Grid Sage Forums

Cogmind => Strategies => Topic started by: zxc on September 14, 2015, 07:49:53 AM

Title: Combat Run Theorycrafting & Open Discussion
Post by: zxc on September 14, 2015, 07:49:53 AM
Everyone is having trouble winning combat runs. So far, it's only been managed once with an amazing run by Adraius (http://www.gridsagegames.com/forums/index.php?topic=245.0). In this thread I want to start a discussion on how we can best achieve success with combat runs.

I think that combat runs are inherently random in nature compared with stealth runs. This is because in combat, parts get damaged and destroyed, and sometimes aren't able to be adequately replaced. Backup items you find depend largely on RNG, and importantly, the number of backup items you can carry depends on how many HCP storage units you find and equip - also largely dependent on RNG.

Key battle theory

I've noticed that my combat runs end shortly after a massive encounter: either on the same floor, or the one directly after. Let me illustrate with some screenshots.

Spoiler (click to show/hide)

This massive encounter destroys my ideal setup to that point, forcing me to adapt to items I scavenge from the battle. E.g. a forcefield + energy storage unit + heavy armour setup might be downgraded to a slab of light armour. Severe general item attrition occurs in that encounter, depleting my store of backup items. Add to that the massive increase in alert level that this battle causes, and you have a recipe for destruction.

My thinking is that this massive battle needs to be delayed as long as possible. -2 or -1, or none at all. If you have it earlier you will simply die from the attrition and alert level.

Many of these battles occur right in the open, because moving is costly, and it seems possible to destroy the small group of enemies before you. However, by the time you destroy them and reach their items, another group has appeared, and another, and you're surrounded. In my last run, this key battle occurred in a small corridor, which vastly limited the damage I took. It still annihilated my items, but I was able to destroy a huge number of enemies, resulting in a peak influence of 2232 on floor -5. It may be the case that moving to a good location to fight in, despite taking significant damage in traveling there, is the right call when there is a strong possibility that this battle may turn into a key battle.

Another important aspect of key battles is the looting that takes place after it (and even during it). Matter tends to run out, and tractor beams become a huge timesaver. Saved time, of course, means safety, as the longer you stay, the higher the chance that enemies find you and extend the battle further. One problem is that tractor beams like to get destroyed during the battle. Maybe carrying a spare and only equipping it for a turn before unequipping it is a good way of retrieving matter and retaining the tractor beam.

Power/Propulsion Evolutions

I've gone with no upgrades to propulsion, like Adraius. This seems correct, as treads are going to be insanely slow no matter how many you have, and their penalty stat for being overburdened is almost negligible compared to their base speed.

I find that additional power sources aren't needed for quite some time. This changes when you acquire a force field, as energy income becomes a priority. Still, I imagine 2-3 is sufficient in Factory and perhaps further.

Weapons/Utilities

One thing that is yet unknown is the optimal number of utilities : weapons. Let's focus on weapons first:

Expanding weapon slots does a few things.

First, it increases the maximum size of your volley, and in such a way that damage/time becomes more efficient. When attacking a key beefy target, such as a behemoth, one volley of three weapons is usually not sufficient to kill it; therefore, multiple volleys are required; however, greater volley size means that fewer volleys are required overall and a lot of time is saved. There are downsides to having a large volley however: overkill becomes a problem. If you destroy an enemy with your first three weapons in the volley, then the other three were effectively wasted time. Larger volley sizes mean greater matter and energy expenditure over time, stressing your resources and heat/energy setup. Finally, a large volley means a lot of enemy attacks can occur directly afterwards, before you have a chance to react; this means that a key piece of armour may get blown off, and then subsequent attacks are not being mitigated adequately. Therefore, large volley sizes mean a loss of control.

Second, it means less weapon swapping is needed during battles. Suppose you are confronted by programmers: you immediately unload with your kinetic weapon volley. Right behind them are some swarmers, and you fire your explosive launcher. Now you are under attack by hunters: you utilise your non-kinetic weapons to dispatch them. It means that you don't need to swap weapons to achieve optimal volley composition for the enemy you're currently facing. Swapping weapons takes about 150  time per weapon slot, and in that time you can suffer damage. A large number of weapon slots also means that when weapons get destroyed in battle, you don't necessarily need to immediately replace them like you would with only a few weapon slots, saving time and therefore reducing damage taken.

I tried a massive volley setup, and it worked for a time until I died in -2. Now I'm leaning towards a smaller volley setup because it allows more room for utilities.

Fewer weapons means more utilities. Utilities offer damage mitigation primarily, and storage units allow for more backup items.

Damage needs to be mitigated. This is obviously an enormous part of a combat run, where you will be taking a huge amount of damage no matter how you handle it. As I see it, the useful utilities for damage mitigation / avoidance are: 1. armour 2. resistance items 3. force fields. Dodging items can be excluded because they can't be utilised by tread-users, and the accuracy reducing utilities are really rare and probably not terribly useful unless stacked.

Armour is a key part of damage mitigation. The problem I find is that there isn't enough armour to go around. Therefore, other items soak up a lot of damage.

Resistance items are probably not very useful unless combined with other effective forms of mitigation, such as armour. Else, they will just get blown off, and they aren't terribly common enough to replace easily. I once found an 80% EM resistance 2x armour in the chute trap map.

Forcefields seem a really important part of damage mitigation, because they don't just soak up damage - they make some of it vanish into thin air. But like resistance items, they are sensitive themselves, and also require a bunch of other items to be most effective: excellent power sources, energy wells, and matter/heat-energy converters. This requires some luck with item generation. Another issue is that you need energy to use weapons (esp. TH/EM) and to swap items. This means you may want to stack kinetic weapons, which is what I've been trying to do.

My dream is to actually have good armour, good forcefields + related items, and resistance items, all equipped at the same time. This would provide outstanding damage mitigation and could be enough to win the game with.

Storage units are clearly important, because you can reduce the effect of RNG by reducing peak performance (because storage units use up utility slots otherwise used by 'useful' utilities) in order to increase average + floor performance (because backup items can be equipped when needed). The problem is that only the HCP storage units are fantastic, and they aren't especially easy to find. One key thing I've realised is that they're rating 1, and very easy to scan/fabricate. Even so, the fabrication system is fickle, only really letting you get one shot at a fabrication per fabricator unless perhaps you have a lot of hacking utilities. I was able to fabricate about three HCP storage units in my last run, but this wasn't enough. The current meta of stacking storage units is I think the best strategy.

The offensive combat-oriented utilities like targeting computers, target analysers, and so on, are obviously cool, but they're rare enough and flimsy enough not to be very useful I think. I dream of stacking these sorts of items and one-shotting everything in sight, but honestly, even if that could be achieved, eventually the parts would get destroyed with little to show for it.

What about sensors, interpreters and optics? Could be useful, but they're very flimsy again, and damage is going to be unavoidable. I once dreamed of having a bunch of penetrating weapons and some sensors in order to hit enemies through walls, but I see now that it's most likely not practical; for one, you'd likely get one volley off before the enemy is in your face and the point of the build is nullified.

What about hacking utilities? They would only be good as a swap, because they are ridiculously flimsy. Adraius was able to purge threat 10 times in his winning run. The problem is that system corruption becomes unavoidable in the late-game, and directly inhibits hacking, which is already quite hard when you wish to make indirect hacks like often necessary for a specific key hack like purge threat. Keeping some 6-8 hacking utilities is a whole lot of spaces that could've been used on other backup items, and sometimes you're just stuck without terminals for a long time, making them rather useless. There is also the matter cost of swapping so many items all the time.

