Why are you bringing it up in a conversation about bots’ efficiency then ?
I pointed out stupid bot behaviours and you first reaction was
4 Likes
Either I’m exceptionally lucky or most players I encounter in Legend and Cata can dance with bosses, at least relatively well, which is as good as or better than the bots who stand in aoe effects from bosses and engage them in terribad positions (with players you can at least tell them to get out and engage the boss elsewhere, most seem to listen, rarely had one or two not do that - with bots if they’re the entire team it’s the same as nobody on the entire team listening).
Sometimes I get the odd QPer that can’t or that positions themselves so the boss can knock them off a ledge every time, but those experiences stand out to me since it doesn’t happen that much.
I have seen three bots go down to a single rat ogre overhead several times lately too, not to mention get gobbled up by a chaos spawn which they never used to do unless cornered, plus lately they seem to eat hits from all manner of trash mob, Exec Kruber Merc is especially guilty of this to the point I moved him to other weapons (and he used to do okay with exec, now he heavies single unarmoured enemies while eating hits from others).
Against a boss alone bots normally work well but they’ve got nothing on humans when it comes to a boss mixed in with other threats. You have to keep them clear as their prioritisation means they’ll eat gas clouds hooks etc. around the boss while failing to deal with the horde - or sometimes some will randomly break targeting to deal with a gasrat but ignore the boss and horde for a few seconds while doing it (had a bot Kruber switch from shield to repeater handgun while standing next to a chaos spawn that had the munchies for him already and with trash chaos enemies attacking him, he died very fast). He shouldn’t have weapon switched in that situation at all.
How is it bots go from worse than players to better than players depending on which post of yours one reads. If they’re better than players as you say surely it’d take less skill to beat a map with bots?
Hahaha, a child’s strike, can you do no better?
6 Likes
Kinda miffed at how people are so damn paranoid that bots are going to start making Cata runs a joke if they get buffed.
No argument here but just a thought feel free to call me out for it:
Maybe, just maybe, people dont like the fact that less skilled players could get a better experience with bots and make it through Legend+ because that would be catering to the casuals!!!
2 Likes
Its more the fear of people getting even more aggressive towards new players. Why give them a chance, if the bot will do better every time?
Not sure I’ve seen anyone arguing for bots remaining brokenly dumb come across as new player friendly with a warm welcoming attitude.
Those same hypothetical new players will sometimes find themselves playing with bots and wondering why the bots are all grouped up standing on top of each other and running face-first into warp fire until all 3 are dead. Or the new player will get downed, the bots will all walk over, stand on top of the downed player, then never revive them.
I’m not saying your wrong; a new player right now that struggles with Champion would still be more useful than a bot if dragged into Legendary games. But having done this with a friend who’s new to the game, I’m not sure he actually has fun beyond just hanging out in voice chat and spending most of the maps dead, just like a bot would.
1 Like
Are there really players that shitty in this community ? In over 1100 hours I’ve encountered one russian dude yelling in his mic and one player who wanted to votekick a player for not taking a tome and proceeding to shoot us when we told him to chill. 2 players, that’s not even a statistic.
1 Like
There are but usually the ones I have experienced were the bad players on the team and mostly seemed to want to be carried. Not saying I haven’t seen new players that didn’t know what they were doing on higher difficulties, but usually the really good players on the team didn’t say much beyond giving them advice, they just silently carried.
But I don’t stay in games where the host is being an arse or half the team are.
As for making the bots too good (not directed at you), why don’t we make them not stupid first and then adjust them from that point? At present we have a big gulf between the bots and intelligent, it is not like there is no room for FS to play around in.
6 Likes
Exactly, it will be much easier to balance them anyway once their behaviour makes sense / is consistent.
I’ve been in bot games where ironbreaker and I have like 200 damage taken, and then the other two bots have like 2,150 and 3,400 damage taken. Oh and we obviously lost with ironbreaker being the last man standing. Looking over me as I get my face bashed in by the local hooligans. His shield up. Just watching. It’s wild.
1 Like
So just asking for curiosity’s sake. Has anybody actually noticed much difference playing with bots after the bot improvement patches a bit before Chaos Wastes? Can’t say I noticed much to any difference myself, but I haven’t had much time to play so would be curious how impactful people think the last lot of bot improvements were.
