Suggesting the release of official gameplay statistics for better balance discussions

Warhammer 40,000: Darktide General Discussion
1

While it’s good to listen to users’ opinions on balance discussions in the forums, I’d prefer that Fatshark base its balance patches primarily on its own statistical data.

Everyone participating in these balance debates here is discussing the issue without knowing various metrics, such as which builds have high or low win rates, which weapons have high or low win rates, which skills have high or low uptime, or what the average DPS is. Of course, I’m no exception.

Personally, I found it hard to understand the recent nerf to the Krak Grenade because, in my experience, Smoke felt like it had better performance and a higher win rate. However, if there had been official data showing that Krak actually had a significantly higher win rate, I would have readily accepted the nerf without any questions, even if it contradicted my personal feel. Having access to these metrics would eliminate unnecessary doubt and help players align with the developers’ intentions.

As such, the meta likely varies by player, and on a larger scale, by region, and everyone has different opinions. So, what I’d like to suggest is
I’d like to ask if Fatshark could release the various statistical data they have.

If we look at the actual statistics generated by players during gameplay, we might find that certain builds, weapons, or skills have surprisingly high win rates—contrary to what we expected—or that some we thought were strong actually have low win rates. We could also see how much the average match duration varies by region and how meta builds differ, which would make it easier to understand the differences in perspective between players from different regions.

I think this could lead to more meaningful discussions rather than users simply arguing based solely on their own experiences.

What do you all think?

2

balancing by numbers, especially winrate, isn’t a good idea because there’s many reasons that can give something a big winrate that have little to do with the thing itself. even usage rates are elevated/killed off by literal memes and youtube algos.

the numbers are only as good as the brain of whoever is actually interpreting them properly. and im not saying people are just stupid but it won’t help people get smarter if they have the numbers either.

it would probably be interesting but it’s also not actually that relevant for these discussions.

3

It’s not that I disagree, but I get the impression that a lot of people on this forum are arguing rather emotionally, based solely on their own experiences.
Of course, I’m not saying we should just nerf a character because its win rate is high or buff it because it’s low, but I think this approach might be somewhat helpful in the balance debate.

4

they for sure do, but i think that’s fine and how balance discussions should go. people only have their own angles to share as feedback, it’s up to the devs to understand where the take comes from and putting it into context.

i think winrates and such would be very misleading to many people because it is legitimately a very complex thing. for example i am certain that arbites with dog has very high winrates compared to most other things because it has possibly the lowest skill floor in the game. yet i dont think it gets mentioned by balance feedback often. but if they showed winrates, and dog arbites is on top, wouldn’t people start focusing on it feedback wise just because they think they should?
meanwhile there are extremely effective builds, straight up OP builds that i think the average player fails with and would have a low winrate as a result, just to mention the opposite. people would feel discouraged or be discouraged by others to suggest nerfing such builds based on that metric

i think this would probably poison discussions a bit if the info was public

also i think fatshark themselves might balance their game better if they had no statistics available to themselves. i think psyker gets buffed a lot because most psyker players aren’t very good and it shows that way on the statistics.

5

Some sort of match API would be cool. There are definitely flaws to relying on these but they’re much better than ancedata.

6

This is exactly what I mentioned in my post.
The meta varies quite a bit depending on the player and the region.

You might expect Arbites with dogs to have a very high win rate,
but I think combining Lone Wolf with Arbites Grenade is a better strategy following the continuous nerfs to dog-related traits.

If two people with such differing opinions were to debate balance, they probably wouldn’t understand each other’s perspective.
Additionally, if Fatshark were to implement a patch based on one person’s opinion, the other person would likely disagree entirely.

In situations like this, I think having objective metrics could be helpful in resolving balance disputes.

Of course, I’m not saying you’re wrong—it’s certainly a valid concern—
but my opinion is that these metrics themselves could provide meaningful assistance

7

I think FS only nerfs weapons n talents that are mostly used by players. Arbite with the talent that limits the maximum damage to 50 definitely got higher winning rate than veteran. But since voc has been more widely used, let’s ignore the actual winning rate then nerf voc.