Tractor beams? I think they're really good actually. They don't serve much of a purpose most of the time, but in key spots they really matter 8) such as during long extended battles and in the aftermath of battles. My earlier idea to keep a tractor beam in inventory and swap it in only for a single turn at a time is looking more and more attractive to me.

A matter container kept in inventory for backup matter and for fabricators seems important as well.
Title: Re: Combat Run Theorycrafting & Open Discussion
Post by: Draco18s on September 14, 2015, 07:55:11 PM
Quote
resulting in a peak influence of 2232

Holy fsk!  My last run (which ended after three prolonged fights across two levels) I had a peak influence of a mere 381.

I got lucky that run, ran into two force field generators.  They were stupidly inefficient as they also protected nearby allies, but at one point, with it AND an ECM enabled I was generating +77 energy a turn, so I could soak a decent chunk of damage (peak energy storage was 707, but at that point it was down at 580 as I'd lost my com.battery) and I was able to run away.  No, I did not equip both at the same time, that would have been dumb (AFAIK).
Title: Re: Combat Run Theorycrafting & Open Discussion
Post by: Happylisk on September 14, 2015, 08:02:00 PM
[reserved]
Title: Re: Combat Run Theorycrafting & Open Discussion
Post by: Reiver on September 14, 2015, 09:38:58 PM
My most recent weekly run was my first real shot at combat runs since the earlier alphas, and things are certainly... different.

Treads have been buffed enough that they're absolutely the way to go. I almost wonder whether legs need a little something (heavens knows what) to keep pace with the trundlin' things. I ended up with three slots, but only after I ran out of treads and needed something to keep moving with legs! So yeah, that part is good.

The biggest thing I notice though is the difference between a stealth build and a combat build is how many propulsion/power vs guns you take - absolutely everyone seems to need to emphasise utilities. And I disagree to a certain extent regarding the accuracy utility - it's effect is dramatic, especially if you're stuck without AoE weapons against Swarmers.

Needless to say, I rather wish the Accuracy Chip was actually a Targeting Pod instead - that way you could have a weapon mount that let you make your guns more accurate... and if they got blown off, you could at least shove an extra gun in the same space to keep your damage output up too. It'd also help make combat builds that little bit more... distinctive, I suppose? :)

I've also noticed a lot of cores you kill off robots are Light, Compact, etc. If a robot has plenty of weight tolerance, is there any reason it doesn't have the heavier versions in place? It'd certainly make scavenging the things easier, and if you're running energy weapons, you'll value the battery space. ;) I also sort of wish the Power slot were a touch more versatile; some days you need plenty of juice to run your lasers, but then you wind up finding a cache of kinetics, your lasers eventually get shot up, and you've now got dead slots.

Mobility-wise, there's lots of strategies to commit to, one way or the other. Weapons can be kitted out aplenty depending on your needs (I still fondly remember using a Greatsword while running a flying build; zoom up and chop their guns off!). Utilities? Utilities are half the game right now. But Power? Take as few slots as you can get away with, keep them filled with the biggest plus you can find, and you're done. Perhaps things like 'toughened' power cores, that are halfway between heavy reactors and armour; you slap 'em on if you need the health more than the spare juice... :)
Title: Re: Combat Run Theorycrafting & Open Discussion
Post by: Draco18s on September 14, 2015, 10:34:34 PM
Oh man, that sounds neat, actually.  Toughened power cores.
And yeah, I can agree with making some slots a little more versatile, e.g. you can put a battery into a power slot OR a utility slot (a battery afterall is just a really crappy generator ;) )

That last run was a combat run with treads, I never found a storage utility larger than medium, but I'm not sure I maintained a good ratio on slots.  Probably did not actually need that third power slot, but at the time I had two big power-hog utilities and the generators I had were only so-so (ion engines).  Ditto the fourth propulsion slot, but I was barely maintaining a weapons loadout at that point and another utility wasn't going to help unless I found one that was worth picking up.

Still, I finally scored higher than my first run.  Got 7200.
Title: Re: Combat Run Theorycrafting & Open Discussion
Post by: zxc on September 14, 2015, 10:42:07 PM
As I see it there are three reasons to upgrade propulsion on a combat build.

1. To increase mass support in order to speed up movement.
2. To have additional high integrity parts to soak up damage.
3. To make it more likely that at the end of a large battle, some propulsion will remain, vastly improving your move speed when looting.

I don't think 1. is an important factor with treads. Using the juicy weapons and HCP storage units, you will certainly go over your support limit anyway, and the penalty stat on treads is small in proportion to the speed stat. This means that there is very little speed gained by expanding mass support.

I think 2. is an interesting thought. Armour is hard to find, and supplementing armour with treads may help somewhat.

I think 3. is extremely important, however this can also be achieved by keeping spare treads in your inventory.

On the balance, it seems like two propulsion slots is still the way to go.
Title: Re: Combat Run Theorycrafting & Open Discussion
Post by: Kyzrati on September 15, 2015, 10:03:57 PM
Quote
resulting in a peak influence of 2232

Holy fsk!  My last run (which ended after three prolonged fights across two levels) I had a peak influence of a mere 381.
zxc's combat experience is going to be quite different from other players, because as I explained in another thread (http://www.gridsagegames.com/forums/index.php?topic=261.msg2265#msg2265) he's using an incredibly efficient method of play that is keeping his perceived threat high. (You can read more about it over there.)

I almost wonder whether legs need a little something (heavens knows what) to keep pace with the trundlin' things.
They got a +50% integrity boost! That's pretty good considering how fast they are, and common. (They can also somewhat benefit from Maneuvering Thrusters, unlike treads.)

especially if you're stuck without AoE weapons against Swarmers.
Positioning is everything. I will literally absorb fire from Swarmers for a while to reposition myself somewhere I can take them all out with ease. But yeah I much prefer to have an AoE weapon :P

Needless to say, I rather wish the Accuracy Chip was actually a Targeting Pod instead - that way you could have a weapon mount that let you make your guns more accurate... and if they got blown off, you could at least shove an extra gun in the same space to keep your damage output up too. It'd also help make combat builds that little bit more... distinctive, I suppose? :)
Hm, I had an idea for a mechanic that wasn't this, but the interface I theorized could potentially work to implement what you're talking about here, and I like that idea a lot...

It would require a lot of work because it doesn't fit into the current systems, but I think as you say it would give combat builds an interesting alternative approach. In a way it would encourage taking more weapon slots, but it would have to be worth more than taking a more generally versatile utility slot.

I've also noticed a lot of cores you kill off robots are Light, Compact, etc. If a robot has plenty of weight tolerance, is there any reason it doesn't have the heavier versions in place?
Sure you might prefer the energy storage, but they don't!

Perhaps things like 'toughened' power cores, that are halfway between heavy reactors and armour; you slap 'em on if you need the health more than the spare juice... :)
In Alpha 3 I already increased all non-light power source integrity by 50%!

The light ones are still fairly weak, but then they don't weigh much... With a combat build I always carry extra non-light power and usually don't need to rely on light variants (unless things went really poorly for some reason, and I'll occasionally attach a battery if needed).

And yeah, I can agree with making some slots a little more versatile, e.g. you can put a battery into a power slot OR a utility slot
Parts must remain single-slot-type, for implementation reasons (that's a set requirement); from some players' point of view allowing multi-type parts would be too confusing, anyway.
Title: Re: Combat Run Theorycrafting & Open Discussion
Post by: Draco18s on September 16, 2015, 11:19:36 AM
And yeah, I can agree with making some slots a little more versatile, e.g. you can put a battery into a power slot OR a utility slot
Parts must remain single-slot-type, for implementation reasons (that's a set requirement); from some players' point of view allowing multi-type parts would be too confusing, anyway.