2 Likes
The only really obvious improvement that I see is that bots are now way less likely to throw themselves off ledges when a gunner or flame rat shows up, at least on the standard maps - I haven’t used bots much in the wastes.
2 Likes
@DezZzO did have this comment, but yeah it’d be interesting to know if anyone else had anything to chime in there.
I do know from having last played prior to GK and having returned post CW that bots, to me, have definitely been much worse overall. I can’t speak to the bot improvement updates.
I suspect if DezZzO is correct that FS accidentally rolled back the bot improvements - I’m not sure how their version management it but it would coincide with how old bugs often got reintroduced in the past in later updates after being fixed.
2 Likes
Thanks for highlighting that comment, I’d forgotten about it.
My last CW run my Saltz bot threw himself off a cliff with literally nothing around then for some reason died about 3 seconds before we could grab him up. So yeah they’re not great with cliffs in CWs lol. Good to hear they’ve been better in adventure maps.
To be honest my dudes, I’m convinced at this point that FS is either incapable or completely unwilling to make meaningful improvements to bots. I am very seriously considering making a post to try and commission a modder to make a V1 style bot mod and just try to get FS to agree to sanction it. Ironically I see that final step with minimal involvement/work from FS as likely being a bigger hurdle than the rest of of it 
4 Likes
I said it before but that is very unlikely to happen. You have to note that the Bot Improvement mod (even before it didn’t get updated anymore) for Vermintide 2 never came close to the level of the Vermintide 1 Bot mod. Why do you think this is? There are two more likely scenarios for this:
- The modder was not able to create a meaningful Bot Improvement Mod due to Vermintide 2 being a more complex or at least more cluttered game.
- Some “Improvements” the bot contained where actually cheats (as written by Robin) and had to be removed for being sanctioned in Vermintide 2 hence resulting in a less effective Bot Improvement mod.
There are also other probable explanations and I will not exclude being wrong about this.
Also, I will be frank. Bot Improvements are very low on my personal priority list and as such I oppose any more improvements here. Why? Where is the harm? Well, bot improvement eats up developer ressources, actually developer time (in contrast to cosmetics which are firstly an artists work or lore post which is … someone’s work). There is other stuff which has higher priority for me where I would like to see these ressources spent.
Bot users have complained and they got their improvement, yet they want more. Some people will never be satisfied no matter how good they would be. Also some “suggestions” here are kinda ridiculous in the perspective of benefit/investment. Self-learning bots? Great idea, just brutally complex. Bots which behave according to their class? Also, fantastic idea but you would need to adapt AI to 20 different careers, not including loadouts (you could probably dumb it down to 4 to 5 archetypes). Bots scaling with player/difficulty? Again, to complext to program.
I did Legend Deeds with the bots this week and they work well and snipe specials because I can’t do it (and Ironbreaker cant do it too). I have not seen much issues with them to be honest. So I also don’t see the point to make them better because they shouldnt be an equal choice in Cataclysm (dont jump all on this one comment, my main gripe is that I want developer ressources spent elsewhere). For Chaos Wastes Bots are more awful by design. But this is mainly an issue to the complex navigation meshs caused by the Ledges and the different pathing options which make certain areas one time there one time not. Remember that bots see the game differently than us. The same goes for AoE attacks.
So yea, I am against further work on Bots. By all means, push for better bots. It is your right, especially if it is important to you. However, do NOT deny my right to advocate AGAINST bot improvements because I want to see the developer time spend elsewhere for (hopefully) stuff that is higher on my personal priority list.
You can try to go the “commissioning” route. However, I think it should go away from outright making them better and more towards the ability to micro them more. Either by having them listening to the social wheel or giving them some kind of Gambit system which would solve most problems people have with bots.
1 Like
I’m assuming you mean Patch 4.2 and Patch 4.2.2. Yes, the changes by those patches are significant and noticeable.
I find that these either all work as described or if not, at least work better than pre-patch. They still occasionally facetank warpfire throwers and ratling gunners, but at a lesser rate than pre-patch. Still needs further tweaking.
“Stick closer when under high threat” works better than pre-patch. Although this behavior can be overridden by the “chase after specials with melee weapons” bug already described above. “Being more aware of pinged targets” seems to work better than pre-patch. “Bots should only use ranged weapon against targets that match weapon type eg. they won’t fire handguns at infantry” definitely works. Previously, they would shoot at Skavenslaves with ranged weapons such as handguns, wasting ammo. “Bots should no longer run off cliffs if they are in ratling gunners fire cone or other long sticky AOE threat” seems to mostly work (in regular mode). Chaos Wastes pathing seems wonky, they could still trip over ledges even if there is no enemy attacking.
