information for noobs
Unreal Tournament still lives and you can download and play it by going here https://www.ut4ever.org/
Discord:
About every hour. Unless I am pugging.
The data comes from https://ut4stats.com/, I use data from the last 9 months.
I'm sorry I ever created this as it has caused much confusion, but it is measure of FlagCaps
, FlagCarrierKills
and KDR
,
although it is completely subjective, if you look at players with the best BlitzPower
you will see that they are the best flag carriers and flag killers.
The recommended balancer DamageShare (BlitzPower)
tries to balance AverageDamageShare
across both teams.
There is a little secret sauce which means that stronger players are slightly more heavily weighted than weaker players.
Meaning that a player with AverageDamageShare=30
> Two players with AverageDamageShare=15
Different weightings were tried, until I got a best fit, see validator
For example
ShX raw = [30.474298447339564] weighted = [31.241915316236472]
JœKehr raw = [16.634956124678347] weighted = [16.862392188646176]
DamageShare
is PlayerDamage / TotalTeamDamage
(ie the share of Damage
contributed by the player). So if everyone was the same skill level, the expected DamageShare
of each player would be 20%.
Very good players have a DamageShare
of 30%.
I tested different variables including Damage
, Kills
, AverageKillsShare
, Score
,FragPower
and BlitzPower
and
found AverageDamageShare
to be the best predictor. It's not perfect, but it is better than relying on what people think.
AverageDamageShare
is the best predictor of team strength, is a real metric (i.e., I don't do any funny weighting to it).
This is the primary thing I use to balance on.
You can check the validator, to see that the checker is right 67% of the time, and is at over 80% correct for what the checker predicts are stacked games
In other words, when the checker predicts 75%:25% the team with 25% is almost certainly going to lose - unless the weaker team has LÔßÔ or Bezy-Boomsa - then anything is possible.
The MVP (Most Valuable Player) score is designed to reward players who make a real impact across multiple areas of the game — not just by racking up kills, but also by grabbing flags and taking out the flag carrier when it matters.
We look at three key actions:
To keep things fair, especially for players who only played a few matches, we use a Bayesian weighted average. That just means we blend your stats with a league average, so someone who had one amazing game doesn’t jump to the top unfairly.
Extra weight is given to:
So your final MVP score reflects not just raw performance, but balanced, team-focused impact over time. 🔥
We use a Bayesian Weighted Average to calculate each player’s impact score. This helps level the playing field between players who’ve played 2 games vs 200.
Here's what we look at per game:
Then we use this formula:
Weighted MVP Score = (RawScore × GamesPlayed + C × m) / (GamesPlayed + m)
This ensures:
Depending on your score and ranking, you might earn:
Hover over your badge to see what you earned it for. 🎖️