, every now and then you’ll get a message along the lines of “Garrison compromised.” Every time you hear this, the alert level has gone up. On the bright side, you won’t face any programmer dispatches as long as you’re in the Garrison.
Garrisons also contain various Garrison-only machines (relays and phase generators). Blowing them up will raise the threat level even more than blowing up normal machinery.
Trojan Hacks
With Alpha 4, there are new hacks called Trojan that you can pull off. Even though this malicious little hacks wreak havoc on the enemy, they do not raise the security level. Trojan away to your hearts content!
One Trojan (Trojan(Detonate)) causes machines to become proximity mines and blow up when enemies are near. It is currently unclear whether you're on the hook for machines and enemies blown up by these detonations.
It is currently unclear whether Brute Force hacks raise the threat level.
4. Things That Indirectly Raise The Threat Level
Machine Lockouts
If you repeatedly try to use any machine for too long, you’ll eventually be detected and locked out of the machine. This will cause an investigation squad to be dispatched to see what’s causing the trouble. Although I don’t believe lock outs by themselves raise the threat level, more enemies for you to blow up means a rising security level.
Brute Force Hacks
Alpha 4 introduced a new hack type called Brute Force Hacks. These hacks get a machine to do something at the cost of overloading the machine and rendering it otherwise inoperable. I don’t know if this raises the security level, but like a lockout it will definitely cause an investigation squad to get dispatched.
Watchers, Haulers, Alarm traps and Operators
Watchers will emit a distress signal if they see you and a combat enemy is in range. Haulers will emit distress signals when attacked, and higher level ones will call in reinforcements if attacked. Operators will use terminals to call in reinforcements if any enemies are nearby. Alarm traps will dispatch an investigation squad to your position. None of these actions will raise the threat level by itself, but again the more you kill the higher the level goes.
Garrisons
If you fight near a Garrison machine, enemies will periodically pop out of it as combat enemies report your presence. Under no circumstances should you fight near a Garrison if avoidable.
5. How to Reduce and Manage The Threat Level
Hack The Security Level Down
There are several hacks that can help you manage the threat level. The most important one (and arguably the most important hack in the game) is Alert(Purge). Every time you pull this hack off, the alert level goes down. I try to do this at least once a level if possible. You cannot do this hack indefinitely on a level. Each time you pull this hack off, the chances of pulling it off successfully decreases. This makes camping forever on a Factory or Research level impossible.
Alert(Check) is a fairly easy hack to pull off that lets you know what the current alert level is. Once you’ve played long enough you’ll get a fairly good feeling where things stand at any given point, but this hack can be useful if you’re not sure how pissed they are.
Finally, there’s Recall(Assault), Recall(Extermination) and Recall(Reinforcements). These hacks are hard to pull off, but if it works it’ll cause a dispatched squad to leave the map and spare you the hassle of fighting them.
Get Off The Level
Going to the next level causes the alert level to drop. This, combined with programmer dispatches and assault squads, is the reason why our primary goal is to get off the level ASAP.
Don’t Get Locked Out Of Every Terminal You See
When I first started playing Cogmind I would hack away at every terminal I came across until lock out. Don’t do this. As noted, this will cause an investigation squad to be dispatched. Unless you’re a dedicated hacker, use level one terminals for your indirect hacking needs, use terminal 2 primarily for direct hacks, and don’t bother with security level 3 (this will probably change around Alpha 5).
Use Transmission Jammers
Watchers and enemy combat unites can send enemies to your position; watchers have a greater range then enemy combat bots. If you have a transmission jammer active, no one can hear them scream. Limiting how many enemies attack you will necessarily limit how high the alert level goes. This method will also stop Garrisons from squirting out enemies if you're fighting near them. If you have to fight near a Garrison, attach a jammer if at all possible
Jammers can also be used to jam alarm traps, which means one less squad for you to fight.
Jammers do not stop Haulers or programmers from calling in reinforcements.
Use Sensors and Optic Arrays
One of the best things you can do in Cogmind is avoid a fight. Since we don’t get experience for enemies, almost always take damage in combat, and typically find our best items from caches, fighting is usually a net loss experience (without even taking the increase to the alert level into account). Sensors and optic utilities can help you see enemies before you see them. Every needless fight you skip indirectly keeps the threat level down.
Don’t Kill The Survivors
If you blow the last weapon of a combat bot, it’ll usually flee (and look for a non combat bot to graft a weapon on). I usually let survivors go. They pose little risk to you at this point, and sparing them keeps the alert level low.
Garrisons
Somewhat counterintuitively, Garrisons can play a role in threat level management. If you go through a Garrison and end up on the same map as where you started (e.g. enter one at -6, and after leaving the Garrison find yourself back at -6), all enemy squads on the level are reduced by one member. Facing smaller squads in late Factory and Research not only means lower increases in the threat level, it also means you’re less likely to die.
Let Them Blow You Up A Bit
This is also counter intuitive, but the game keeps track of how many parts you lose from combat. Getting parts blown off can actually reduce the threat level. This represents the AI concluding it’s doing a decent job taking you out and adjusting its response accordingly.
To illustrate, if you have some armor and some spare armor in your inventory, consider letting the 1st armor get entirely destroyed rather than swapping it out for a fresh one before it reaches that point. Item attrition can make the enemy go easier on you.
Allies Getting Blown Up
The security level goes down when your allies die. This helps mitigate some of the downsides of using allies, but probably can’t be realistically used as an independent strategy keep the threat level down.
Super Secret Tech
Finally, there is apparently a super secret way to set the security level to zero (at a cost). As far as I know, no player has figured it out yet. I suspect the answer is in Waste but that's just a hunch. I’ll add the method to this list with the appropriate spoiler tags once we know.
As we learn new things that cause the alert level to rise and fall, I'll update the guide.
13
Ideas / An option to prompt Y/N before taking the stairs
« on: November 03, 2015, 06:37:53 AM »
Currently, the moment you touch the stairs you go the next level. Usually, that's exactly what you want. Sometimes however you might want to spend a bit more time on the level. A couple of times, usually due to combat situations, I have inadvertently stepped on a stair tile when I didn't want to leave the level just yet. It might be worthwhile to add a toggleable option to give a Y/N prompt before leaving a level. This is consistent with most other roguelikes, where stepping on an exit tile isn't enough; you usually have to hit a command to leave the level (even in games with no backtracking like DoomRL). Obviously, this is super low priority.
