We’ve incorporated all the rest of 2013 today in one big batch. That’s 27 tournaments added at once, including Pro Tours Gatecrash and Dragon’s Maze. Having several weeks off from new tournaments let me work on 2013 as a group which expedited the curating process. I might try to do something similar in the weeks around the release of Amonkhet. There’s about 1.014 million matches in the database now, and in a couple of weeks we’ll hit tournament number 200. I believe at the moment there have been 625 relevant tournaments total, so roughly a third now are available for your perusal. (Relevant being GPs, PTs, and Worlds. I’m thinking about tackling national championships too though.) The older tournaments are (a) smaller on average and (b) may not have usable records kept, so I’m guessing it’s actually more like 40% to 45% of the possible database that’s up now.

Our data set is starting to reach a size where you can see peaks from different “eras” emerging. With each passing match an old result has a smaller and smaller effect on your current rating, so adding old data isn’t useful if all you care about is what the current ratings are. But ratings as of several years ago are closer to the front of the dataset, so new additions to the front will still influence them. Huey hit a peak upon reaching 11-0 at Pro Tour Magic 2015 which has grown to become the fourth-highest mark ever attained (2305 at the time of writing). Other high points, like Sam Black’s high at GP Toronto 2013, will be interesting to track once tournaments from 2012 start getting added.

Despite what I said in the previous paragraph, if you’re keeping careful attention you may have noticed that there’s been a “rising tide” effect as I’ve added more results to the beginning of the data set. For example, LSV’s rating was 2227 right when the World Championships were done, but since I’ve added a year to the front end of the data his rating has risen to 2258. This is almost entirely due to indirect contributions, rather than direct ones. Adding these tournaments has increased the rating of LSV’s opponents, and so matches closer to the present look more impressive than they did previously. All boats rise due to this effect — even my rating went up about three points from the beginning of working on this cluster of tournaments. We’ll continue to see this as we continue going back in time, since most people don’t have a transcript long enough to make early tournaments irrelevant. There are ways to correct for this inflation, but I think it’s a problem to tackle at a later date.

In the last update, I said that I’d update the table of draw streaks each time I completed a year. The table in the entry below is now current. We found the beginning of some long streaks, like Christian Calcano’s and Joe Demestrio’s. Owen said on Twitter that he can’t remember taking a draw, so if there is one it’ll be a while before we come across it. How far back does LSV’s streak go? In a little bit we’ll peel back another layer.