Announcing the 7th Edition Rolling Meta Project

People seemed to like Requiem for an Edition, which was gratifying to see, but one of the most common requests was for this to also take place for 7th Edition.

So I’m going to do it. Simple enough, right?

The approach for this one is going to be a little different however. Rather than picking on “iconic” tournament, like I did with the LVO, I’m going to be continually updating the picture of what the tournament scene is like – which means there’s definitely going to be some armies that wax and wane, etc. How I’m going to do this involves some tricky statistics (for those of you who are interested: Bayesian Logistic Regression), but basically, I’m going to be calculating the odds that a particular army ends up in a “Top 5” placement. But the way the model is going to be set up is going to take into account prior information, so the estimates get more precise as time goes on, and lots of little tournaments can count as much as one big one. This offers a few advantages:

  • People take this prior information into account when designing their armies. They don’t go “Gee, I wonder who will be good at this tournament…” they go “Eldar are kicking ass and taking names, I better come prepared…”
  • This allows small local tournaments to be reflected in the results, but gives greater weight to the “Big Guys” like the LVO, BAO, NOVA Open, etc.
  • It makes codex releases easy to handle. New Space Wolves codex drops? Just reset the “prior information” for Space Wolves, and you clear the set for them – while keeping the information for everyone else.

What comes out of this is what’s called an “Odds Ratio” (OR), which is the ratio of a particular army placing vs. all armies placing. So 1.00 means you have an exactly even chance, and if every army has an OR of 1.00, the game is perfectly balanced. Similarly, an OR of 2.00 means you’re twice as likely to win, and an OR of 0.50 means you’re half as likely to win as everyone else.

I’ll be updating things about once a month, more frequently when the big tournaments hit. Right now, the only results I’m reporting are the ones on Torrent of Fire, but if you’re a tournament organizer and would like your information to be included, please get in touch.

Large tournaments will also probably get the full breakdown like I did for the LVO, so stay tuned! Easiest way to get updates for new posts is to either subscribe to the RSS feed, or follow me @VarianceHammer on Twitter.

A note on smaller tournaments: A Top 5 placement in a 5 man tournament doesn’t much mean anything, now does it? For tournaments smaller than 20 people, Top 5 will be substituted for only the winning army. The reason I’m not calculating just the winning army for everything is a sneaking suspicion that once you’re in that very top tier of players, dice rolls and particular vagaries determine who actually wins, and it will be very hard to control for that. Besides, my suspicion is very few players, even competitive players, go to win and would be dissatisfied with doing very, very well.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.