14
Strategies / Combat Cogmind Strategy Series #1: Finding the Stairs
« on: September 23, 2015, 10:56:19 AM »
This is the first installment in a series of various posts I am going to write up providing strategy advice for combat Cogmind runs. The goal of the series is to cover all of the basic issues in a combat run to be eventually assembled into a single strategy guide. This will help new players (who don’t mind being spoiled about mechanics) dive right into the game once the game is officially released.
I’m doing it in installments primarily because I’m still new to the game myself, even with a combat win under the belt. It’s also easier approaching a comprehensive guide if I break it into chunks. Please feel free to point out things I am missing or wrong about in the comments. This will help ensure that the final guide is as accurate and as thorough as possible.
1. The Main Goal.
When you enter a floor for the first time there is always one main goal, followed by various subgoals of secondary importance. The subgoals will vary from floor to floor and will vary based on your playstyle. Some people might want to make sure they have a schematic for a high capacity storage unit before leaving Materials. Other people might want to make sure they have a force field or powered armor before entering Research. These are all perfectly fine subgoals and reflect the wealth of options Cogmind gives you.
By definition, however, subgoals are secondary to the main goal. On every single level of this game there is always one main goal that takes precedence over everything else: find the stairs and get off the level as soon as possible.
The first half of that sentence is pretty self-evident. After all, the goal of the game is to escape. The second portion of that sentence is what you should focus on. The longer you linger on a floor, especially after Materials, the greater your chances of death. Time is of the essence.
A brief tangential discussion about the AI death spiral is now necessary, which will be covered in greater depth in the “Security Level” chapter. Briefly, there is a global security level in this game. The higher it goes, the greater the resistance the game will send after you. You always want the security level to be as low as possible. Killing enemies, however, raises the security level. This can and will lead to nasty feedback loops.
Once you get to factory, there is a steady stream of programmers sent after you periodically – the “food clock” of the game. Even if you’re capable of swatting them down, the mere act of killing them raises the threat level. If you dillydally, you’ll soon have extermination squads raining down upon you. You can’t dillydally. Find the stairs and get off the level as soon as possible.
This problem is magnified by combat builds. You’re designed to kill, not to run or hide, so you’ll necessarily be raising the security level. You’ll be using treads or legs, so you’ll be slow. The more time it takes you to get from point A to point B means the more programmer dispatches you’ll face. These two facts combined means that blindly wandering around is a great way to die. You need a plan to reliably find the stairs.
Below are various stair finding techniques. Mixing and matching these tricks based on the utilities you find is key to getting through Factory and Research quickly and efficiently.
2. Map sense
One way to find exits without the help of any utilities or hacks is simply acquiring what I call “map sense.” As you play the game more, you’ll see that there’s a pattern to where exits are. Generally speaking, exits tend to be in hallways, not in rooms; they tend to be along the perimeters of the maps, not towards the middle; and they are often in corner areas. These are generalizations. As a rule of thumb, however, traveling down hallways as opposed to in and out of rooms will help you find the exit faster.
It’s worthwhile traveling down these corridors hugging walls as opposed to in the center of the corridor. There are two reasons for this. One, as you pass doors they will open and you can peak in to see if there are any worthwhile items. Second, enemy patrols tend to move along corridors (most likely because these are the paths you have to take to get to the exits). If you’re hugging the wall, it’ll be easier to move into rooms to assume more favorable firing positions.
I’ll be expanding this section later with screenshots of various sample exits in Materials, Factory, and Research.
3. Hacking
Possibly the best way to find the exit. I’m not going to go into a full blown explanation of how hacking works, which will have to wait for its own chapter, but I’ll just touch upon a few preliminary things.
There are two types of hacks at a terminal: direct and indirect hacks. A direct hack happens when you select from the list of options offered by the terminal. The perk of this type of hack is you know your success chance. The downside is that you’re limited to what the terminal offers. An indirect hack occurs when you hit z and then type in the hack you want manually. The perk of this type of hack is that you can try whatever hack you want at any terminal. The downside is that is has a -15% penalty to success, and you can’t see your success rate.
Hacks that help you find the exit
The crowning achievement of hacks is Access(Main). A successful Access(Main) hack will show you all exits on the map. You can hold alt and scroll around with the arrow keys to see where the exits are in relation to you once you pull this hack off. Unsurprisingly, it is one of the hardest hacks and has a very low chance of success. Indirect Access(Main) hacks will often fail. If you see a terminal offering this choice (Level Access Points) and the success rate is above 15%, go for it. Even better if you have hacking suites.
I would not indirect hack Access(Main) unless I had lots of deep network scanners (which increase the success chances of indirect hacks), or if the Alert level was low. If the alert level is not low, you’re better off doing an Alert(Purge).
All of the above applies to Access(Branch), which shows you where the branch exits are, except that this hack is usually a bit easier to pull off. Currently not very useful with so many braches locked, except for Storage, but will become very useful indeed once the branches are added.
Another worthwhile hack in your quest for the exit: Index(Terminals) shows you all the terminals on the level. Knowing where they are gives you more opportunities to do direct or indirect Access(Main) hacks.
4. Drone Bays
Drone bays are one of my favorite ways to find the stairs. A drone bay is a 2-slot utility (the fact that they are 2 slots is a good hint that they’re a great item). When you equip it and activate it, it’ll deploy one drone each turn. The bay holds a maximum of 2 drones, so in 2 turns it will have emptied its payload. If you want to only deploy one drone, turn it off or put it back in your inventory after it has deployed the first drone.
Drones are little robots that move very quickly. Although they do not have armor, enemies have a -30% chance to hit them. Coupled with their fast speed, which means they won’t be in an enemies line of sight for long, they can have surprisingly long life spans.
Using Drones
To use a deployed drone, hold shift and right click on it, then tell it to explore. This will cause the drone to zip around the map using an exploration algorithm. You share vision with the drone so you’ll see what it discovers. If you hold the alt key you can scroll around the map. It’s good to periodically do this to check on your drone and to see what they’ve discovered. Hit Enter to recenter the map on your cogmind.
In the best case scenario, your intrepid drone will go far and discover the exit. Even if it doesn’t, it can reveal some portion of the map for you, find item caches, and give you an idea where enemy patrols are. If you know a room is empty or contains items you don’t need, you can bypass that room entirely. This furthers our goal of getting off the level as soon as possible.
