@MiGoon_Fox-3 yet another one lmao
1 Like
bmpt72 barely postive in KD even with the 10.3 release date stats included
i didint expect it to be this bad lmao
1 Like
Gaijin determines BR changes based on aggregate performance statistics across all players. This methodology has a critical flaw: it doesn’t separate player skill tiers.
The Leopard 1A5 Example:
The Leopard 1A5 at 9.3 BR is being proposed for a buff to 9.0. I say this as someone who genuinely loves these vehicles. I own handmade models of them. This isn’t competitive frustration. It’s concern that a vehicle I care about is being misrepresented by flawed data.
Here’s what those statistics don’t show:
- Experienced players perform excellently with Leopard 1A5 because they understand its playstyle: long range sniping, mobility-based flanking, avoiding direct brawling
- Inexperienced players consistently misuse it as a brawler, dying repeatedly to vehicles it was never designed to fight head-on
- The aggregate stat combines both groups, producing misleading “poor performance” data
The Real Issue
War Thunder’s player population includes:
- Veterans who understand vehicle-specific playstyles
- Players who skipped the learning curve through premium vehicles
- New players who haven’t developed playstyle awareness yet
- Players who watched a 2-minute YouTube video on how Leopard tanks actually work vs. those who haven’t
When a Leopard 1A5 player uses it as a brawler against T-72s instead of a long-range sniper, the resulting death isn’t a vehicle performance issue. It’s a playstyle issue.
Gaijin’s statistics can’t distinguish between these two scenarios.
As someone with a background in sociology and data analysis, the misuse of aggregate statistics without controlling for skill variables is a well-documented methodological flaw. It produces misleading conclusions regardless of the field it’s applied to. War Thunder’s BR decisions are no exception
The Cascade Effect
This creates a predictable and documented cycle:
- Aggregate stats show poor performance (due to misuse)
- Gaijin interprets this as vehicle underperforming
- BR buffed to lower bracket
- Skilled players immediately exploit the now-undertiered vehicle
- Gaijin panics and reverses or overcompensates
- Russian/other nation vehicles get compensatory BR changes
- The whole BR ecosystem shifts based on one flawed decision
We’ve seen this pattern repeatedly: vehicles get buffed, get exploited, get nerfed, trigger compensatory changes elsewhere. The root cause is always the same: aggregate data without skill filtering.
What Gaijin Already Has
Gaijin stores detailed player statistics in profiles: K/D ratios, win rates, vehicle-specific performance, match scores. The data to separate skilled from unskilled players already exists. It’s allegedly just not being applied to BR decisions.
Why This Matters
Bad BR decisions based on flawed methodology affects the whole game development:
- Skilled players get their vehicles undertiered then overtiered repeatedly
- New players face vehicles that were buffed for wrong reasons
- Other nations get compensatory changes based on flawed premises
- The BR ecosystem becomes increasingly unstable
I may submit this as a formal Suggestion. Community discussion and additional examples welcome.
(Please read the post fully before answering)
Should Gaijin use skill-tiered performance data rather than aggregate stats for BR decisions?
- Yes, separate all player tiers for BR evaluation
- Yes, but only for vehicles with complex playstyle requirements
- Yes, but use different tier criteria (explain below)
- No, aggregate stats are sufficient (please explain)
- I don’t know / I don’t care
0
voters
higher than 0.75 kills per spawn is above average.
This killed COD
1 Like
Vehicle: JH-7A (JH-7A Prototype)
Gamemode: Ground Realistic
BR Change: 12.0 > 11.7 and remove LS-6(IR) bombs.
Reason: China doesn’t have solid attacker for their popular 11.7 set-up and for higher BR there is much better options.
2 Likes
Why are you using all time stats? is it really that difficult to post the proper stats?
1 Like
damn we as a commiunity are horrible if the bar is that low
2 Likes
cod has battle rating style balancing? never knew that, i thought you could always use any gun in any match.
“This isn’t competitive frustration. It’s concern that a vehicle I care about is being misrepresented by flawed data.”
it couldn’t be more obvious unless the prompt was included too
2 Likes
actually above median not average, but that means better than “the average player”.
its thanks to mostly high amount of suicides & J outs even in ground battles and 1-4 players hogging most of the kills in the match.
Like, I myself use AI for translating longer texts as english isnt my first language, but at least these are my own thoughts, just slapping chatgpt output out here without any editing is terrible.
3 Likes
Vehicle: OTOMATIC
Gamemode: Ground Simulator
BR Change: Add it to 10_2 preset
Reason: Italy’s only option is sub-par Sidam-Mistral manpad carrier which is useless against vehicles that matter. On the other hand OTOMATIC has probably most inflated BR in the game, it is barely any more effective than 2S38 so there is no reason that Italy shouln’t get a better SPAAG.
Its almost like its well documented that vehicles in major nations have worse stats than vehicles in minor nations, even when they are identical:
The numbers get scewed dven more when the vehicle is OP:
1 Like
7 Likes
I laughed at the XM800T comment. 7.3, lmao.
2 Likes
Mistake, didnt check. Picture he posted is correct for lastest stats.
Zzz, the game is balanced by stats. So yeah, the maps don’t suit the Leopard 1A5/A1A1. They get eaten. Not that this is even going to be a buff for these vehicles, it will just put them in more matches with autocannon and APHE slingers that melt these Leos on sight. I wouldn’t worry about any compression issues.
1 Like
Vehicle: SANTAL
Gamemode: Ground Realistic
BR Change: 9.3 > 9.7-10.3 and change Mistral-1 to Mistral-2
Reason: France lacks any decent SAM in 10.0-11.0 range. AMD-30 DCA is alright at ~9.0-9.3 BR range, where SANTAL isn’t needed.
5 Likes