Launch App
Back to Blog

New Rating System: What Changed and Why

We found bugs in how ratings were calculated. Here's what we fixed, how it affects your rating, and the data to prove it.

Your rating is changing. It’s changing for the better.

We found real problems with how player ratings were calculated. Not rounding errors. Actual bugs that made ratings less accurate and less fair. We rewrote the system, ran the numbers on all 1,128 rated players, and this post walks through what was broken, what we fixed, and the data that proves it works.

What Was Wrong

Five things were broken.

1. The butterfly effect. Adding a new tournament recalculated every player’s rating from scratch. Your rating could shift even if you weren’t in that tournament.

2. Time travel. Future tournament results could influence calculations for older tournaments. A result from last Saturday could change how a tournament from three months ago was scored.

3. Decay punishment. The old system applied a 4-month decay window. Take a break and your past wins faded to nearly nothing. Your skill didn’t change, but your rating said it did.

4. Ties counted as losses. Two players draw, both lose rating.

5. No incremental updates. Everything recalculated from scratch every time. Adding one tournament could shift everyone’s rating unpredictably.

What We Fixed

The new system processes tournaments in chronological order, exactly once. Each tournament uses ratings as they exist at that moment. No decay, no multi-pass recalculation, no time travel.

Old SystemNew System
ProcessingRecalculates everything 5 timesProcesses each tournament once, in order
DecayRecent events count more, old ones fadeNo decay. Your results are permanent
TiesCounted as a loss for both playersProperly scored as a draw (0.5)
Adding a tournamentShifts every player’s ratingOnly affects players from that date forward
ConsistencyChanges depending on when you checkDeterministic. Same data, same result

If you played well six months ago, those results now count at full value. If you tied someone, you get credit for the draw. Your rating won’t shift because someone ran a tournament in another state.

The Numbers

We ran both algorithms on all 1,128 rated players and compared the results.

The old system (orange) spread ratings from 1246 to 1760, a 514-point spread. The new system (blue) compresses ratings into a proper bell curve from 1364 to 1645, a 281-point spread. That tighter distribution means the differences between players are real, not noise from a broken algorithm.

About half the players moved up, half moved down. The average rating change was +1.3 points. The total amount of rating in the system barely changed. What changed is how it’s distributed. Players who were unfairly boosted came down. Players who were unfairly penalized came up.

Who Moved and Why

Here’s how the top 10 shook out:

New RankPlayerOld RankChange
#1Rodrigo Ezequiel Lezcano#10
#2Yoel D.Buigues#4+2
#3Bucs1211#29+26
#4Vpassion#20+16
#5Kouhaiken#9+4
#6juan cruz “faye”#45+39
#7BeelzeBoy#6-1
#8Gonzalo “Gono” Delaguardia#49+41
#9Klammeh#7-2
#9Vee#22+13

A few patterns stand out.

Players who dropped typically had few events and benefited from the old decay weighting. Showing up recently was worth more than a strong historical record. That artificial boost is gone. If your rating dropped, the old system was giving you too much credit for recency.

Players who climbed had consistent results over a longer period. The old system penalized them for having older tournaments in their history. Those results were decaying toward zero. Now they count at full value.

The number one player stayed number one. If you’re the best, the algorithm reflects that regardless of methodology.

Does Playing More Help?

No.

The old system rewarded frequency because of decay. The more recently and often you played, the more your rating benefited. Not because you were better, but because your results hadn’t faded yet.

The new system doesn’t care how often you play. It cares how well you play when you do. There’s no meaningful correlation between events attended and rank change. Whether you’ve played 5 tournaments or 50, the system evaluates your results the same way.

What Your Rating Means Now

RatingWhat It Means
1600+Top tier. Consistently beating strong competition
1550-1600Strong player. Regularly finishing near the top
1500-1550Above average. Winning more than you lose
1450-1500Average. Competitive but still developing
Below 1450Getting started. Every tournament helps you improve

The range is tighter now (281 points vs 514). Every point matters more. A 50-point gap between two players is more meaningful than it used to be.

This is the foundation we’re building on. Head-to-head records, rating history graphs, confidence intervals. They’ll all be built on a system we trust. Questions or feedback? Come find us on Discord.