Never send two drones out to explore at the same time. They’ll follow the same exploration path, which is a waste of a drone. Either deploy one drone and keep the other one in the bay, or wait a few turns before telling the second to explore. This will maximize how much you get out of each drone.
When a drone finds the exit, you’ll get a beeping noise that lets you know, just as if you had found the stairs yourself. When this happens, command your drone to return. If you have your drone bay equipped, you can scoop up a returning drone, allowing you to use it on the next level. For this reason, don’t discard an empty drone bay immediately after discarding it. I would only discard a drone bay if both drones associated with it were dead (and I either already had a drone bay schematic or did not intend to do any fabricating – see below).
It’s typically best to not hoard drone bays for too long, especially since they take up so much space. If I find drone bays, I’ll deploy one immediately and one each time I get to a new level. A real benefit (exploration now) is often better than a hypothetical benefit (saving them for later in the hopes that you can speed through Research). Getting to Research in strong shape is better than limping into Research with some drones.
Getting Drones
There are 3 ways to get drones. The first is to just randomly find a cache. Always a great moment, but luck dependent and nothing you can rely on. You can consider doing the “Index(Components)” hack at a terminal in Factory to see if there are any nearby drone caches. However, this is probably not the best use of a hack. If you do find a drone cache, I’d immediately deploy some, and save the rest for the next immediate levels.
The second way to get drones is to visit Storage, a branch off of Materials or early Factory. There is a very high chance of finding a drone cache in Storage. This is one level where the Index(Components) hack is a very good idea. If you find the drone cache, deploy some to see what else Storage has, and save as many as you to help get your through Factory
The third way to get drones is to fabricate them. To do this, you first need their schematic and then a fabricator. I would not bother trying to get the schematic through indirect hacks, since it’s a rating 5 item. Instead, if you find a drone bay bring it by a scanalyzer. That’s a much easier way to get the schematic than hacking. For this reason, I’ll keep even an empty drone bay around until I’ve gotten it scanned.
Once you have the schematic, the drone bay costs 40 material to fabricate. Since it’s a rating 5 item, having a hacking suite handy can help make sure you pull it off (hacking suites increase the chances of all direct hacks, and scanning/fabbing count as direct hacks).
5. Programmers (and Extermination Squads)
Ah programmers, the bane of your cogmind’s existence. They’ll corrupt you, turn your allies against you, and they won’t stop coming. They’ll also help you find the stairs if you know what you’re doing.
Once you get to Factory, every now and then you’ll get a message that says “rogue bot found, programmers report to [your location].” This message means that a certain number of programmers have been dispatched against you. They automatically know where you are and they’ll make a beeline to you until they find you.
They also spawn from the stairs. Therefore, if you know which direction they’re coming from, you know that there’s a stair case in that direction. This is extremely useful information.
When you know a programming team is en route, you want to prepare accordingly. You don’t want to be in the middle of a large hallway. If a group of hunters or grunts joins the party when the programmers arrive, things can go south quickly. The best place to fight them is in narrow corridors or in rooms with only 1 entrance.
If you’re in a corridor, pay attention to where they came from when they show up. Once you kill them, head in that direction. Rather than wait inside a room (which can prevent you from knowing where they came from), I sometimes like to wait right outside of rooms, popping into it once they show up. This lets me get the benefit of the room while letting me know where they came from. If you’re doing this, consider using an optical array so you can spot them before they get into firing range.
Besides giving you hints about where the exit is, programmers can also give information about the distance to it. Once you see the programmer dispatch warning, take note of how many turns pass until they get to you. If it’s very quick, that’s a strong hint that there’s an exit close by.
Extermination squads also come from the stairs, so take note of where they come from for the same reason. Theoretically, if you were by a terminal right when an extermination squad was dispatched, Enumerate(Exterminations) could possibly help you find the stairs. However, if the security level is high enough for these squads to be coming down, you should probably be purging the threat level rather than anything else at a terminal.
Alpha 4 (Garrisons) Update: The above tactic has been heavily, though not totally, nerfed in Alpha 4. Alpha 4 saw the addition of new machines called Garrisons. Enemies dispatched to the map (programmer dispatches, investigation/reinforcement/extermination/assault squads) have a chance to be deployed from Garrison access points rather than the stairs. Accordingly, if you play follow the programmer now, there's a very decent chance you'll find a Garrison rather than the stairs.
Enemies still have a chance of spawning from the stairs all the same, so this approach can work if desperate. Also, if you blow up garrisons as you find them you'll increase the chances that enemy spawns come from the stairs. On the other hand, this will raise the security level (though there will be a corresponding decrease in the frequency of programmer dispatches).
Besides blowing up garrisons, another possible approach to making this strategy viable again is to pull off an Index(Garrisons) hack. If you know where every Garrison is, this can help inform you whether a recently dispatched enemy came from a garrison or came from the stairs.
6. Sentries
Sentries in hallways are very often sitting on top of stairs. This fact is less important in Factory, which has pretty straightforward layouts. It becomes more important in Research, which is a more of a maze. Knowing where Sentries are can provide you with possible places to check out in your quest for the stairs.
There are two ways to search for Sentries beyond of your line of sight. First, you could use an advanced sensor array coupled with a signal interpreter. The sensory array lets you know where robots are well outside of your line of sight. The signal interpreter would let you know which of these robots are sentries. Spotting sentries could give you hints about where to go. Kyzrait made some gifs demonstrating this (along with an explanation of terrain scanners) in an informative blog post that can be found here:
http://www.gridsagegames.com/blog/2014/11/information-warfare/
You can either have these sensors on at all times or occasionally swap to them. Although a combat cogmind might be loathe to fill 2 utility slots with sensors, which would be useless in a battle, the sensors would enable you to see enemies coming so you could either avoid them, or swamp to more combat related utilities before battle begins. If that’s not your style, you could periodically switch to the sensors and signal array, and put them right back after sending out a quick scan to check for Sentries.
The second way to know where Sentries are lurking is the “Enumerate(Security)” hack. This hack will show you Sentries, but only those in the local area. I would not do this hack if the Alert level is high, since you’d be better of with an “Alert(Purge).” However, if I were on Research with a low security level and no hacking utilities, I’d consider it. An attempt at Access(Main) will almost certainly not work absent significant utilities or luck. A successful “Enumerate(Security)” is less helpful, but it’s better than nothing.
