Development Info Underrail 2: Infusion Dev Log #9: New Combat System

[Ol' Willy]

Turd based gameplay for simulating tactical engagements is abomination that turns RPGs into slow, banal, exploitable slogs for no lifers and he wrote entire post explaining it only to miss the point, forest for the trees indeed.

There is a reason why wargames with tactical depth ten times more than in average RPG are all turn based

Please share more pleb opinions

 

[Ol' Willy]

Not a fan of such game systems but I see a lot of potential. Given the basic premise of the system it's obvious to expect the feats that will allow you to use it to your advantage. Time warping psi powers may get even more OP

 

[Oreshnik Missile]

Single biggest problem in Underrail 1 combat was cheesing the AI with LOS abuse. I should think that has been eradicated.

 

[BlackPlate]

Single biggest problem in Underrail 1 combat was cheesing the AI with LOS abuse. I should think that has been eradicated.

Are you referring to breaking LOS by running around corners or obstacles? That is a pretty good simulation of being fast and fighting tactically. It can get wonky and cheesy, but there is no better way to simulate that with the current combat system.

 

[Stavrophore]

They should at least throw flashbangs near the corners if you hid behind them.

 

[Crescent Hawk]

Over the years, however, I came to realize that, while it's a solid way to approach party-based combat, it has some significant disadvantages when it comes to single character RPGs

He was this close to getting how to fix it properly, but missed the forest for the trees.

I do agree with you yeah, even if in the end I will probably enjoy my time with it. Like this the game just feels like a blueprint for a System Shock type game. A party based game would be way better. Nevertheless Iam impressed by the looks and ambiance.

 

[Cleveland Mark Blakemore]

That engine looks superb.

Throwing out turn-based combat for this game makes no sense. Making it just another clickalot rogue-like totally removes the appeal of the original game.

Otherwise, that lighting engine is pretty nice. The prototype looks good.

 

[Crispy]

Looks neat.

Not sure why you play at such a low resolution, though. Framerate problems?

 

[BlackPlate]

They should at least throw flashbangs near the corners if you hid behind them.

I would love to see that and more, but it's probably a never-ending spiral of 'just one more thing' that gets harder and harder to program the AI to do in any acceptable capacity. Also entering the CDDA autism-tier of 'simulating' til it gets bogged down and becomes unfun. I am happy with what we have, it doesn't become ezmode on Dominating even if you know all the little things.
The new combat looks like some bizzaro shit to me, might just be strange enough to be great and fix this.

 

[Stavrophore]

They should at least throw flashbangs near the corners if you hid behind them.

I would love to see that and more, but it's probably a never-ending spiral of 'just one more thing' that gets harder and harder to program the AI to do in any acceptable capacity. Also entering the CDDA autism-tier of 'simulating' til it gets bogged down and becomes unfun. I am happy with what we have, it doesn't become ezmode on Dominating even if you know all the little things.
The new combat looks like some bizzaro shit to me, might just be strange enough to be great and fix this.

How is programming ai to do a wide berth of corners and throwing flashbangs with a simple logic of [if enemy behind corner && no enemies inside grenade radius with reticle throw placed at corner = throw grenade] something unreasonable? The problem with making combat realistic and hard and AI acting believable is that UR games feature just one character. Too smart AI or too well usage of disabling weapons will mean GG for most chars that dont have thick skull or some perk like that. Similarily AI could do a wide berth and employ suppression fire, ie. shoot at the corner and PC would just get penalty to AP/MP when going out of the corner instead of getting disabled. There are ways to deal with breaking LoS, but it doesn't mean they will make game enjoyable for the given game constraints like one player char. Everything has to be balanced properly, realism should not be goal in itself, the enjoyable gameplay loop should be. If realism doesn't infringe upon that too much, then by all means go for realism, otherwise please do not try to fit something that won't fit.

 

[Ol' Willy]

They should at least throw flashbangs near the corners if you hid behind them.