8

There was a recent poll; it would be interesting to see those results too.

9

While I do 100% agree they should defenitely release the ((files)) stats, there’s many issues with releasing things like winrate at just face value

The main issue is that winrate is affected not only by how powerful the loadout is, but how good the player is, and this hurts both sides

Builds/weapons that are mega OP and extremely popular will have lower winrates than they deserve because there’s players who are bad, learning and/or had bad luck with the team

Meanwhile, builds/weapons that are on the weaker side which nearly no one uses, will have HIGHER winrates than what they deserve because the players commiting to this are experienced and gamesense alone can win a lot more games

Now all of a sudden people will argue “See this is not OP, it has a 45% winrate!” and “Well if this has a 70% winrate maybe it doesn’t need to be buffed and you just need to get good…”

10

That’s right. I agree.
Although I used win rate as an example, I didn’t mean that we should make a patch based solely on win rate.

11

I’ve said it in a different thread, but win-rate is a god awful metric to balance around and tells you absolutely nothing about the performance of said weapon, blitz, ability, or talents.

Let’s say they released the data for 200,000 Veteran games. Of those: 100,000 used Krak and 100,000 used Smoke. Let’s say Smoke had a 42% win rate and Krak had a 38% win rate amongst their respective matches. What did you learn? You think you learned that Smoke > Krak, but what you’re missing is that of the 100,000 Smoke games the 42,000 that won had a premade team with meta builds and the remaining 48,000 were pug failures while the 38,000 Krak wins were all solo PUG carries.

Do you think Krak still worse than Smoke with this new information? I don’t. Smoke needed premade meta builds to win and Krak could Pug and win. The weapons, builds, talents, teammates, missions, AI Director’s Attitude, etc, all play a part in win rate. It’s impossible to determine the performance of a single piece of a puzzle without the whole puzzle. You may as well be balancing based on a roulette wheel

Edit: I recognize that you’re asking for more than just winrate to be released, but even then you can’t extrapolate anything worthwhile from these data sets. Usage rate of specific combinations would be the most you’d really get, and that’s an even worse metric to balance around.

12

I know, it’s just that many will inmediatly jump into it as “arguments” and it’s gonna be tiring explaining to them why the info is skewed

13

Why do you think only people who’ve pre-made teams use smoke?

I play Havoc a lot with strangers, and I actually find smoke more useful in those situations.

Of course, it’s only natural that some people, like you, don’t agree.
Seeing this makes me curious about the stats comparing when you play with a pre-made team versus when you don’t.

I admit that the stats I’m looking for might be inappropriate for a balance debate, but since we’ve never done this before, I thought it might be worth a try.

Players have been discussing balance based on their personal experiences, but if you ask me whether that has actually contributed to achieving good balance, I honestly don’t know.

14

I don’t, that was an example fabricated data set to show how there’s so much more to the picture than just “how often is it picked and how often does it win when picked,” and how just looking at that data can be misleading without further information.

For example: Play 5 games with the exact same loadout each time and tell me which specific aspect of your loadout was the key to your win or loss. If you lost because your 3 teammates ran off and died, that doesn’t matter, specifically your loadout is what we’re looking at.

See the issue?

If we want valid data sets, you’d need the information about the entire loadout of every team member for every match and each team members performance average with the specific loadout they’re using. After all, Timmy may have a 6.3 W/L ratio, but if he’s won every match with Loadout A, and lost every match with Loadout B, then your match with Timmy using Loadout B would need to reflect his Loadout B performance and not his overall performance. We’d also need to know the spawn rates of enemy types, the damage dealt to each enemy type individually per player, the number of hordes and specials and monsters spawned, and even with all of that you couldn’t possibly say “Krak’s need nerfed”

15

I would love to see an array of statistics. As long as we can all agree that we can’t adjust by numbers. Something like win rate doesn’t even matter that much in this game, because the real issue with

Op things vs non-OP things in a PVE game like this is really:

Does this thing remove other players ability to interact with the content as intended

But stats can be useful at identifying outliers too.