7. Terrain Scanners
Terrain scanners are interesting little items. When activated, over time they reveal walls, doors, and stairs in a certain area. By themselves, this takes awhile. With a terrain scan processor, however, they scan far more quickly. With a nice improved terrain scanner and improved scan processor, you can uncover the map and even find stairs before they enter your field of vision. Very handy stuff.
The more basic terrain scanners have their time and place but aren't much to write home about. Once you get to Long Ranged Terrain Scanners, however, you have one hell of an item indeed. With this scanners range, you can know most of the surrounding area in no time provided you have a good processor. In Factory and Research, this is a great way to find stairs. I consider long ranged or experimental terrain scanners to be one of the best utilities in the game and highly recommend you pocket one if you find it.
However, unlike a sensor array, a terrain scanner won’t let you know if enemies are coming. If you bump into some enemies with your terrain scanner stuff on, you’ll have to waste 2 turns to get to peak combat efficiency. This is a draw back, but not the biggest one in the world. Keeping your terrain scanner on in combat not only reduces your combat effectiveness, but also exposes the scanner to damage. I once had a potential winning run go down in flames because I lost my scanner after foolishly failing to swap it out in combat. Good scanners are precious - protect them!
Once could also consider using terrain scanners in conjunction with sensors. With this approach, you can run terrain scanners and sensors at all times except combat, see enemies coming in advance, and swap out our sensors and scanners to full combat utilities when needed. I can see this getting tedious however.
Its important to take advantage of the fact that terrain scan processors stack. The more you have going, the faster they reveal the surrounding area. Having multiple processors and the resulting swapping wouldn’t cost you too much matter if done sparingly, but keeping multiple scan processors would cost you inventory real estate.
8. Garrisons
Garrisons are a new map added in Alpha 4. They are explained in two dev blog posts (http://www.gridsagegames.com/blog/2015/11/garrisons/ and http://www.gridsagegames.com/blog/2015/11/garrison-access/). In a nutshell, they are machines that enemy squads can spawn from. If you hack them, you can actually enter them if you so choose. This will lead to a zone filled to the brim with enemies, and it will also raise the security level.
There are multiple possible benefits to entering a garrison, but perhaps the biggest one is that the exit in a Garrison has a chance to lead to the next level (though it also has a chance of returning you to the complex level you were originally on). In practice, if I had a very strong combat build and found a Garrison early on in Factory -5 or higher, I would strongly consider going in. I would not recommend going into the Garrison if you've been on the level for awhile, since a) the security level will be getting high and going into the Garrison will only make things worse, and b) if you've been on the level for awhile you'll probably find the exit soon enough.
There are a few tips to stair finding. I'll add more as I discover them. However you hunt for the stairs, the faster you find them the better off you’ll be.
I’m doing it in installments primarily because I’m still new to the game myself, even with a combat win under the belt. It’s also easier approaching a comprehensive guide if I break it into chunks. Please feel free to point out things I am missing or wrong about in the comments. This will help ensure that the final guide is as accurate and as thorough as possible.
1. The Main Goal.
When you enter a floor for the first time there is always one main goal, followed by various subgoals of secondary importance. The subgoals will vary from floor to floor and will vary based on your playstyle. Some people might want to make sure they have a schematic for a high capacity storage unit before leaving Materials. Other people might want to make sure they have a force field or powered armor before entering Research. These are all perfectly fine subgoals and reflect the wealth of options Cogmind gives you.
By definition, however, subgoals are secondary to the main goal. On every single level of this game there is always one main goal that takes precedence over everything else: find the stairs and get off the level as soon as possible.
The first half of that sentence is pretty self-evident. After all, the goal of the game is to escape. The second portion of that sentence is what you should focus on. The longer you linger on a floor, especially after Materials, the greater your chances of death. Time is of the essence.
A brief tangential discussion about the AI death spiral is now necessary, which will be covered in greater depth in the “Security Level” chapter. Briefly, there is a global security level in this game. The higher it goes, the greater the resistance the game will send after you. You always want the security level to be as low as possible. Killing enemies, however, raises the security level. This can and will lead to nasty feedback loops.
Once you get to factory, there is a steady stream of programmers sent after you periodically – the “food clock” of the game. Even if you’re capable of swatting them down, the mere act of killing them raises the threat level. If you dillydally, you’ll soon have extermination squads raining down upon you. You can’t dillydally. Find the stairs and get off the level as soon as possible.
This problem is magnified by combat builds. You’re designed to kill, not to run or hide, so you’ll necessarily be raising the security level. You’ll be using treads or legs, so you’ll be slow. The more time it takes you to get from point A to point B means the more programmer dispatches you’ll face. These two facts combined means that blindly wandering around is a great way to die. You need a plan to reliably find the stairs.
Below are various stair finding techniques. Mixing and matching these tricks based on the utilities you find is key to getting through Factory and Research quickly and efficiently.
2. Map sense
One way to find exits without the help of any utilities or hacks is simply acquiring what I call “map sense.” As you play the game more, you’ll see that there’s a pattern to where exits are. Generally speaking, exits tend to be in hallways, not in rooms; they tend to be along the perimeters of the maps, not towards the middle; and they are often in corner areas. These are generalizations. As a rule of thumb, however, traveling down hallways as opposed to in and out of rooms will help you find the exit faster.
It’s worthwhile traveling down these corridors hugging walls as opposed to in the center of the corridor. There are two reasons for this. One, as you pass doors they will open and you can peak in to see if there are any worthwhile items. Second, enemy patrols tend to move along corridors (most likely because these are the paths you have to take to get to the exits). If you’re hugging the wall, it’ll be easier to move into rooms to assume more favorable firing positions.
I’ll be expanding this section later with screenshots of various sample exits in Materials, Factory, and Research.
3. Hacking
Possibly the best way to find the exit. I’m not going to go into a full blown explanation of how hacking works, which will have to wait for its own chapter, but I’ll just touch upon a few preliminary things.
There are two types of hacks at a terminal: direct and indirect hacks. A direct hack happens when you select from the list of options offered by the terminal. The perk of this type of hack is you know your success chance. The downside is that you’re limited to what the terminal offers. An indirect hack occurs when you hit z and then type in the hack you want manually. The perk of this type of hack is that you can try whatever hack you want at any terminal. The downside is that is has a -15% penalty to success, and you can’t see your success rate.
