Everyone is having trouble winning combat runs. So far, it's only been managed once with
an amazing run by Adraius. 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 theoryI'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.
Images are my desktop size - 1920x1080 - so you may need to right click them and view them in a new tab else the items list will be cut off.
1
2
3
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 EvolutionsI'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/UtilitiesOne 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
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.
