Hammer of Math: Elo and the World Championships of Warhammer 2025

Jeremy “Curie” here from Stat Check with a post-WCW Elo update. One of the most valuable uses of any ranking system (aside from bragging rights) is to use it as a method to predict the outcome of any one game. Elo (see this article for more details) is derived from an expected value based on the difference between two player ranks, and updates itself after each round using this expected value. Back in January of 2024 we reviewed the accuracy of the system using LVO 2024 as our test case.

Now that another two years of data has been put into the system, we took another stab at this but using the World Championships of Warhammer 2025 (WCW 2025), an annual event that has taken over as the premier competitive event of the year with many of the strongest players in the world in attendance.

WCW 2025

The World Championships of Warhammer recently completed its third installment (fourth if you count the much smaller Grand Finale in 2021) in Atlanta with over 400 players in attendance for the Warhammer 40k event. It was won by Art of War’s Richard Siegler piloting an Adeptus Mechanicus list (go check out our Competitive Innovations article covering it), burning a path through a truly stacked field of competition.

Elo Prediction Accuracy

First things first - a simple look at the whole dataset (4,472 games across 8 rounds of competition) and how accurate Elo is at predicting the winner of the match. As the table below shows, the elo rankings of the players in attendance reflect the expected value quite closely - the higher the difference in player rankings, the more likely the higher ranked player will win.

[table id=130 /]

The win rate also follows the Elo theoretical win rate (the final column in this table) quite closely, showing that the training data used up to this point (GT+ events since February 2022) is sufficient to predict these trends. Overall, the higher-ranked player won 62.6% of the time.

The Championship Bracket

The format of WCW 2025 results in the final four rounds being played in a swiss format with the winner of each group seeded into this top bracket. The top 16 from this year were:

Alexander Jessen - Thousand Sons (Grand Coven) - Elo: 1722.9
Dan Sammons - Adeptus Mechanicus (Haloscreed Battle Clade) - Elo: 2040.2
Folger Pyles - Aeldari (Warhost) - Elo: 2075
Giorgio Castelli - Deathwatch (Black Spear Task Force) - Elo: 1766.9
Jack Harpster - Adepta Sororitas (Champions of Faith) - Elo: 1777.6
Jack Murphy - Adeptus Mechanicus (Skitarii Hunter Cohort) - Elo: 1919.4
Jean-Loup GENIN - Adepta Sororitas (Hallowed Martyrs) - Elo: 1753.3
John Lennon - Emperor's Children (Coterie of the Conceited) - Elo: 2017.3
Kyle McCord - Dark Angels (Wrath of the Rock) - Elo: 1883.7
Leif Westermann - Aeldari (Aspect Host) - Elo: 1843.9
Liam VSL - Aeldari (Aspect Host) - Elo: 2020.7
Mark Hertel - Space Marines (Gladius Task Force) - Elo: 2011.7
Nick Nanavati - Aeldari (Aspect Host) - Elo: 1718.6
Richard Siegler - Adeptus Mechanicus (Skitarii Hunter Cohort) - Elo: 1833.1
Steve Trimble - Drukhari (Spectacle of Spite) - Elo: 1852.6
Travis King - Adeptus Mechanicus (Skitarii Hunter Cohort) - Elo: 1836.5

There were 32 games played in this bracket, in which the higher-ranked player won 18 times - a small but useful indicator that Elo is quite good at predicting outcomes, even at the higher levels of play.

Credit: Robert "TheChirurgeon" Jones

The Champion

Most interestingly is that Richard Siegler’s championship run had two of these games - in his matches against Jack Murphy and Liam VSL he was the underdog. One of the challenges with Elo that has been discussed before is the need for consistent recent data - which we don’t have as much for Richard Siegler as we do for Jack Murphy and Liam VSL. At the start of the event Richard had an Elo of 1833, which ranked him at #113 in the world, but by the end of the event he had jumped to an Elo of 1947 and a global rank of #22. If the Elo post-event was used to predict those same matches, it remains favoured to Liam VSL (and watching back the finals it’s clear that those two players are very closely matched) while Richard gains a slight edge against Jack Murphy.

What’s Next?

Stat Check is working closely with the Goonhammer team to improve the Elo system and provide more tools to players to make use of this data - watch this space for more about this in the coming weeks!

Have any questions or feedback? Drop us a note in the comments below or email us at contact@goonhammer.com. Want articles like this linked in your inbox every Monday morning? Sign up for our newsletter. And don't forget that you can support us on Patreon for backer rewards like early video content, Administratum access, an ad-free experience on our website and more.