Hacks that help you find the exit
The crowning achievement of hacks is Access(Main). A successful Access(Main) hack will show you all exits on the map. You can hold alt and scroll around with the arrow keys to see where the exits are in relation to you once you pull this hack off. Unsurprisingly, it is one of the hardest hacks and has a very low chance of success. Indirect Access(Main) hacks will often fail. If you see a terminal offering this choice (Level Access Points) and the success rate is above 15%, go for it. Even better if you have hacking suites.
I would not indirect hack Access(Main) unless I had lots of deep network scanners (which increase the success chances of indirect hacks), or if the Alert level was low. If the alert level is not low, you’re better off doing an Alert(Purge).
All of the above applies to Access(Branch), which shows you where the branch exits are, except that this hack is usually a bit easier to pull off. Currently not very useful with so many braches locked, except for Storage, but will become very useful indeed once the branches are added.
Another worthwhile hack in your quest for the exit: Index(Terminals) shows you all the terminals on the level. Knowing where they are gives you more opportunities to do direct or indirect Access(Main) hacks.
4. Drone Bays
Drone bays are one of my favorite ways to find the stairs. A drone bay is a 2-slot utility (the fact that they are 2 slots is a good hint that they’re a great item). When you equip it and activate it, it’ll deploy one drone each turn. The bay holds a maximum of 2 drones, so in 2 turns it will have emptied its payload. If you want to only deploy one drone, turn it off or put it back in your inventory after it has deployed the first drone.
Drones are little robots that move very quickly. Although they do not have armor, enemies have a -30% chance to hit them. Coupled with their fast speed, which means they won’t be in an enemies line of sight for long, they can have surprisingly long life spans.
Using Drones
To use a deployed drone, hold shift and right click on it, then tell it to explore. This will cause the drone to zip around the map using an exploration algorithm. You share vision with the drone so you’ll see what it discovers. If you hold the alt key you can scroll around the map. It’s good to periodically do this to check on your drone and to see what they’ve discovered. Hit Enter to recenter the map on your cogmind.
In the best case scenario, your intrepid drone will go far and discover the exit. Even if it doesn’t, it can reveal some portion of the map for you, find item caches, and give you an idea where enemy patrols are. If you know a room is empty or contains items you don’t need, you can bypass that room entirely. This furthers our goal of getting off the level as soon as possible.
Never send two drones out to explore at the same time. They’ll follow the same exploration path, which is a waste of a drone. Either deploy one drone and keep the other one in the bay, or wait a few turns before telling the second to explore. This will maximize how much you get out of each drone.
When a drone finds the exit, you’ll get a beeping noise that lets you know, just as if you had found the stairs yourself. When this happens, command your drone to return. If you have your drone bay equipped, you can scoop up a returning drone, allowing you to use it on the next level. For this reason, don’t discard an empty drone bay immediately after discarding it. I would only discard a drone bay if both drones associated with it were dead (and I either already had a drone bay schematic or did not intend to do any fabricating – see below).
It’s typically best to not hoard drone bays for too long, especially since they take up so much space. If I find drone bays, I’ll deploy one immediately and one each time I get to a new level. A real benefit (exploration now) is often better than a hypothetical benefit (saving them for later in the hopes that you can speed through Research). Getting to Research in strong shape is better than limping into Research with some drones.
Getting Drones
There are 3 ways to get drones. The first is to just randomly find a cache. Always a great moment, but luck dependent and nothing you can rely on. You can consider doing the “Index(Components)” hack at a terminal in Factory to see if there are any nearby drone caches. However, this is probably not the best use of a hack. If you do find a drone cache, I’d immediately deploy some, and save the rest for the next immediate levels.
The second way to get drones is to visit Storage, a branch off of Materials or early Factory. There is a very high chance of finding a drone cache in Storage. This is one level where the Index(Components) hack is a very good idea. If you find the drone cache, deploy some to see what else Storage has, and save as many as you to help get your through Factory
The third way to get drones is to fabricate them. To do this, you first need their schematic and then a fabricator. I would not bother trying to get the schematic through indirect hacks, since it’s a rating 5 item. Instead, if you find a drone bay bring it by a scanalyzer. That’s a much easier way to get the schematic than hacking. For this reason, I’ll keep even an empty drone bay around until I’ve gotten it scanned.
Once you have the schematic, the drone bay costs 40 material to fabricate. Since it’s a rating 5 item, having a hacking suite handy can help make sure you pull it off (hacking suites increase the chances of all direct hacks, and scanning/fabbing count as direct hacks).
5. Programmers (and Extermination Squads)
Ah programmers, the bane of your cogmind’s existence. They’ll corrupt you, turn your allies against you, and they won’t stop coming. They’ll also help you find the stairs if you know what you’re doing.
Once you get to Factory, every now and then you’ll get a message that says “rogue bot found, programmers report to [your location].” This message means that a certain number of programmers have been dispatched against you. They automatically know where you are and they’ll make a beeline to you until they find you.
They also spawn from the stairs. Therefore, if you know which direction they’re coming from, you know that there’s a stair case in that direction. This is extremely useful information.
When you know a programming team is en route, you want to prepare accordingly. You don’t want to be in the middle of a large hallway. If a group of hunters or grunts joins the party when the programmers arrive, things can go south quickly. The best place to fight them is in narrow corridors or in rooms with only 1 entrance.
If you’re in a corridor, pay attention to where they came from when they show up. Once you kill them, head in that direction. Rather than wait inside a room (which can prevent you from knowing where they came from), I sometimes like to wait right outside of rooms, popping into it once they show up. This lets me get the benefit of the room while letting me know where they came from. If you’re doing this, consider using an optical array so you can spot them before they get into firing range.
Besides giving you hints about where the exit is, programmers can also give information about the distance to it. Once you see the programmer dispatch warning, take note of how many turns pass until they get to you. If it’s very quick, that’s a strong hint that there’s an exit close by.
Extermination squads also come from the stairs, so take note of where they come from for the same reason. Theoretically, if you were by a terminal right when an extermination squad was dispatched, Enumerate(Exterminations) could possibly help you find the stairs. However, if the security level is high enough for these squads to be coming down, you should probably be purging the threat level rather than anything else at a terminal.
Alpha 4 (Garrisons) Update: The above tactic has been heavily, though not totally, nerfed in Alpha 4. Alpha 4 saw the addition of new machines called Garrisons. Enemies dispatched to the map (programmer dispatches, investigation/reinforcement/extermination/assault squads) have a chance to be deployed from Garrison access points rather than the stairs. Accordingly, if you play follow the programmer now, there's a very decent chance you'll find a Garrison rather than the stairs.