Oh I figured. :)
That was more of an "ideal world" pipe-dream comment.

I'm a (game) programmer myself and I know quite well the limitations that one can inadvertently code oneself into.
Title: Re: Combat Run Theorycrafting & Open Discussion
Post by: Kyzrati on September 16, 2015, 08:13:32 PM
I'm a (game) programmer myself and I know quite well the limitations that one can inadvertently code oneself into.
To ensure progress on the game itself rather than endlessly architecting the source, I enforced a number of assumptions from the very beginning--this was one of them.

I also decided to use the 7DRL code base, which was written under duress, but had the advantage of making other such assumptions that have served to limit me from engaging in endless scope creep ;). Another example would be that utilities can only have a single ability, or at least abilities that are all attached to a single enumerated type with an effect degree based off no more than a single integer value.

Overall, I'm grateful for the limitations that a 7DRL engine has imposed on development, though it obviously hasn't put a complete stop to scope creep. For the features that would be really beneficial, there are always ways around limitations :) (but you have to make sure you absolutely must have them!)
Title: Re: Combat Run Theorycrafting & Open Discussion
Post by: Draco18s on September 17, 2015, 08:52:44 AM
Definitely :)
Title: Re: Combat Run Theorycrafting & Open Discussion
Post by: zxc on September 18, 2015, 12:28:11 AM
Some updates to my thoughts on combat runs and some new ideas, after winning one:

Scanalysers/Fabricators

I think by far the most important item to fabricate is the HCP. Storage Unit. Having a large supply really makes a big difference. Also, it's only rating 1, so it's very easy to scan and fabricate. I made several during my winning run but I still ran low on them and had to use large ones that I found as well. If you really need storage, kill the high tier Recyclers (R) as they carry Medium Storage Units.

Hacking

The number one hack by far is Alert(Purge). I think this was the main thing that got me past the threshold and onto victory in that run. Access(Main) is excellent as well, but primarily useful in Research and only once your threat is low. Other useful hacks include Index(Terminals), Index(Fabricators), and simply Alert(Check). Possibly recalling assault forces and reinforcements as well, though I didn't get to make use of them.

I carried with me every Deep Network Scanner I could find (about four or five in total by the end). This helped a great deal with the manual hacks, and meant I was better off using manual hacks than the direct hacks when they were available. Assuming the regular Hacking Suites only help with direct hacks, you're best off either stacking Hacking Suites or Deep Network Scanners (preferably the latter as they are always useful).

Evolutions

Three power sources were plenty by -1. Certainly no more is necessary, because even with a force field my energy was barely taking a dent when under attack, so I ended up ditching my Exp. Energy Well. The starting two propulsion slots is enough. No change of opinion there. Five weapons seemed a good number by the end. It seems to be a sweet spot between a powerful enough volley and a short enough volley to limit overkill.

Weapons

I prefer kinetic weapons because they seem to do reliable damage to everything while not needing much energy and not generating much heat. This allows for very few utilities spent on heat dissipation, and allows for the maximum energy supply to be devoted to mitigating damage via a force field. Kinetic weapon focus also makes sense with treads because they are the only weapons with recoil. Switching to EM weapons for Behemoths is excellent. Programmers still overheat extremely easily so using thermal weapons against them is also extremely viable.

Enemies

The worst enemies are the assault squads that come with a Saboteur, several Grunts and a Bruiser or Duelist. Saboteurs are especially important to destroy immediately because they blow off several components, which could include storage units, which is extremely annoying and tough to deal with. Hunters are also pretty tough to kill and very annoying because 1. they resist kinetic damage 2. they have high crit weapons which like to destroy your items esp. storage units 3. they fire through walls which is tough to deal with when your move speed is low (even moreso when your treads get destroyed and your move delay is 2k+). On the flipside, Hunters drop very useful kinetic weapons.

Items

Keeping a Lrn. Tractor Beam in your inventory and swapping it in to pick up a ton of matter around the place and swapping it out again is an awesome tactic I encourage everyone to try. It saves a heap of time and there is no possibility of losing the item in combat (plus you get to use a more combat-oriented utility). I also encourage doing the same with a matter storage unit: keep one in your inventory and swap it in temporarily when you need matter, and otherwise make use of it for fabricators.

Positioning

The 1x tile wide long corridors are perfect for fighting in. Ideally you also want to fight deep inside it, not just along the outskirts, because enemies will approach to get closer and once they enter the corridor they block the fire of their allies. They also like to run off and try to come around the other side when the way is blocked, which is actually helpful sometimes because it gives you time to prepare, and sometimes you can dispatch the original enemies fast enough that you are only ever fighting on one side at a time.

When programmers are dispatched, staying at a narrow corridor to deal with them is a good tactic. Problems arise when you are caught in the open by them, and then other enemies get involved as well, and you are suddenly thrown into a huge battle with very poor positioning.

When moving around the map, especially without an excellent launcher at hand, narrow areas and passageways should be preferred. The open area is very dangerous, although often necessary to cross at times. Never try to use launchers when you're near walls.

Targeting Min-Maxing

If you move and an enemy comes into view, don't immediately fire at volley at them. Instead, rest one turn (or even better, make a useful item swap such as putting on the appropriate damage resistance item for that enemy). This will improve your targeting accuracy by about 10%. This tip is most relevant when you have a sizeable volley of say, five or more weapons, with a long fire delay.
Title: Re: Combat Run Theorycrafting & Open Discussion
Post by: Reiver on September 20, 2015, 07:09:59 PM
Needless to say, I rather wish the Accuracy Chip was actually a Targeting Pod instead - that way you could have a weapon mount that let you make your guns more accurate... and if they got blown off, you could at least shove an extra gun in the same space to keep your damage output up too. It'd also help make combat builds that little bit more... distinctive, I suppose? :)
Hm, I had an idea for a mechanic that wasn't this, but the interface I theorized could potentially work to implement what you're talking about here, and I like that idea a lot...

It would require a lot of work because it doesn't fit into the current systems, but I think as you say it would give combat builds an interesting alternative approach. In a way it would encourage taking more weapon slots, but it would have to be worth more than taking a more generally versatile utility slot.

While I'm absolutely positive that whatever mechanic you're thinking up is going to be subtle and brilliant, if the time:benefit is not in your favour, straight-up +accuracy targeting pods would not be in the least bit unwelcome.

As zxc points out, combat builds really only want half a dozen weapon slots at the moment - the slot type is self-balancing, because recoil, heat and volley time mean that you start getting diminishing returns on more slots very quickly. Meanwhile, utilities can do it all - armour, storage, scanners, stealth, hackware and combat efficiency (via +accuracy items)?

Yeah, I'd really prefer to  be given a reason to 'load up' on gun slots. Even if their sole purpose is 'make your shooting better' in the same way the purpose of Movement is 'make your moving faster', it'd be nice if there wasn't a utilitiy that did a better job at improving the gun salvo you already had more than an actual gun slot, y'know? It'd be like there being Utilities that outright improved your move speed without the loss of versatility and heat output of actually installing more VTOL thrusters...

