Clustering Twilight Imperium Factions into Tiers

Readers of this blog know that I’m fascinated with the intersection of data and my favorite board game, so I was delighted to come across this Reddit thread, in which a user shares an interesting graphic he created with R:

The poster explains the concept and his process as follows:

There are a lot of “subjective” tier lists out there, so I thought to make one that was more based on stats

Data is from this table of victory points and win rates since May 2022 which is the date of the last codex entry.

Keleres was averaged together. The XXcha variation is D tier but only has 31 games entered while the others are solidly F tier.

One notices immediately that Yassaril is in a league of its own above all the other factions in the current meta, outdoing even famously high performers such as Jol-Nar, Mahact, and Titans. One might immediately conclude that Yassaril are unbalanced relative to the other factions, but this is not necessarily the case.

First of all, this just reflects the overall meta of the game since the release of Codex 3 in May 2022. These stats can and likely will change whenever the next Codex or expansion is released.

Secondly, this might just be the result of the community finally getting a firm grip on the game after several years of playing, particularly in regards to action cards, which I personally find to be the most difficult aspect of the game. A player with sufficient mastery over and experience with action cards can make Yassaril’s special abilities truly shine.

Finally, another poster makes a good point in the comments in response to a concern with how useful the Y-axis (Average Victory Points) really is if there’s only a spread of 1.1 victory points between the lowest- and highest-rated factions:

What you’re seeing is a VP floor - which is just the sign of a well designed game. It suggests that all factions are essentially “in it” for most games, and that high tier factions are only capable of generating 1-2 extra VP per game over low tier factions, as opposed to fully blowing out other factions.

Deviation from the trend line likely suggest “Feast or Famine” gameplay - a la the Winnu, a faction that’s is clearly designed to score points in a specific way. Their moderate win rate suggests that their kit is sufficient to accomplish that goal fairly regularly, but their sharp departure from the trend indicates that when blocked, they fall off significantly.

In other words, though some factions do better than others, TI4 is far from a pre-determined roulette of the same six or seven factions. More game designers (tabletop or otherwise) should take a closer look at Twilight Imperium and take notes on proper balance between asymmetric factions and/or builds.