Enemies still have a chance of spawning from the stairs all the same, so this approach can work if desperate. Also, if you blow up garrisons as you find them you'll increase the chances that enemy spawns come from the stairs. On the other hand, this will raise the security level (though there will be a corresponding decrease in the frequency of programmer dispatches).
Besides blowing up garrisons, another possible approach to making this strategy viable again is to pull off an Index(Garrisons) hack. If you know where every Garrison is, this can help inform you whether a recently dispatched enemy came from a garrison or came from the stairs.
6. Sentries
Sentries in hallways are very often sitting on top of stairs. This fact is less important in Factory, which has pretty straightforward layouts. It becomes more important in Research, which is a more of a maze. Knowing where Sentries are can provide you with possible places to check out in your quest for the stairs.
There are two ways to search for Sentries beyond of your line of sight. First, you could use an advanced sensor array coupled with a signal interpreter. The sensory array lets you know where robots are well outside of your line of sight. The signal interpreter would let you know which of these robots are sentries. Spotting sentries could give you hints about where to go. Kyzrait made some gifs demonstrating this (along with an explanation of terrain scanners) in an informative blog post that can be found here:
http://www.gridsagegames.com/blog/2014/11/information-warfare/
You can either have these sensors on at all times or occasionally swap to them. Although a combat cogmind might be loathe to fill 2 utility slots with sensors, which would be useless in a battle, the sensors would enable you to see enemies coming so you could either avoid them, or swamp to more combat related utilities before battle begins. If that’s not your style, you could periodically switch to the sensors and signal array, and put them right back after sending out a quick scan to check for Sentries.
The second way to know where Sentries are lurking is the “Enumerate(Security)” hack. This hack will show you Sentries, but only those in the local area. I would not do this hack if the Alert level is high, since you’d be better of with an “Alert(Purge).” However, if I were on Research with a low security level and no hacking utilities, I’d consider it. An attempt at Access(Main) will almost certainly not work absent significant utilities or luck. A successful “Enumerate(Security)” is less helpful, but it’s better than nothing.
7. Terrain Scanners
Terrain scanners are interesting little items. When activated, over time they reveal walls, doors, and stairs in a certain area. By themselves, this takes awhile. With a terrain scan processor, however, they scan far more quickly. With a nice improved terrain scanner and improved scan processor, you can uncover the map and even find stairs before they enter your field of vision. Very handy stuff.
The more basic terrain scanners have their time and place but aren't much to write home about. Once you get to Long Ranged Terrain Scanners, however, you have one hell of an item indeed. With this scanners range, you can know most of the surrounding area in no time provided you have a good processor. In Factory and Research, this is a great way to find stairs. I consider long ranged or experimental terrain scanners to be one of the best utilities in the game and highly recommend you pocket one if you find it.
However, unlike a sensor array, a terrain scanner won’t let you know if enemies are coming. If you bump into some enemies with your terrain scanner stuff on, you’ll have to waste 2 turns to get to peak combat efficiency. This is a draw back, but not the biggest one in the world. Keeping your terrain scanner on in combat not only reduces your combat effectiveness, but also exposes the scanner to damage. I once had a potential winning run go down in flames because I lost my scanner after foolishly failing to swap it out in combat. Good scanners are precious - protect them!
Once could also consider using terrain scanners in conjunction with sensors. With this approach, you can run terrain scanners and sensors at all times except combat, see enemies coming in advance, and swap out our sensors and scanners to full combat utilities when needed. I can see this getting tedious however.
Its important to take advantage of the fact that terrain scan processors stack. The more you have going, the faster they reveal the surrounding area. Having multiple processors and the resulting swapping wouldn’t cost you too much matter if done sparingly, but keeping multiple scan processors would cost you inventory real estate.
8. Garrisons
Garrisons are a new map added in Alpha 4. They are explained in two dev blog posts (http://www.gridsagegames.com/blog/2015/11/garrisons/ and http://www.gridsagegames.com/blog/2015/11/garrison-access/). In a nutshell, they are machines that enemy squads can spawn from. If you hack them, you can actually enter them if you so choose. This will lead to a zone filled to the brim with enemies, and it will also raise the security level.
There are multiple possible benefits to entering a garrison, but perhaps the biggest one is that the exit in a Garrison has a chance to lead to the next level (though it also has a chance of returning you to the complex level you were originally on). In practice, if I had a very strong combat build and found a Garrison early on in Factory -5 or higher, I would strongly consider going in. I would not recommend going into the Garrison if you've been on the level for awhile, since a) the security level will be getting high and going into the Garrison will only make things worse, and b) if you've been on the level for awhile you'll probably find the exit soon enough.
There are a few tips to stair finding. I'll add more as I discover them. However you hunt for the stairs, the faster you find them the better off you’ll be.
15
Ideas / Balancing EM [spoilers]
« on: September 20, 2015, 09:45:52 AM »
Spoilers ahead.
Based on a discussion in the competition thread, it looks pretty clear that late game EM weaponry is objectively the best. (http://www.gridsagegames.com/forums/index.php?topic=261.125). K said that there's going to be some tweaking of the damage type. This thread is for people with combat experience to weigh in on how to tweak EM without nerfing it into the ground. I'll consider the various different ways EM works and how it could be changed.
1. Enemy resistances and vulnerabilities.
Currently, there are 3 major sources of enemies in the game: grunts, hunters, and programmers. Yeah, there are other bots out there, but you're only encountering melee dudes every so often, sentries and behemoths aren't ending your runs, and swarmers ain't no thang as long as you have a launcher. These 3 are the main opposition.
As zxc showed (http://www.gridsagegames.com/forums/index.php?topic=259.0), grunts and hunters are vulnerable to EM, while programmers are resistant to it. The programmer resistance is key to game balance; without it, there'd be no reason not to run all EM all the time. The grunt and hunter vulnerability makes EM the best weapon for taking these guys on. One could argue that you could remove the grunt vulnerability, so that hunters are weak to EM; grunts are weak to kinetic; and programmers are weak to thermal (a pending balance tweak), but that's getting kind of rock paper scissors. I think removing anyone's EM vulnerability is not the solution (though I think the pending increase to programmer EM resistance is a good idea).
2. EM and Security Level
Besides damaging an enemy, EM causes corruption. Corrupt an enemy enough and it collapses into bits irrespective of its core integrity. This is what makes EM so damn good.