It's up to you, of course. It just feels odd that once you reach a certain point, the correct way to boost your firepower is to install more utility slots. The weapon slots had enough drawbacks as it was.
Title: Re: Combat Run Theorycrafting & Open Discussion
Post by: Kyzrati on September 20, 2015, 10:35:42 PM
It's up to you, of course. It just feels odd that once you reach a certain point, the correct way to boost your firepower is to install more utility slots. The weapon slots had enough drawbacks as it was.
I think this makes a lot of sense as is, really. Logically speaking, advanced warfare is all about improving your additional capabilities on top of a small selection of weapons you're carrying.

I don't really mind that utility slots are the superior "go to" slot when everything else is sufficient for your needs. It even comes closer to one of my original concepts for the game (before the 7DRL)--that there is no such a thing as "slots" and you can equip as many or as little of anything as you want. (That said, I'm still glad the basic types were differentiated.)

From a mechanics standpoint, however, I can see the appeal of greater differentiation for combat builds.

I'm not sure about this other feature, and it might be really close to what you were thinking (?), but I've started a new thread for it here (http://www.gridsagegames.com/forums/index.php?topic=275.0).
Title: Re: Combat Run Theorycrafting & Open Discussion
Post by: Adraius on October 04, 2015, 06:21:21 PM
Impressive thread; I concur with just about everything you've put forth.  I already see the big value in keeping an Lrn Tractor Beam in storage for post-battle cleanup, and will at least give fabricating Hcp. Storage Units a try if I can't locate any larger than a Med. version.

I'd like to differ, however, on the usefulness of combat utilities, Targeting Computers in particular; although combat utilities are flimsy, their very small coverage means that they can last a long while. Think of HP/coverage as 'effective hitpoints' - I can't check at the moment, but I think they might have more than your average armor plate, although they are more heavily affected by RNG.  And as Reiver says, the effect is pretty dramatic - going from 60% to-hit to 72% is effectively a 20% increase in applied damage with none of the weight, energy/mass consumption, or heat generation of another weapon.  As for Targeting Analyzers, stack a few good ones on weapons with a significant crit chance, and you can get upwards of a 50% chance to crit.  Unfortunately, I'm less convinced of the value of relying on crit-kills or critting parts off of enemies when you want to be dealing massive damage to kill them regardless - and while Target Analyzers are pretty common (loot Programmers), Targeting Computers are rather scarce.

Also, I'd like to point out that system corruption is massive pain in the later levels, but not always unavoidable; I won my combat run with 0% corruption thanks to a pair of items - System Recalibrators? - that each purged 1 point of corruption every 150 turns, double the efficiency of the baseline units.  They were a lucky find, but if they are a low enough rarity to be feasibly fabricated, they may be an option; without high system corruption, hacking and, well, everything remains more viable.

EDIT:
Another big factor in the success of combat runs, IMO, is simply map familiarity.  You are, almost by definition, slow as hell; knowing which direction an exit is most likely to be in based off of the types of spaces, corridors, and rooms you see and what you know of the map's shape so far is very important in minimizing time spent and attrition incurred on levels where your build is already close to optimal. (as an aside, refreshing this familiarity after three weeks away is going to suck...)

In another thread, the ability to be able to see the layout of maps in a large image post-run was brought up.  That would probably help a lot in this department, although I'm not sure it's a good addition in game design terms; I think perhaps map familiarity is something best built up by playing the game - and sharing our discoveries socially - and not by dissecting images.
Title: Re: Combat Run Theorycrafting & Open Discussion
Post by: Kyzrati on October 04, 2015, 08:02:44 PM
Good points. Coverage is an especially important mechanic that shouldn't be overlooked.

I occasionally switch over to the 'c' mode visualization to make sure my important utilities have sufficiently relative low coverage that they won't go boom as soon as I meet an enemy (in situations like that, combat builds may choose to be a little more wary of engagements, and I'll even put something back in my inventory if I really want to keep it safe until better protection can be found). Another good solution that Happylisk likes to use is Utility Shielding, to prevent the odd shot from taking out a low-integrity part.

In another thread, the ability to be able to see the layout of maps in a large image post-run was brought up.  That would probably help a lot in this department, although I'm not sure it's a good addition in game design terms; I think perhaps map familiarity is something best built up by playing the game - and sharing our discoveries socially - and not by dissecting images.
I was just writing about this for a blog post yesterday (the next). I wouldn't want to reveal the whole map--that's not going to happen--but what would you say to outputting an image of what the explored part of the map looked like?
Title: Re: Combat Run Theorycrafting & Open Discussion
Post by: Adraius on October 04, 2015, 08:15:17 PM
Good points. Coverage is an especially important mechanic that shouldn't be overlooked.

I occasionally switch over to the 'c' mode visualization to make sure my important utilities have sufficiently relative low coverage that they won't go boom as soon as I meet an enemy (in situations like that, combat builds may choose to be a little more wary of engagements, and I'll even put something back in my inventory if I really want to keep it safe until better protection can be found). Another good solution that Happylisk likes to use is Utility Shielding, to prevent the odd shot from taking out a low-integrity part.

In another thread, the ability to be able to see the layout of maps in a large image post-run was brought up.  That would probably help a lot in this department, although I'm not sure it's a good addition in game design terms; I think perhaps map familiarity is something best built up by playing the game - and sharing our discoveries socially - and not by dissecting images.
I was just writing about this for a blog post yesterday (the next). I wouldn't want to reveal the whole map--that's not going to happen--but what would you say to outputting an image of what the explored part of the map looked like?
I think that strikes an excellent compromise - play the game, get a good view of what you uncovered and how your choices led you to uncover it.  Yes please!

An awesome add-on would be an overlay showing the path you took, but I imagine that would be a lot more coding.
Title: Re: Combat Run Theorycrafting & Open Discussion
Post by: Kyzrati on October 04, 2015, 08:42:46 PM
An awesome add-on would be an overlay showing the path you took, but I imagine that would be a lot more coding.
Not a lot more! In fact, I bring up that very possibility in the blog post. After looking at some example shots, I think that would make such a map a lot more meaningful, and interesting!
Title: Re: Combat Run Theorycrafting & Open Discussion
Post by: Draco18s on October 04, 2015, 09:49:16 PM
I'd look over my post-run maps, absolutely.
Title: Re: Combat Run Theorycrafting & Open Discussion
Post by: zxc on October 04, 2015, 10:40:06 PM
I'd like to differ, however, on the usefulness of combat utilities, Targeting Computers in particular; although combat utilities are flimsy, their very small coverage means that they can last a long while. Think of HP/coverage as 'effective hitpoints' - I can't check at the moment, but I think they might have more than your average armor plate, although they are more heavily affected by RNG.  And as Reiver says, the effect is pretty dramatic - going from 60% to-hit to 72% is effectively a 20% increase in applied damage with none of the weight, energy/mass consumption, or heat generation of another weapon.  As for Targeting Analyzers, stack a few good ones on weapons with a significant crit chance, and you can get upwards of a 50% chance to crit.  Unfortunately, I'm less convinced of the value of relying on crit-kills or critting parts off of enemies when you want to be dealing massive damage to kill them regardless - and while Target Analyzers are pretty common (loot Programmers), Targeting Computers are rather scarce.
Don't get me wrong - I do like those offensive utilities very much. I equip them whenever I can. I'm just saying they're fragile enough that they get blown off quite quickly and therefore aren't extremely useful in the grand scheme of things compared with say, force fields, storage, and so on. That first post was also written before I won my combat game, in which I was using about three utility slots for offensive utilities by the end. Basically, any left-over slots after storage, armour, force field, energy utilities if needed, heat dissipation, and damage reduction for the specific enemies I'm facing. I also agree that targeting computers are the best offensive utilities.