I would love to see that and more, but it's probably a never-ending spiral of 'just one more thing' that gets harder and harder to program the AI to do in any acceptable capacity. Also entering the CDDA autism-tier of 'simulating' til it gets bogged down and becomes unfun. I am happy with what we have, it doesn't become ezmode on Dominating even if you know all the little things.
The new combat looks like some bizzaro shit to me, might just be strange enough to be great and fix this.

How is programming ai to do a wide berth of corners and throwing flashbangs with a simple logic of [if enemy behind corner && no enemies inside grenade radius with reticle throw placed at corner = throw grenade] something unreasonable? The problem with making combat realistic and hard and AI acting believable is that UR games feature just one character. Too smart AI or too well usage of disabling weapons will mean GG for most chars that dont have thick skull or some perk like that. Similarily AI could do a wide berth and employ suppression fire, ie. shoot at the corner and PC would just get penalty to AP/MP when going out of the corner instead of getting disabled. There are ways to deal with breaking LoS, but it doesn't mean they will make game enjoyable for the given game constraints like one player char. Everything has to be balanced properly, realism should not be goal in itself, the enjoyable gameplay loop should be. If realism doesn't infringe upon that too much, then by all means go for realism, otherwise please do not try to fit something that won't fit.

Have you ever played Underrail? You either can tank stuff thrown your way, or you have the means to not be targeted in the first place. In the latter case, if you failed and have shit on you, you need to have a plan b to disengage - or just reload

 

[Stavrophore]

Have you ever played Underrail? You either can tank stuff thrown your way, or you have the means to not be targeted in the first place. In the latter case, if you failed and have shit on you, you need to have a plan b to disengage - or just reload

The issue we were discussing is to make the ai smarter and less cheesy, like hiding behind pillar 1 tile wide and they literally doing nothing but conga line going to you. Not debating whether you can tank, evade or whatever, just wondering how we can make AI more engaging.

 

[Ryzer]

Single biggest problem in Underrail 1 combat was cheesing the AI with LOS abuse. I should think that has been eradicated.

"Single" lol, you are wayyy too nice.

 

[Roguey]

What overbalancing does to a mf.

 

[BlackPlate]

How is programming ai to do a wide berth of corners and throwing flashbangs with a simple logic of [if enemy behind corner && no enemies inside grenade radius with reticle throw placed at corner = throw grenade] something unreasonable?

It isn't, that is why I said I would love to see it and similar solutions, but you yourself answered in your own post why it probably isn't in there.
To add to that, you would have to account for enemies that don't have non-targetable AoE attacks they could use to flush you away from corners, and even if you add something like suppressive fire penalty to AP/MP, the solution would be just moving deeper in around corner to avoid it. It would add more to the usual list of moves you would have to do in order to be safe, but such things can quickly enter the realm of tedium rather than any real danger to you. The more you try to add to combat depth the more you have to compensate for player just doing the bare minimum to avoid it and if that is even worth it, and every new thing itself has to be something that doesn't add to the cheese.
This is what I meant when I said 'just one more thing' and 'unfun'. 'nades around the corner is good, but it effects a relatively small amount of combat encounters, and as a counterpoint, that itself could be used to deplete grenades from enemies in a safer environment. You would need to overhaul the game's combat in order for it to be impactful, which Infusion seems to do with its tit-for-tat combat rather than full turns.
I wonder just what exactly are we discussing here?

 

[damager]

Don't make every enemy wear ten granades and give the player abilities and feats to counter such tactics (like throwing granades coming around the corner right back). Also Sperr-Feuer should deplete Amo

 

[Israfael]

Sounds a lot like more nuanced SPM combat from Brigade E5: New (Jagged) Alliance, did you play it perchance, Styg ? There everything happens simulatneously in real time, but all actions have certain time allocations and the outcomes are dynamic (i.e. if you shoot when the enemy's muzzle flashes during a night time encounter, you'll get higher chance to hit and vice versa).