It also gets you out of the AI death spiral. As you kill more stuff, the alert level rises causing more enemies to be sent out, causing you to kill more stuff, causing the alert level to rise, causing [etc etc etc] until the AI really drops the hammer on you. Managing alert level is one of the most important things of a run.
Enemies corrupted to death do not raise the security level. This helps you get out of the death spiral despite going in guns blazing. This also lets you keep the level low without having to invest in a lot of hacking. A great perk to EM.
On the downside, corrupted enemies don't go to your score. So currently an EM heavy run will score lower than a kinetic run. This seems right: the game is meant to reward tougher playstyles, so EM heavy runs scoring lower fits in with this nicely. Nevertheless, I think this is one area where EM should be changed.
Solution One: A robot corrupted to death has the same effect on the alert level as one whose core is destroyed. I don't like this solution. It completely removes the security level distinction, and this is a nice little feature that makes EM stand out. If this were implemented, the EM scoring penalty should be removed.
Solution Two: Cause all enemy meltdown/corruptions to raise the alert level, but 50% less or so than how much destroying the core would raise it. Obviously the AI should wonder what's up if a bunch of robots start melting down or shutting off. This makes kinetic weapons the riskiest weapons in terms of security levels, which seems right. This would be a slight nerf to EM without rendering the weapon unpalatable.
3. EM and Salvage.
This, bar none, is what I think makes EM the most OP. When you drop a micro nuke on a hunter, it'll get the job done but odds are there'll be nothing left to salvage. If you corrupt an enemy to death, it explodes into all of its wonderful bits. I used this in research to turn sentries into refilling stations. Fry one down, and you got tread replacements, armor, and railguns. It really helps keep you filled with the meat and potatoes of a combat run.
It gets insane though once you throw hunters into the mix. Corrupt a terminator and voila, you got an advanced targeting computer and an advanced cloaking device, two of the best utilities. Once I switched to heavy use of EM, I began looking forward to hunters. That right there shows something is off.
Solution One: Have any part coming off a corrupted robot have a 10-25% chance of being red (unusable). This reflects it being rendered useless after it got fried. I don't like this blanket solution, esp. because it doesn't make sense that EM would effect things like armor or treads in this way. Instead...
Solution Two: Have light parts/utilities (targeting computers, sensors, cloaking devices, etc) have a chance to be fried and unusuable if they resulted from corruption. This makes sense; if you destroy a robot with coruscating EM energy, wouldn't you expect its delicate bits to be ruined? With this change, EM weapons would be the king of salvaging armor, treads, and guns, but would suck for getting delicate utilities. I imagine you could even re-use the bashing code here, where chance of frying is in inverse proportion to integrity.
So a long winded post, but I think EM should be balanced with a bit more security level consequences and frying light utilities, in addition to the already reduced score it necessarily results in.
I'm sure there's other ways to balance it that I haven't thought of.
E: oops, I realize this should not be in the strategy subforum. Sorry!
Based on a discussion in the competition thread, it looks pretty clear that late game EM weaponry is objectively the best. (http://www.gridsagegames.com/forums/index.php?topic=261.125). K said that there's going to be some tweaking of the damage type. This thread is for people with combat experience to weigh in on how to tweak EM without nerfing it into the ground. I'll consider the various different ways EM works and how it could be changed.
1. Enemy resistances and vulnerabilities.
Currently, there are 3 major sources of enemies in the game: grunts, hunters, and programmers. Yeah, there are other bots out there, but you're only encountering melee dudes every so often, sentries and behemoths aren't ending your runs, and swarmers ain't no thang as long as you have a launcher. These 3 are the main opposition.
As zxc showed (http://www.gridsagegames.com/forums/index.php?topic=259.0), grunts and hunters are vulnerable to EM, while programmers are resistant to it. The programmer resistance is key to game balance; without it, there'd be no reason not to run all EM all the time. The grunt and hunter vulnerability makes EM the best weapon for taking these guys on. One could argue that you could remove the grunt vulnerability, so that hunters are weak to EM; grunts are weak to kinetic; and programmers are weak to thermal (a pending balance tweak), but that's getting kind of rock paper scissors. I think removing anyone's EM vulnerability is not the solution (though I think the pending increase to programmer EM resistance is a good idea).
2. EM and Security Level
Besides damaging an enemy, EM causes corruption. Corrupt an enemy enough and it collapses into bits irrespective of its core integrity. This is what makes EM so damn good.
It also gets you out of the AI death spiral. As you kill more stuff, the alert level rises causing more enemies to be sent out, causing you to kill more stuff, causing the alert level to rise, causing [etc etc etc] until the AI really drops the hammer on you. Managing alert level is one of the most important things of a run.
Enemies corrupted to death do not raise the security level. This helps you get out of the death spiral despite going in guns blazing. This also lets you keep the level low without having to invest in a lot of hacking. A great perk to EM.
On the downside, corrupted enemies don't go to your score. So currently an EM heavy run will score lower than a kinetic run. This seems right: the game is meant to reward tougher playstyles, so EM heavy runs scoring lower fits in with this nicely. Nevertheless, I think this is one area where EM should be changed.
Solution One: A robot corrupted to death has the same effect on the alert level as one whose core is destroyed. I don't like this solution. It completely removes the security level distinction, and this is a nice little feature that makes EM stand out. If this were implemented, the EM scoring penalty should be removed.
Solution Two: Cause all enemy meltdown/corruptions to raise the alert level, but 50% less or so than how much destroying the core would raise it. Obviously the AI should wonder what's up if a bunch of robots start melting down or shutting off. This makes kinetic weapons the riskiest weapons in terms of security levels, which seems right. This would be a slight nerf to EM without rendering the weapon unpalatable.
3. EM and Salvage.
This, bar none, is what I think makes EM the most OP. When you drop a micro nuke on a hunter, it'll get the job done but odds are there'll be nothing left to salvage. If you corrupt an enemy to death, it explodes into all of its wonderful bits. I used this in research to turn sentries into refilling stations. Fry one down, and you got tread replacements, armor, and railguns. It really helps keep you filled with the meat and potatoes of a combat run.
It gets insane though once you throw hunters into the mix. Corrupt a terminator and voila, you got an advanced targeting computer and an advanced cloaking device, two of the best utilities. Once I switched to heavy use of EM, I began looking forward to hunters. That right there shows something is off.