Edit: Also, comparing integrity/coverage of offensive utilities vs armour doesn't make sense as you want armour to have both high integrity AND high coverage. In practice, flimsy offensive utilities last 1-3 hits, and the way to make them useful is to simply have a steady supply of them from enemies you kill and stockpiles you find. Hunters often have target analysers IIRC, but I don't think any enemies have targeting computers. These utilities just don't last a minute without armour.

Also, I'd like to point out that system corruption is massive pain in the later levels, but not always unavoidable; I won my combat run with 0% corruption thanks to a pair of items - System Recalibrators? - that each purged 1 point of corruption every 150 turns, double the efficiency of the baseline units.  They were a lucky find, but if they are a low enough rarity to be feasibly fabricated, they may be an option; without high system corruption, hacking and, well, everything remains more viable.
I also found that system corruption was pretty much a non-issue in my winning run. It seems that massive system corruption is more a symptom of a feedback spiral leading to a loss than an unavoidable factor. My highest system corruption was just 17%, reached on the hardest floor, -3. I only made use of the purging utilities on -1 because simply escaping ASAP is the best way of purging corruption. I would not recommend fabricating these items as you'd get much more use out of fabricating HCP storage units.

One thought I ought to mention, unrelated to the above: When running out of utilities to equip, I would just equip anything I could scavenge, even if it would be completely useless for me. This included heat sinks, for instance. The reason is that these flimsy, useless items can still soak up hits for other, more important parts. Small tip for anyone who doesn't already do this.
Title: Re: Combat Run Theorycrafting & Open Discussion
Post by: Decker on October 12, 2015, 12:08:30 PM
At long last, I got a combat win!

My strategy is similar to the ones you guys use, with a few major differences.

Main goals:
1) Keep alert level low.
2) Keep corruption under control.
3) Keep a high inventory with all the spare parts needed.

Finding the stairs early is useful for 1) and 2), but sometimes it pays off to linger on the map longer to purge alert and find better parts (especially for Hpc. Storage Units).

My combat build focuses on four aspects.
a) Reduce incoming damage.
b) Soak up more damage per part.
c) Keep a steady supply of spare parts.
d) Stay mobile.

It turns out that there is a strong synergy between these four aspects.

Reducing the incoming damage is difficult. The only effective way I found is to use force fields. Cutting the damage by half really makes a difference.

Counter-intuitively, utilities like targetting computers, cloaking devices and shielding actually make you more vulnerable in combat when used before Research. This is due to a combination of factors: fragility, reduced coverage, rarity, resource drain and marginal effectiveness.

Fragile parts tend to get blown off quickly, even if you go out of your way to increase your total coverage to protect the fragile parts. Even parts like Improved Utility Shielding, which has a reported coverage of 0 (?) and presumably block 66% of the damage to itself, somehow do not survive long in combat.

Fragile parts are typically rare. Enemies usually don't bring them to you for use at your convenience like they do with the more mundane parts. Replacing them in a problem, especially since they are high-churn parts. Carrying a bunch of them in your inventory takes space away for the more vital items such as power, propulsion, heat sinks, and weapons.

Fragile parts have both low coverage and low integrity, meaning that they don't protect (cover) each other, do not protect force fields, and don't help absorbing incoming damage. Moreover those parts drain resources (energy, mass), do not contribute back resources, and take slots away from the parts that do contribute resources. To top it off, the effect of those parts is generally limited. Resistances in particular just don't cut it.

In contrast, parts for power, propulsion and weapons have a lot going for them. All these parts are commonly found laying around and on enemies. They are thus easier to replace.

Power parts have 60+ coverage, decent integrity, and give you (reserve) power for your force fields and for large volley of energy weapons.

Propulsion parts have 80+ coverage, high integrity, and keep you mobile. Morever, they have no mass, do not drain energy, and don't generate heat when not moving.

Weapons have 100+ coverage, decent to high integrity depending on the weapon, and reduce the time it takes to kill the enemy, thereby reducing indirectly the amount of damage taken.

Hence, the mundane parts protect force fields and Hpc. Storage Units, soap up lots of damage and generate the resources you needed. Before Research, I forego on the fragile items entirely and focus on keeping a high stash of the vital spare parts in case I run in a dry spell or get bogged down in a protracted battle. You can survive without fragile parts, but you can't survive without power, propulsion, weapons, heat sinks and storage.

Unlike most other players, I use 4 propulsion slots: 3 legs, plus 1 treads. Treads are uncommon, but since I attach just one of them at a time, churn is rarely a problem. Legs are carried by grunts so replacing them is usually straightforward.

Using 3 legs give you a speed rating of 160 to 190 (using four legs make you slower). I try to keep the speed to 160 since it helps a lot with the food clock, it reduces the incoming damage in tactical repositionings, it helps to flee from an area that the AI keeps reinforcing, and it helps to escape from chute traps. When I need to fire a weapon with recoil, I switch from legs to treads. This is an instant action, though it is very tedious to do repeatedly.

Thus, with 4 propulsion slots you get high coverage, high integrity, the speed + availability of legs, the recoil dampening of treads, all of that at no cost to mass, energy or heat.

Before Research, I keep the utility slots to a minimum. 4 Hpc. Storage Units, 2 heat sinks, 1 armor, 1 force field. That's all I need, really. The armor lasts longer since the mundate parts take some of the damage. I use 3 power slots, the rest is weapon slots. I finished the game with only 3 power slots, but I regretted it because I ran out of juice in tight spots.

I think that 5 weapon slots is all you need by the end game. Weapons have a lot of mass so they tend to make your speed climb to 190. Also, getting more means that you shift the churn toward the weapon category due to increased coverage, which is not necessarily a good thing. Volleys with 5 weapons tend to overkill, drain a lot of your energy, and leave you unable to respond quickly when some part blows up. Morever, launchers don't need lots of slots, and launchers are the most effective weapons.

Unlike many players, I tend to use thermal weapons most. Heat and energy are less of a problem given my focus on using mundane parts. Thermal damage is neutral, so it can handle whatever the AI throws at you without requiring a part switch. This is especially useful when the shit hits the fan and you're fighting off a mixture of grunts, hunters and programmers. Morever, thermal weapons don't consume matter so you can use your launchers liberally knowing that you can keep on fighting without matter when things are grim.

By the end game, an all-EM volley corrupts the enemies fastest, but you're in trouble if you meet programmers. An all-thermal volley helps enemies cook themselves to death. It's especially good when you're surrounded and the enemies spend several turns sniping at you. They'll be near the boiling point, they just need a little help from you to melt down. In other cases, the raw thermal damage is sufficient to kill most enemies eventually. The heat makes your current enemy a little less accurate and more vulnerable to your next volley.

I find that the most grindy part of the game is to get your hands on enough Hcp. Storage Units to survive up to the end game. You can download schematics starting at -9, but the food clock starts at -7. Effectively you have two levels to download the schematic, get a matter storage unit, and stumble on a low-level fab. Index(Fabricators) rarely works. Then you sit on the fab and milk it repeatedly. By the time you hit Factory, the combination of high-level terminals/fabs, food clock and corruption means that fabbing (hacking, really) is not really an option anymore. From this point on you have to rely on luck to find the storage units or the hackware to make fabbing possible again. I really hope this gets addressed.

The hardest part of the game for me is Research. Access is comparatively easy. I didn't know what to do when I first got there and ended up circling the map twice until I found the hackware necessary to get out. I finished with 1000+ HP and no corruption. The enemies mostly keep to themselves there. I guess that's the best location to maximize your score -- just keep shooting and purging.