Bot healing works better than pre-patch. Most of the changes here are the same as the ones in Patch 4.2 - they work better than pre-patch but could occasionally not work too well. For example, even if reviving behavior is better than pre-patch, there are still times where they will ignore revives. Needs further tweaking.
A core bot bug that is left unfixed is the behavior where bots have a suicidal drive to chase after specials with melee weapons, which still overrides many other bot behavior.
I’ve already replied to this above. That post by Robin was from 2017 (before Vermintide 2 was even announced) and is specifically about Vermintide 1. It is hardly relevant to V2.
There isn’t much modding activity on bot mods by V2 modders even on unsanctioned (and potentially “cheaty”) mods. There are only 2 bot improvement mods that are a continuation of the V1 bot improvement mod’s features: Bot Improvements - Combat and Bot Improvements - Impulse Control. The latter has been “Sanctioned Status: [applied]” since 2018.
I think the actual reason for V2 bot improvements mods having less functionality than their V1 counterparts is lack of modder interest. The modder of Bot Improvements - Impulse Control pretty much sums it up in their comment made back in 2019 (emphasis mine):
I haven’t tested this with recent versions of the game, so the chances are it got broken at some point by an update to the game. In particular, if you’re on the beta branch I wouldn’t be surprised if it doesn’t work.
I probably won’t try to fix it unless there’s a reasonable chance it will get sanctioned, and I don’t [think] that’s likely for the foreseeable future.
They are correct. Fatshark has not reviewed new mods for sanctioning in a long time. Modders lost interest as a result.
[Citation Needed] Complete speculation. We don’t know that. I’ve seen no comment like that at all from a bot mod author.
6 Likes
In each of my posts I made clear that Robin’s comment is about Vermintide 1. And yes, it is relevant for the discussion because people have drawn comparisons between the Vermintide 2 bots (modded and unmodded) and the modded Vermintide 1 bots. To which I reply that the modded Vermintide 1 bots obviously are better because they cheated. As such the efficiency of Vermintide 2 bots should never be compared to the modded Vermintide 1 bots. That is all I say. If you don’t draw this comparison then good for you.
That only explains why there is not much activity NOW. It does not explain why the initial and sanctioned Bot Improvement mod was noticeably worse than the Vermintide 1 mod. There has to be a reason why the later created mod for Vermintide 2 has been worse. It is 100 % not that FS stopped sanctioning because that wasn’t apparent at that point in time.
Of course it is speculation. I even wrote that in my comment O.o It is one of two potential explanations why the initial bot improvement mod fell so short in comparison to the one we had in Vermintide 1. It does not exclude the true reason being another option which … I even wrote in my comment.
1 Like
I understand and appreciate this perspective. It’s partially why I was trying to think of a method that would outsource a large amount of the work involved (at my own and other like minded community members’ expense, no less). I get that the approval process would still require a decent amount of resources but it seemed like the closest to a happy medium I could concoct.
Hence why I asked what peoples’ experience with the improvements actually are. I’m aware having mostly played CWs recently my view on them will be skewed, but the adventure maps I have played the impression I got was that there wasn’t a real or noticeable improvement in how they behaved. Seeing mixed results on how people have found them since the updates, but certainly welcome more perspectives on their performance nowadays.
Regarding your various points interpreting FS’s motives and what they would and would not be willing to sanction, fair enough to speculate but to be honest it isn’t really worth anything to me unless it’s coming from FS themself or someone like Grim who might have first hand experience with this. I certainly wouldn’t bother trying to go very far with this route without confirmation from a dev that they’d be willing to sanction it. That would obviously be foolish.
You’ll hopefully notice I didn’t jump on that one point you said not to despite the fact I somewhat disagree 
3 Likes
What is a ‘cheat’? There is only code. Bots either behave well and predictably or they don’t. Labels like ‘cheats’ mean nothing. There is no reason Vermintide 2 bots shouldn’t be allowed to emulate human player behaviour a bit more realistically and actually snipe specials (like the VT1 mod) instead of charging off into the far distance to engage them in melee or standing in a torrent of green flame doing nothing.