Solution One: Have any part coming off a corrupted robot have a 10-25% chance of being red (unusable). This reflects it being rendered useless after it got fried. I don't like this blanket solution, esp. because it doesn't make sense that EM would effect things like armor or treads in this way. Instead...
Solution Two: Have light parts/utilities (targeting computers, sensors, cloaking devices, etc) have a chance to be fried and unusuable if they resulted from corruption. This makes sense; if you destroy a robot with coruscating EM energy, wouldn't you expect its delicate bits to be ruined? With this change, EM weapons would be the king of salvaging armor, treads, and guns, but would suck for getting delicate utilities. I imagine you could even re-use the bashing code here, where chance of frying is in inverse proportion to integrity.
So a long winded post, but I think EM should be balanced with a bit more security level consequences and frying light utilities, in addition to the already reduced score it necessarily results in.
I'm sure there's other ways to balance it that I haven't thought of.
E: oops, I realize this should not be in the strategy subforum. Sorry!
16
General Discussion / What a live action Alert(Purge) looks like
« on: September 16, 2015, 12:57:09 PM »17
General Discussion / Ambush Traps: some thoughts
« on: September 16, 2015, 07:27:58 AM »
Apparently all I'm good for is complaining to the admin cogmind about traps.
So, Ambush traps. I had 2 runs basically ended by them yesterday. In the first one, on -7 I stepped on an ambush trap in a room. Out pop a grunt and a pest (who cares) and 2 brawlers and a duelist (oh shit). The melee enemies were like 2 tiles away or so, were on me immediately, and proceed to bash my processors in and slice off my weapons respectively. I was literally naked in 4 turns.
I'll readily admit this entirely preventable - I was in the room cause it had a terminal, I saw one ambush trap, correctly surmised it was a minefield layout, and went in hoping for the best. So, stupidity on my part.
Second run, I'm walking around -6 and in the middle of the hallway step on one. Same thing happens, I suddenly have 2 melee guys right next to me. My only option was to slap on a launcher and blow em up pointblank at extreme cost to my build.
So. Most other traps hinder you but not extremely so (bit of damage here, bit of corruption there). Ambush traps have the very real potential to strip your parts and force you to flee naked. Even though you can always "come back," this is far more likely to precipitate a death march. My game design view is stepping on a trap in an empty room should never have a realistic chance of totally wrecking you.
My suggestions:
-Have enemies spawn a bit farther away (esp. melee enemies) to give you more of a fighting chance.
-Perhaps decrease melee spawn rates in ambush trap groups, at least until upper factory
-
So, Ambush traps. I had 2 runs basically ended by them yesterday. In the first one, on -7 I stepped on an ambush trap in a room. Out pop a grunt and a pest (who cares) and 2 brawlers and a duelist (oh shit). The melee enemies were like 2 tiles away or so, were on me immediately, and proceed to bash my processors in and slice off my weapons respectively. I was literally naked in 4 turns.
I'll readily admit this entirely preventable - I was in the room cause it had a terminal, I saw one ambush trap, correctly surmised it was a minefield layout, and went in hoping for the best. So, stupidity on my part.
Second run, I'm walking around -6 and in the middle of the hallway step on one. Same thing happens, I suddenly have 2 melee guys right next to me. My only option was to slap on a launcher and blow em up pointblank at extreme cost to my build.
So. Most other traps hinder you but not extremely so (bit of damage here, bit of corruption there). Ambush traps have the very real potential to strip your parts and force you to flee naked. Even though you can always "come back," this is far more likely to precipitate a death march. My game design view is stepping on a trap in an empty room should never have a realistic chance of totally wrecking you.
My suggestions:
-Have enemies spawn a bit farther away (esp. melee enemies) to give you more of a fighting chance.
-Perhaps decrease melee spawn rates in ambush trap groups, at least until upper factory
Spoiler (click to show/hide)
18
Fixed Bugs & Non-Bugs / Display bug: cogmind disappears
« on: September 13, 2015, 06:08:21 AM »
How to reproduce: If you're directly above a staircase and hover over it with the cursor so the green box appears, your cogmind disappears until the box is gone. This is true for almost any square adjacent to a staircase, expect to the center left of it.
19
General Discussion / Blade Traps: Some thoughts
« on: September 05, 2015, 01:05:19 PM »
I think blade traps need some work. I just had the pleasure of a storage unit being sliced off, which spilled my stuff everywhere. It was a blade trap array of some sorts, so going after my severed bits caused more stuff to get sliced off. Since my stuff was covering the floor I couldn't even see the traps so I once or twice blundered onto one yet again. On the next level, the exact same thing immediately happened again literally a few steps from the stairs and I quit the game out of frustration.
I can imagine the odd emergent scenario where blade traps add something very interesting (having propulsion sliced off while fleeing from an enemy for example). However, having something sliced off in the middle of am empty corridor is just annoying. When that something is a storage unit it's especially unfun and just serves to generate a whole bunch of key presses.
I'm not sure what to suggest. I think once there's a "are you sure you want to step on that known trap" feature, it'll be slightly less annoying. Regardless, I ate a lot of explosive, EMP, and alarm traps in my seed game this morning, and none were frustrating or tedious like blade traps.
It's possible I'm just being a baby. On the other hand, a feature that routinely leads to you having to pick up half your inventory is just not that fun.
I can imagine the odd emergent scenario where blade traps add something very interesting (having propulsion sliced off while fleeing from an enemy for example). However, having something sliced off in the middle of am empty corridor is just annoying. When that something is a storage unit it's especially unfun and just serves to generate a whole bunch of key presses.
I'm not sure what to suggest. I think once there's a "are you sure you want to step on that known trap" feature, it'll be slightly less annoying. Regardless, I ate a lot of explosive, EMP, and alarm traps in my seed game this morning, and none were frustrating or tedious like blade traps.
It's possible I'm just being a baby. On the other hand, a feature that routinely leads to you having to pick up half your inventory is just not that fun.
20
Ideas / When pressing "f", have cursor default to last robot targeted
« on: August 01, 2015, 04:01:39 PM »
Let's say you want to blow up the recycler in the starting zone to get a storage unit. You press f and fire. Mr. Recycler doesn't die, but instead moves away. You press "f" again. Your cursor will be placed on where he was, not where he is now, forcing you to move the cursor before firing again. The default should be last targeted robot, not last targeted square.
Pages: [1]