16

I think equip rate is the most honest, people don’t all gravitate towards something for love of the larp…

17

It’s not that I’m not interested in overall stats for things, but just more fundamentally I wish we had more knowledge feedback we’re giving to FS is seen and considered and there was more back and forth between the community and devs. I’ve been writing feedback in these forums for years now with little explicit sense anything gets through. Not that I’m a perfect skilled infallible player that they should listen to me about everything, but rather, it would be nice to discuss and hear more from them what their design intentions are more frequently.

Without explicit data I don’t have that much of a foundation to comment on hardcore balance issues outside of big outliers that I feel like dumb down the game (which should be taken care of, this is supposed to be a challenging game, at least at its highest levels), which is exactly why like 90% of the balance feedback I suggest comes not from a place of this is too strong/too weak, but rather, what are the boring or lacking unique attribute weapons/skills/traits/blessings, etc. Because at the end of the day, having a build that feels unique to me that plays differently than someone else’s and testing myself to succeed with it is the greatest pleasure hardcore challenge games like this can provide. And still to this day, I think there’s a gigantic amount of the above that exists in Darktide, way more than other games in the genre. I’m really hoping this latest balance patch is just the first in more frequent experiments

18

thats exactly the point and i guess a great divide between western and eastern mentality.

@Taezzang

i go on a limb here and “assume” (though always the worst approach given its moniker, but still)
you take smoke as a strong option cause "the team wins and wonder why krak is considered op in comparison.

likewise i “assume” as long as the “team wins” in that situation no one cares about the means.

counter example and the essence of the quote above: a krak vet gets all crushers in one or two nades and the rest stand there :eggplant: in hand

THIS right here is the difference in mentality.
when playing with randoms or without a group of people that like/respect each other in real life, sharing time with strangers is about “fairness”.

strange as it sounds, a teams win isnt worth :poop: if you don’t feel like you attributed your considerable share to it or exceeded the others contributions.

even in coop theres competition and usually games mechanics have it each participant has his means to take his fill and feel accomplished.

there’s no satisfaction to be found in watching others playing the game for you, neither does one tolerate it for too long if certain actions gets you “more than you are owed” in comparison to common player inputs such as aim/movement/pacing.

and above all theres no “we/we have won” with randos, theres the “how did i fare when doing the same map for the umpteenth time with these peeps/loadouts”

both, the krak and the smoke playstyle are mutually exclusively playstyles and require each unique adjustments/balancing,

for one group of players goes :“im a good player cause my smoke kept the ranged from all else, even if i didn’t do any kills”

and the other group of players rather approaches the game as in :“look at this pathetic wannabe, no aim, no skill but thinks hes the hot :poop: by throwing kraks all match”

dont get me wrong, the “later” matches do get won, they do so though by individuals doing their utmost to perform at their max, not giving potential away so another stranger does the job for him.

with that in mind, stuff like the krak/plasma/dueling sword/hive scum stuff etc isnt a boon to the “team” its an insult to the individual bringing his time and effort into something that rewards “lesser” options with more participation in the game.

sure, social/cultural differences, still they are the common denominator in this equation here.

19

I think numbers are only useful as a rough reference point.

For example, imagine two Veteran players joining Auric with DS, Krak grenades, and VoC equipped.

One player only has Malice-level skill. Even with a fully optimized meta setup, they still struggle heavily in Auric and end up with an awful winrate. (Everyone has seen the phenomenon of players with Havoc 40 titles constantly going down in regular Auric missions.)

The other player actually has the skill level required for Auric. With the same meta setup, the game becomes trivial unless an accident happens. Their winrate would probably exceed 80%.

That’s why statistics don’t mean much unless you also account for player skill, and player skill varies massively from person to person. Realistically, it’s impossible to fully account for all of that.

Ever since the game’s balance started falling apart, the average skill level of Auric players has dropped dramatically. Most players are honestly below Malice level now, and the constant quitting is another major issue.

So in summary, data should only be treated as a reference. I think the most reliable balance discussions come from players who genuinely understand the game at a high level.

20

I’d love to see these numbers just because they’re cool. Please release them, Fatshark? :smiley: I love statistics.