Well, I’d argue that making bots even marginally less buggy is a good investment of Fatshark’s resources because I assume that there will be bots in Darktide and any other co-op game that they release and that any improvements they can make will translate to those future games. In contrast, adding more maps to VT2 benefits only VT2.
I’ll be interested in seeing how Darktide bots behave, especially how they handle the balance between using ranged and melee weapons if Darktide turns out to be more shooter oriented.
2 Likes
Well we can’t know for certain unless FS managed to do a survey of existing players and players who have quit, but judging by the amount of complaints about bots I see I imagine it is putting people off the game.
Except the bots are significantly worse since I came back using the same builds as prior. They’re a little less bad vs specials, although I still see them try to facetank fire rats and ratlings, but their melee abilities seem to have been worsened, they eat all kinds of hits they didn’t before. Exec Kruber is the worst offender, he uses lights on armoured and heavies on unarmoured when he worked fine previously, I had to switch him to a much more defensive setup. He used to use Exec properly and well.
This is a fair point. imo using ‘Come Here’ with bots should force them to come over and leash them nearer to you, even if they were chasing specials or a boss. Being able to hotkey ‘Come Here’ would help a lot as well if this was done.
I find Kruber with the Executioner is far worse however, he uses lights on elites and heavies on unarmoured fodder, so not everything seems to be working.
I haven’t noticed them sticking closer at all, they always push to engage bosses in terrible spots and on +Specials Chaos Wastes maps and Cataclysm with its abundance of specials they’re always spending their time dashing away from the team. It is zero problem when I play e.g. Bounty Hunter (although some places like the first hut on Against the Grain are a problem due to LoS blocks), but on GK and other careers depending on my weapon choices it means I have to go in as if it is a solo mission for the most part.
Weirdest of all I see them switch to ranged sometimes vs specials now, usually when there are other enemies around, but don’t use it until they are point blank. I had Merc Kruber terminator-sense a gasrat that spawned in and switch to repeater while the chaos spawn was targeting him amd right next to him, along with a fanatic. He died instantly.
Meanwhile when they have a safe gap around themselves they often insist on meleeing the specials instead.
I heard about it after (I started VT with the VT2 open beta): The VT1 mod framework (from the same guy who made the bot improvements mod) contained plenty of cheats you could activate - free xp, instant mission win, perfect rolls, no overcharge etc etc. Basically, you could alter game functionality with it, some was simple QoL, some was outright cheating. I’m not sure if/what bot related cheats it had, but the potential was there to turn bots unbeatable.
Someone made this later, but I have no idea how it good it is/was.
All that said, VT2 bots definitely could be better, they have been better/worse at various points. The flaw with the VT2 bot improvements mod isn’t that it didn’t work, it is that the modder hasn’t updated the actual bot AI in ages afaik. I’m not even sure it was updated post-WoM.
Pre-WoM it definitely helped a lot more than it does now, it was actually worth taking back then. Apart from tagging enemies and leashing bots more to you it no longer makes them that much better than vanilla - some of their priorities it started to mess up so I stopped using ages ago.
3 Likes
Nobody in this thread is advocating for game breaking changes. They just want bots that are slightly more functional/less bug-ridden, so that they better mimic playing with a human player. VT1 bots mods seem an apt comparison by which to evaluate bot behaviour in VT2 (not xp hacks, perfect rolls, etc.). I am comparing VT2 bots to the VT1 Different Bots Mod.
I think most players would be happy if bots would ping elites, snipe distant specials more frequently and consistently, rez properly, not focus solely on a boss when surrounded by a horde, not fall off ledges, and scale with difficulty so Recruit is still a “challenge”. If we are still limited to 7 viable bot classes and a handful of builds then that’s fine. I just wanna be able to play cata with bots without having to kill 50+% of the specials to win, as this limits me to playing ranged classes because of the ammo drain.
5 Likes
I know. I’m just providing context to the VT1 mods, since those did originally involve actual cheats/alterations to the game that went beyond improving bot AI.
It’d be disingenuous to claim VT2 bots can’t be improved without cheating, they definitely can - they have been better at points, like when the VT2 bot mod still worked better, or when Kruber bot still knew which combos to use when with his melee (in general bot melee priorities/combos have felt more wonky recently than before).
The fact FS did an update to try and improve them would imply FS are, on some level, interested in improving them too, or at least recognise the advantages to doing so and would be capable of doing so.
3 Likes