Thanks for the game!
Title: Re: Combat Run Theorycrafting & Open Discussion
Post by: zxc on October 12, 2015, 01:01:42 PM
Stellar first post Decker!

Force-fields are certainly one of the most important items, yet you can't rely on having one until you start coming across behemoths.

Regarding part integrity, the fragile utilities you disdain are perfectly useful while you can maintain a steady stream of replacements. However, I have noticed that in all my combat run losses, the real killer is running out of core items i.e. weapons, power, propulsion, and of course storage. The points you make regarding maximising coverage per part equipped make a great deal of sense, and would let your hard to replace items like storage and force fields last much longer. Replacing fragile utilities (that aren't force fields) isn't a huge deal because they would not be a core aspect of the build, so you can just equip the next best item available.

Resistance items should become more useful the later in the game you are, as the damage they nullify should begin to outweigh the loss in coverage/integrity from equipping them. One slot to reduce damage by ~20% is surely worthwhile.

Utility Shields (and the similar Propulsion/Weapons/Power Shields) absorb damage that would otherwise be inflicted on other utilities, hence the 0 coverage and why they don't seem to last long in combat. I think they are more useful for stealth/speed builds than combat, as lots of armour will make them rather redundant (at least, the utility shields certainly).

That's a very cool propulsion setup. It certainly seems to get the best of all worlds right now. Given how difficult it can be to find armour sometimes, I'm leaning towards more propulsion slots myself. Plus, getting stuck with no propulsion and high mass is really, really bad. In the next version(s) I believe Kyzrati is going to make treads give a recoil reduction bonus per active tread, which would nerf this particular setup.

Two heat sinks I think aren't necessary when relying mostly on kinetic weapons. With your strategy of stacking thermal weapons it makes more sense.

I love how now there are winners that advocate for each of the three main damage types - myself with kinetic weapons, Happylisk and Kyzrati with electromagnetic, and now yourself with thermal weapons. It goes to show there really is a lot to this game and it's not so simple to solve. There are excellent points to make in support of stacking each of these damage types. I do believe a slight buff to thermal damage is coming as well.

I think everyone agrees the hardest part of the game is Research right now. This is especially so due to the current bug on Access, where enemies don't seek you out and alert levels appear to have no effect.

Some changes will be coming at some point to storage and item attrition. Have a look at some of the threads in the Ideas sub-forum.

I'd like to have a look at your morgue file if you don't mind.
Title: Re: Combat Run Theorycrafting & Open Discussion
Post by: Decker on October 12, 2015, 01:42:57 PM

Resistance items should become more useful the later in the game you are, as the damage they nullify should begin to outweigh the loss in coverage/integrity from equipping them. One slot to reduce damage by ~20% is surely worthwhile.


20% reduction is certainly nothing to sneeze at, but you'd need 3 slots if you wanted it to work accross all three damage types (unless you swap parts).

Quote

Utility Shields (and the similar Propulsion/Weapons/Power Shields) absorb damage that would otherwise be inflicted on other utilities, hence the 0 coverage and why they don't seem to last long in combat. I think they are more useful for stealth/speed builds than combat, as lots of armour will make them rather redundant (at least, the utility shields certainly).


Thanks for the explanation! That's one less mystery.

Quote

I'd like to have a look at your morgue file if you don't mind.


Certainly. Although you'll see that I haven't been using thermal all that much since I didn't get enough power slots and I was afraid (wrongly I believe) of running out of thermal weapons due to churn by overrepresentation. The kinetic weapons I found had higher integrity so I kept them around longer than I should have.

Thanks for the guides and explanations you've posted in these boards. It was very helpful to me!
Title: Re: Combat Run Theorycrafting & Open Discussion
Post by: Kyzrati on October 12, 2015, 07:33:04 PM
Welcome, Decker, and thank you for the excellent strategy overview!

This morning before checking the forums I went to put up the first leaderboard data after the reset and saw "Decker  Surface  20619" ... ... "Who the heck is Decker?!" :)

When I need to fire a weapon with recoil, I switch from legs to treads. This is an instant action, though it is very tedious to do repeatedly.
We'll be getting a new method for simultaneous activation/deactivation of all propulsion, as well as cycling through propulsion types, which would help reduce the tedium there. As zxc says, though, there will be a per-tread effect on recoil coming, so that tactic will be somewhat less effective (though still viable, I think, and I was wondering if and when someone would start doing that =p).

Your focus on thermal weapons is great to hear (again like zxc says :)), because preferences among skilled players are apparently all over the place for good reasons in each case. The affect on enemy susceptibility to hits after being blasted by beams is a good consideration. Thermal will even be getting a little better in the next version, so bonus for you there! Thermal weapons can stand to be better since using them requires energy you need for other parts, especially force fields.

I do need to address the meltdown issue, since it's still too easy to fry enemies that way. That should be addressed in Alpha 4.

I find that the most grindy part of the game is to get your hands on enough Hcp. Storage Units to survive up to the end game. You can download schematics starting at -9, but the food clock starts at -7. Effectively you have two levels to download the schematic, get a matter storage unit, and stumble on a low-level fab. Index(Fabricators) rarely works. Then you sit on the fab and milk it repeatedly. By the time you hit Factory, the combination of high-level terminals/fabs, food clock and corruption means that fabbing (hacking, really) is not really an option anymore. From this point on you have to rely on luck to find the storage units or the hackware to make fabbing possible again. I really hope this gets addressed.
Yep, it will be addressed from multiple angles. The first and most obvious change will be that the effect of corruption on hacking will be vastly reduced.

Congratulations on your win!
Title: Re: Combat Run Theorycrafting & Open Discussion
Post by: zxc on October 12, 2015, 10:27:03 PM
20% reduction is certainly nothing to sneeze at, but you'd need 3 slots if you wanted it to work accross all three damage types (unless you swap parts).
I am of course a great advocate of part swapping. 100 time units spent to reduce damage received by 20% is well worth it (especially when you can do this after previously moving, letting your accuracy increase). Also, you often know what you are about to face e.g. programmers getting dispatched, so you can equip the best resist item ahead of time.

The kinetic weapons I found had higher integrity so I kept them around longer than I should have.
That might be another point for kinetic weapons. Another reason I like them is because they are just so common and easy to replace. My kinetic weapons would rarely get destroyed because I would often replace damaged ones with brand new ones off dead enemies.

Thanks for the guides and explanations you've posted in these boards. It was very helpful to me!
Sure thing! I love theory-crafting in a game with depth like this and that is also new and not worked out. I think the biggest change I'm going to make in light of your post is to try using more propulsion slots. They are a poor man's armour (and go well with armour in any case), help prevent scenarios where you run out of propulsion, and reduce your move delay. I think if you are lucky with items, it is still optimal to have the minimum number of propulsion slots, as armour has better coverage and integrity still. But it's rare to be so fortunate with RNG.

Oh and I forgot to say: congrats on your win :)
Title: Re: Combat Run Theorycrafting & Open Discussion
Post by: Kyzrati on October 13, 2015, 05:41:45 AM
Oh hey, I only just now noticed your avatar, zxc.

How nice of mendonca to make the perfect custom avatar just for you 8)
Title: Re: Combat Run Theorycrafting & Open Discussion
Post by: Decker on October 13, 2015, 06:34:02 AM
I read in the manual that the hit chance is 10% lower if the target is flying. I am unclear on the interpretation of "flying" in this context. Does having an active hover unit count as flying? Do you need to be moving? Does that apply to shots against Cogmind?

