Win rates can be misleading; one of the reasons we frequently look at the Tournaments in Winning Position (TiWP) stat, and the reason Stat Check look at OverRep is because it’s very likely that a faction’s popularity will affect its overall win percentages – that is, very popular factions are likely to be dragged down toward average results thanks to having much larger player bases with a wider spread of skill levels (this is the marines conundrum), while less popular factions will rise or fall on the back of a very small number of players.
Something that came up this week in the Goonhammer Discord was the idea of how the meta changes for good players vs. bad players vs. everyone in the middle. I can’t take full credit for the original idea – shout out to the Nurgler and tldr on our Discord for kicking off the conversation around looking at ELO results by win rate this week.
About the Sample
For this sample, we’re using a mix of data from Stat-Check’s meta dashboard and player ELO data, and Tabletop battles scoring data collected at 40kstats. For our breakout of “good players” vs. “Everyone else,” we’ve used a cutoff value of 1750, which roughly maps to the top 200 or so players and a sample of 2,770 games since the Pariah Nexus updates.
Comparing Win Rates by Faction
The first thing we need to note before we dive into this is that, as expected, win rates are just much higher for the top ELO players. Even writing this as a well-below 1750 player, my win rate at events in Tenth Edition is 79.2%, and we’re talking about players who routinely win 80% of their games at five-round events. As a result, it’s difficult to compare win rates directly – instead there are two things we want to look at with these stats:
- the relative increase or decrease over the average win rate for the group, i.e. 79% for top ELO Players compared to 46% for sub-1750 players, and
- The representation of the faction – when it comes to top players, a faction being selected more or less often creates an implied endorsement of the strength of that faction.
So without further ado, let’s look at the stats:
First off, let’s offer a final caveat: Those player and game counts for top-Elo players are small. They were always going to be, based on how we had to do our cutoff. That’s going to make statistical significance really hard to come by – we’re going to have to trust that with this particular group, the relative stability of tournament outcomes will make smaller changes in win rates more significant. Additionally note that, because events tend to pair winning players against each other, players in the 1750 group tend to play harder games and against players with higher Elo rankings.
Next, let’s talk general observations:
- I’ve sorted this list by the win rate for top-Elo players, and looking at this, we can see that Black Templars, Sisters of Battle, Space Marines, Space Wolves, and AdMech top the charts (Deathwatch are also in there too but their days are tragically numbered so let’s not dwell on those).
- For sub-1750 players, Thousand Sons, Drukhari, Space Wolves, Adepta Sororitas, and Blood Angels top the chart, with Thousand Sons showing a sizeable lead.
- In terms of Representation, the Thousand Sons see the biggest bump of any faction to the top Elo crew – almost 5% higher. Adepta Sororitas aren’t far behind them, however. Knights, Custodes, and World Eaters see the biggest drop offs.
- The biggest differences between our two samples here are for Space Marines, who tend to do a lot better in the hands of top players, and Knights, who tend to do a lot worse. Well, relatively speaking – average players aren’t better with knights, but they do get better relative outcomes.
- For top players right now the game’s worst factions are Knights of all varieties, Votann, and Genestealer Cults – not even Innes can prop them up.
This is a really cool set of results to look at, and suggests that there are some armies which are overpowered at the game’s lower levels but more effectively countered at upper levels of play, or being held back by other top armies.
If we use BCP IDs to combine Elo scores and game scoring data, we can look at how VP scored vary by player type and faction. Generally speaking, the High-Elo players average just over 10 VP more per game than Sub-1750 players, scoring an average of 59 VP per game.
Here we can see some pretty large gaps, where top players are just putting up substantially higher points per game than the sub-1750 group when it comes to Black Templars and Space Wolves, and putting up worse scores with Thousand Sons, World Eaters, and Orks. That suggests these armies have a higher skill floor, and make it easier for mediocre players to put up better scores and records with these armies, but not necessarily win games against top armies. Likewise, we can suggest the same for Imperial Knights and Chaos Knights, which were removed from the above chart due to insufficient data numbers, but what we had showed substantially lower scores.
There’s some more diving you can do here – we could, for example, look at the middle band of players, with Elo scores in the 35th to 65th percentile, and compare results against other players in that band. When you do that you get some fun results, like Genestealer Cults suddenly becoming a menace in a way that shows how they can be very strong if you’ve never played against them before. And similarly, Genestealer Cults among top Elo players do well against other top Elo players. But let’s get to the really good stuff, i.e.
What Does the Bad Player Meta Look Like?
That’s an interesting – and in some ways more fun – question. Paring back our win rates to just look at players in the bottom 25th percentile of results shows some fun changes. In this domain we’re looking at win rates which mostly over between 25 and 30 percent overall, and players are both more susceptible to tricks and struggling in a first game against a particular faction and less savvy with their own armies.
Here Imperial Knights, World Eaters, and Chaos Daemons jump to the top of the heap, and that scans given how tricky they can be to play against – they’re armies which players can often struggle with in a first encounter. Here Orks, Marines, and Genestealer Cults are among the game’s worst armies, with the GSC being far too difficult to make work for players at this level.
The army which most consistently performed well at every level? Space Wolves. While the Thousand Sons and Adepta Sororitas may be the game’s best armies right now, you can argue that their limitations and difficulty of play curb their results and make them difficult to pilot well in less experienced hands. In contrast it’s a lot easier for anybody to pick up an army of T6, 4-wound, 4++ save thunderwolves, point them at the enemy, and go to town on them.
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.