 

[Stavrophore]

Jagged Alliance despite turn based, still uses interrupt mechanics, where if both your character and enemy find themselves in field of vision, the one with higher interrupt chance will act, its recalculated after each action. Idk about brigade i've played it long time ago for brief amount of time, but i believe its realtime with pause.

 

[Israfael]

Idk about brigade i've played it long time ago for brief amount of time, but i believe its realtime with pause.

It is realtime, yes, but it has extensive trigger system that covers almost everything in the game and it has action queue so you can basically play it as simultaneous turn-based game (the devs called it 'smart pause mode')

 

[AdolfSatan]

As someone freshly off playing Underrail for the (almost) first time, I think Styg’s self-critiques are spot-on: combat is fun, but there are plenty of situations that fall into the same pit that KOTC2 in which it is essentially defined by who wins initiative with no chance to make up for it. Looking forward to this new iteration.

 

[Grampy_Bone]

I find this funny because when I pointed out in the Underrail thread that the game hinges on first turn alpha strikes I got ratio'd by 'fake news' reactions and such.

 

[Stavrophore]

I find this funny because when I pointed out in the Underrail thread that the game hinges on first turn alpha strikes I got ratio'd by 'fake news' reactions and such.

I mean that's both true and untrue. Often groups are too big to destroy them within first turn even if you have big alpha , you then need some crowd control, escape so lots of MP, or using stasis.

 

[The people of Duckburg]

I'm not too happy about the new combat system. Sure, it makes alpha strikes less important, but it has its own set of problems when it comes to game balance. For one thing, it's highly likely that glass cannon builds are going to be very weak, possibly even non-viable. You spot enemies, try to take out one enemy or CC some of them, only to get critically shot and killed or CCed by another enemy in the group right as you finish your action. It seems likely that it will be impossible or close to it to avoid taking damage regularly with a system like this. You're at even more of a disadvantage in group fights, excluding ones where you act last of course, since your ability to exit cover, hit one enemy and return to cover without taking damage seems to no longer exist. The system reminds me of RTwP, except with automatic pausing. And I really dislike the chaos that RTwP brings with it.

Obviously, this is all speculation. I could be entirely off the mark. We'll see. At least the increased interactivity should be a very nice improvement. Hopefully we can still speed up the game to not have to spend a fourth of the playtime cutting fences, though.

 

[Sweeper]

Turd based gameplay for simulating tactical engagements is abomination that turns RPGs into slow, banal, exploitable slogs for no lifers and he wrote entire post explaining it only to miss the point, forest for the trees indeed.

There is a reason why wargames with tactical depth ten times more than in average RPG are all turn based

Please share more pleb opinions

That's because they're easier to make, not because TB is inherently superior.

I find this funny because when I pointed out in the Underrail thread that the game hinges on first turn alpha strikes I got ratio'd by 'fake news' reactions and such.

That's probably because there are tanky builds that don't rely on needing to kill/CC everything first turn.

 

[Grampy_Bone]

That's probably because there are tanky builds that don't rely on needing to kill/CC everything first turn.

The game's developer disagrees with you. Even when I played Tin Can I still pretty much had to either kill or lock down every enemy the instant they came into range. The thread regulars also commonly tell everyone to play stealth regardless of build in order to guarantee first strikes.

In any case, Underrail's action pool is absurdly high for a single-player turn based game. I can only think of one other TB game where the player can make 10+ attacks per round (Arcanum) and it's also hilariously weighted in favor of first strikes. It's the problem of wanting to give the player a wide array of options per turn but also maintain system parity between player/enemies. The real solution is to make it party-based but as others have pointed out that's not in Underrail's DNA. You could also acknowledge that enemies are loot pinatas and not other players and not give everyone 50 AP plus 20 move points but that doesn't appeal to a certain kind of design autism.

I'm not too happy about the new combat system. Sure, it makes alpha strikes less important, but it has its own set of problems when it comes to game balance. For one thing, it's highly likely that glass cannon builds are going to be very weak, possibly even non-viable.

If it turns out anything like other time-unit based semi-real time games (e.g. ADOM) it will make speed the end all stat that dominates all other choices.