My current propulsion setup is 3 legs plus 1 hover unit that I fetch from fallen programmers. My idea is to switch to the hover unit in combat to reduce incoming damage (paired with an +dodge item eventually, which IIRC defines flying as hover or flying propulsion). I activated full stats and there appears to be no difference whether the hover unit is active or not.
Title: Re: Combat Run Theorycrafting & Open Discussion
Post by: Kyzrati on October 13, 2015, 06:46:57 AM
"Flying" is when it says that in your HUD, i.e. you (or they) have active flight propulsion. So hover doesn't count; it's not as mobile. For the flight bonus you don't have to be moving--that's a separate modifier.

Pretty much all mechanics throughout the game work identically between Cogmind and other robots (corruption is a notable exception--you're inherently more resistant to its accumulation than they are).

Maneuvering Thrusters do work for either flight or hover, though, and even legs (to a lesser extent).

Your strategy would have to use a flight unit to get the highest bonus you're looking for.
Title: Re: Combat Run Theorycrafting & Open Discussion
Post by: Kyzrati on October 13, 2015, 07:01:17 AM
That said, in light of that particular usage it might be necessary to change the rule such that flying builds only get the dodge bonus if they moved the previous turn, or perhaps get half the bonus (+5%) if they were stationary.

I don't mind Cogminds getting benefits by swapping around free propulsion, but a combat build carrying one flight unit for the dodge bonus is a bit too meta, and outright better than either Cloaking Devices or Maneuvering Thrusters, not to mention even more effective when combined with them.

Too meta, don't you all think? Or are there enough mitigating circumstances in this case that it's not unquestionably the best choice? (E.g. flight parts are weak, requires extra propulsion rather than a more versatile utility slot...)
Title: Re: Combat Run Theorycrafting & Open Discussion
Post by: Decker on October 13, 2015, 07:30:14 AM
I'll let you know once I tested it in real conditions. So far I see three problems with this setup:
1) Flight units are rare.
2) Flight units are fragile and have high coverage (so are cloaking devices and maneuvering thrusters).
3) Flight units provide no recoil bonus.

Title: Re: Combat Run Theorycrafting & Open Discussion
Post by: Kyzrati on October 13, 2015, 07:45:50 AM
1) Every Swarmer carries two, and Swarmers are not rare ;).
2) I was looking at the stats and while this is true, the listed alternatives have about twice as much coverage on average, so are twice as likely to be hit.
3) Very good point, and definitely a deal-killer for kinetic users. Viable for thermal/EM, though.

In any case, we'll see how you fare :D
Title: Re: Combat Run Theorycrafting & Open Discussion
Post by: Happylisk on October 13, 2015, 10:42:52 AM
A late game build with all EM/thermal and 2 treads + 1 flight unit to switch to during combat is pretty cheddary.   Cloaking devices and thrusters are rare - flight units are def. more common, and I would gladly swap out one less utility slot to make my third propulsion slot a substitute for cloaking.  I'm definitely in favor of changing that. 

You could make the flight dodge penalty not apply if you're overweight. 
Title: Re: Combat Run Theorycrafting & Open Discussion
Post by: zxc on October 13, 2015, 09:14:33 PM
Oh hey, I only just now noticed your avatar, zxc.

How nice of mendonca to make the perfect custom avatar just for you 8)

It really is 8)

You could make the flight dodge bonus not apply if you're overweight. 

This is a nice, simple solution. One side-effect I can think of is when you have a flight build and come across some swarmers, and you equip your explosive launcher which puts you overweight and loses you the dodging bonus. That might not be a bad thing however.
Title: Re: Combat Run Theorycrafting & Open Discussion
Post by: Kyzrati on October 13, 2015, 10:18:58 PM
Elegant solution, indeed. It could have the stated launcher reaction side effect, but I agree that's not really a bad thing.

This talk is also making me want to have a composite defensive mod indicator on the HUD...
Title: Re: Combat Run Theorycrafting & Open Discussion
Post by: zxc on October 13, 2015, 10:24:16 PM
This talk is also making me want to have a composite defensive mod indicator on the HUD...
Now that you've brought it up, we will not be satisfied until it's added!

I can live without it for some time though. Not a critical feature, but worth thinking about where it would go.
Title: Re: Combat Run Theorycrafting & Open Discussion
Post by: Kyzrati on October 13, 2015, 10:30:07 PM
Yes, I'm wary of saying anything unless I intend to add it ;)

I believe such a feature would be very helpful for players who are still learning the game--less so for those who already know about the various factors that contribute to evasion. Not critical, though, which is why it wasn't included to begin with.
Title: Re: Combat Run Theorycrafting & Open Discussion
Post by: Decker on October 14, 2015, 08:29:00 AM
Upon testing, I think that the 10% dodge bonus for using one flight unit in a combat build is a little abusive. I found biomechanical wings which had a coverage of 5 and integrity of 20. With such low coverage, those wings last for a very long time even with low integrity. Losing some coverage/integrity and recoil benefits is an issue however.

I like Happilisk's solution to the problem. I think the bonus should only be reduced by half when overweight though. I also like the solution of the bonus not fully applying when stationary.

I found maneuvering thrusters with coverage of 80 and integrity of 10. That amounted to 6% of my total coverage. So the part lasts for about 16 incoming shots on average. IMO that's useless for a combat build. I suggest boosting the integrity a bit and massively reducing the coverage (same for regular cloaking devices and many of the other utilities). Perhaps making the thrusters uses a propulsion slot could lead to interesting dilemmas on slot usage.

More generally, I find the combat utilities only become useful in the late game. Earlier in the game, Cogmind is slot-starved, total coverage is lower, and utilities provide a pesky bonus like 4% targetting, 5% cloak, 6% dodge. In the late game, Cogmind has better coverage so utilities last longer, and do provide a meaningful bonus like 12% targetting, 20% cloak.

I tried the flight unit trick along with two advanced cloaking devices (+50% total dodge bonus) and it did make a noticeable difference. Hunters had about 35% hit chance. I should have waited and collected 3 cloaking devices and a force field to really make the most of it. Individually, I think advanced cloaking devices do not last long enough to be really useful.

If you find the right gears, it's fun to fight on wings. I fought a while on four slots of wings, killing and outrunning grunts. There are not enough wings around to keep this going forever though. Eventually I went back to legs.
Title: Re: Combat Run Theorycrafting & Open Discussion
Post by: fernsauce on October 14, 2015, 11:35:27 AM
Maybe 50% bonus if you're overweight, but only if you are still within like twice your weight capacity. Anything more than that and you presumably aren't going to be doing much maneuvering.

Do cloaking units stack? I was pretty sure they didn't, but I honestly haven't really tried to test it. That would seem a little broken though.

Fighting with wings is interesting. You really want to make the most of your high movement, fighting only in locations that heavily advantage you like narrow chokepoints and using the cleaner bots as robot shields. Point blank shots with overloaded energy weapons / multishot weapons can instantly destroy a lot of the less durable robots like programmers if you build well. Ideally, you use your better tactical positioning and overkill to minimize damage taken, especially with items like Particle Chargers and such. But it's still hard because no matter what you have a lot of items that if destroyed can really cripple you, and trying to overcome weight problems with redistributors just means you have more equipped items that are liabilities (and weight redistributors that give more than 10 mass support are difficult to replace).

