Emperor Arc Meta-Game Update #1

Since there were no tournaments this weekend (that I’m aware of), I thought about skipping the weekly stats update and just go straight to the Quarriors! review, but I decided instead to go ahead and post a data-set of the arc so far (if you want to get a jump on that Quarriors thing, the review is already up on BoardGameGeek, so you can go read it and watch all the Quarriors fans jump all over it).

This data, I’ll note, is mostly for curiosity. I don’t think it really tells us anything significant to smash tournaments from completely different environments together. You’ll get some vague things, like who was more consistently good across the arc (not that the arc is that long at this point), but even that is kind of iffy because some environments have many more data points than other environments, and so the data is skewed towards them. It accurately reflects what happened, but what does that all really mean? Maybe next week I’ll do this data as an average of the different environments.

Also note that, as I mentioned in the comments to last week’s update, the chi-squared standardized residual is not really a helpful tool for comparing across environments. It functions well within a particular environment, but what made it different from prior figures I had been using was that it could actually tell us when the data became statistically significant. The problem is that this means that the CSSR numbers get bigger even when the relative performance between the Clans is otherwise identical. So a CSSR of 1.23 might mean a much stronger performance in one environment than the other, it’s just that one environment didn’t include as many data points.

So, I’m just going back to the “relative chance of making the cut.” It’s not entirely intuitive for the un-mathy, but it can be compared across arcs. The main thing to remember is that the values get pretty high pretty easily.

For the environment as a whole, the figure in the “relative chance of making the cut” column is the average of the absolute values of the clan figures. So if you had one Clan at +15% and another Clan at -15%, the environmental figure would just be 15%.

All Environments
Players % of Field Made Cut Won % Made Cut Relative Chance of Making Cut
Crab 549 13.68% 147 11 26.78% 30.09%
Crane 340 8.47% 50 2 14.71% -28.55%
Dragon 437 10.89% 93 12 21.28% 3.39%
Lion 488 12.16% 124 10 25.41% 23.45%
Mantis 420 10.47% 93 8 22.14% 7.58%
Phoenix 422 10.52% 87 5 20.62% 0.16%
Scorpion 522 13.01% 100 8 19.16% -6.93%
Spider 361 9.00% 54 2 14.96% -27.33%
Unicorn 396 9.87% 80 8 20.20% -1.85%
Unalign 77 1.92% 8 0 10.39%
4012 836 20.58% 14.37%

Make the Cut Ranks
1) Crab – 27%
2) Lion – 25%
3) Mantis – 22%
4) Dragon – 21%
5) Phoenix – 21%
AVERAGE – 21%
6) Unicorn – 20%
7) Scorpion – 19%
8 ) Spider 15%
9) Crane – 15%
9a) Unaligned – 10%

Oh, hey, wait, I don’t have a full chart generated for taking the average per environment per clan (instead of just lumping all the data points in a big pool), but apparently I had already set it up to calculate the make the cut rate for that scenario. For that way of looking at the data (environments are equally weighted, regardless of number of tournaments), we get:

Make the Cut Ranks
1) Crab – 26%
2) Lion – 25%
3) Mantis – 22%
4) Unicorn – 22%
5) Dragon – 21%
AVERAGE – 21%
6) Phoenix – 19%
7) Scorpion – 18%
8 ) Spider – 17%
9) Crane – 15%
9a) Unaligned – 10%

3 thoughts on “Emperor Arc Meta-Game Update #1

  1. As a giant nerd, I think it’d be interesting to see the data for T4 appearances (or to a lesser extent T2 appearances). Obvs people place too much emphasis in the actual winner of Koteis; I think T4 might be a stronger indication of relative strength when you really get to the upper echelons of competitiveness. Thanks though for these posts. They’re always interesting.

  2. I have all T2 data from the 2012 Kotei season. Plus all Top X from the same season, but I can’t dissect it to show Top 4 from a Top 8 or 16.

Leave a Reply