The upside, of course, is that you can do a hell of a lot more with 20-50 move delay than you can with the 200 or so you'll get with a legs / treads build, including avoiding nastier encounters and hitting up way more interactive objects and having more time to loot from stashes.
Title: Re: Combat Run Theorycrafting & Open Discussion
Post by: Kyzrati on October 14, 2015, 07:07:09 PM
Regarding the flight bonus, I'd most likely want to go the simple route, rather than adding too many conditions or special cases. You either have a bonus, or you don't. How often are you guys running flight builds overweight, anyway? As soon as you're overweight you lose a significant amount of speed, the whole purpose of flying in the first place. The nerf is targeted purely at abuse by heavy combat builds, with a side effect that seems to make sense for fliers, too.

I found maneuvering thrusters with coverage of 80 and integrity of 10. That amounted to 6% of my total coverage. So the part lasts for about 16 incoming shots on average. IMO that's useless for a combat build.
You could say they're not meant for combat builds :). All evasion-enabling parts in general have relatively high coverage compared to other devices, because otherwise they're waaaay too good.

This is the same with Cloaking Devices. If you can manage to keep them protected (utility shielding? force fields? repairs? spares?) they're quite good, but designing them last longer on their own would turn them into a "must-have" part. We want to keep the number of general must-have parts to an absolute minimum.

The design trade off was: Do we make a part that provides a smaller unexciting benefit and lasts a long time, or one with a really noticeable benefit that you'll have to work to keep functioning?

Do cloaking units stack? I was pretty sure they didn't, but I honestly haven't really tried to test it.
Absolutely not. As you say that would be too effective.
Title: Re: Combat Run Theorycrafting & Open Discussion
Post by: Decker on October 14, 2015, 07:13:03 PM
Quote

Do cloaking units stack? I was pretty sure they didn't, but I honestly haven't really tried to test it.


... d'oh! They don't, according to the description. Sorry for the disinformation :-(
[Ninja'ed by Kyzrati]

Thus, if I understand correctly, I have to combine flight unit + cloaking + manoeuvering thrusters to get the minimum enemy hit chance. "Dodging" combines additively with cloaking, right? It's not a second, separate roll?
Title: Re: Combat Run Theorycrafting & Open Discussion
Post by: Kyzrati on October 14, 2015, 07:16:29 PM
Correct, all those separate sources do stack additively.
Title: Re: Combat Run Theorycrafting & Open Discussion
Post by: Decker on October 19, 2015, 07:47:50 AM
I'm getting the hang of stealth builds. With a speed delay of 5, the game gets much easier. Whoom!
Spoiler (click to show/hide)

However, the RNG is usually not so generous and I've learned not to rely on achieving those high speeds too much. The main issue that comes up times and again is the lack of heat sinks / cooling units. They're quite rare by Research/Access. I typically have everything I need except the cooling, so I can't speed ahead and lift a lot of mass. Hover units don't require much cooling, but finding fast ones is RNG-dependent.

So where do you get those heat sinks? One option is to fight for them but that's risky and anticlimatic for a stealth build.

Two suggestions for the game:
1) Make the turn timer fractional, so it's possible to know when the turn end and when it's safe to run past enemies without them noticing. Right now I work around this by dancing in front of a door until I see the turn tick.
2) Have the 'wait' action flush the "time energy" bank to zero exactly, so the end of turn is predictable.
Title: Re: Combat Run Theorycrafting & Open Discussion
Post by: fernsauce on October 19, 2015, 02:53:15 PM
Jesus Christ that is a lot of propulsion slots. I didn't think you could get below 6 delay, but I guess hover units are magic. I tend to kinda neglect them in sneaky builds just because not being able to hop over robots always ends up getting me killed.

I'm pretty sure speeds below 10 are generally unintended (b/c even the best flight prototypes should be 10 speed at most). That said, 10 move delay is more than enough. I tried a speedrun recently and got through Factory / Research with >10 move delay in under 1000 turns, although I ended up getting killed in -1 because of some very poor decisions.

I can't say I've ever had a problem finding cooling. Programmers, sentries, and transport escorts all have em. As long as you aren't fighting a lot, you're not gonna lose them to attrition much, and you can get pretty good units before you leave Materials. Although you really don't want to fight programmers because they're assholes, but still, sometimes you gotta do what you gotta do. Just don't fight Hunters ever because they will always fuck you up.

Turn timer is kind of frustrating. Actually, I would personally prefer if it was *more* RNG. I'm not sure how, but it does seem kind of bad that you can just be very careful and reliably ninja past enemies. Ducking through someone's LoS should be risky, or at least riskier than it is now.
Title: Re: Combat Run Theorycrafting & Open Discussion
Post by: Decker on October 19, 2015, 08:34:13 PM
Thanks for the info, fernsauce. I did take the habit of carrying along a good cooling unit in case I need one later. I'll try to do as you say and ambush one enemy.

Quote
I tend to kinda neglect them in sneaky builds just because not being able to hop over robots always ends up getting me killed.

That and you lose the nice +10 dodge bonus you get from flying. What I do is have one flight unit equipped at all time, so that if there's a slowpoke blocking the way in a corridor, I can still hop over it, albeit slowly. At least there's now a shield between me and my pursuers.
Title: Re: Combat Run Theorycrafting & Open Discussion
Post by: Kyzrati on October 19, 2015, 11:27:22 PM
Wow, Decker, you've set a new high speed record under the current system. And certainly a new record for propulsion slots :P

As long as you can stay powered and cool, I can see that being a pretty zippy way to get around since even if you occasionally lose some propulsion the effect is minimal.

I did intend for speeds below 10 to be attainable, so I don't see anything wrong with that if the player can manage it. The theoretical limit is 3, it's just probably more or less impossible to get the necessary parts. We'll see how that does or does not change when I update the propulsion mechanics. (Side note: This thread seems to have veered off course and crashed into our Propulsion Rebalancing (http://www.gridsagegames.com/forums/index.php?topic=274.0) topic.)

Like fernsauce says, cooling is pretty easy to get via ambushes. Fighting is dangerous, but as a fast mover you can almost always do it on your own terms. (Heck, for an indirect method wait for a squad to come by and reprogram some traps, then pick up the leftovers ;))

Of course, you won't get the best stuff that way. There's also the option of fabrication, which I still plan to further improve as a system, but may not be so suitable for a build attempting to focus purely on speed, anyway.

It's interesting to hear two opposing opinions regarding the ability to pass through LOS undetected. Early on I thought about adding a chance to be discovered even while zipping through LOS, but decided the results would end up appearing pretty random, anyway (and more difficult to code), so I left it as is which keeps it somewhat unknown (except in the case of Sentries, mentioned below).

I like the idea that you can never be quite sure if you'll make it, and have to deal with the consequences if you don't.

1) Make the turn timer fractional, so it's possible to know when the turn end and when it's safe to run past enemies without them noticing. Right now I work around this by dancing in front of a door until I see the turn tick.
This theoretically should not work, because hostile robot turns have nothing to do with the turn timer--like yourself they can take their turn in the middle of game turns.

The only time it will work is with Sentries (and ones who have never moved, at that--once they've moved all bets are off), since they just sit around and wait from the beginning of the map, matching their turn exactly to the game turn counter.

As such, your second suggestion would really only be useful in the Sentry scenario, so not something generally applicable enough that we can make it a thing.

Regarding the wait command, there are a few new features to come later, including things like "wait X turns," "wait until something interesting happens," "wait until the actual tick a hostile comes into view" (i.e. "tactical overwatch (http://www.gridsagegames.com/forums/index.php?topic=234.0)"). These features will be either easy to execute or outright automatic where they can be.

What I can do is attach your suggestion to the "wait X turns" feature and the command will always stop on an actual game turn, meaning you could use it to wait 1 turn and get your desired effect.