Welcome to Elo 2.0! See the most recent blog post to learn more.
2016 September 13 (Adam)

Refreshed the database again. Highlights from this update:

  1. Five new tournaments were added: Grands Prix Chicago, Moscow, Manchester, and Atlanta 2014, and Pro Tour Journey into Nyx. (Pop quiz: who won PTJOU?) Some things I was forced to ponder: why was Moscow only 14 rounds? Why are the two halves of round five of GP Chicago the same (including the standings, to ensure maximum difficulty in reconstructing the results)? Why have a GP the week after the Pro Tour in the same city?
  2. I discovered that round 12 and round 13 of GP Porto Alegre 2015 were copies of each other. I reconstructed round 13 and we got this fixed. If you happen to notice a player playing the same opponent in two rounds in a row, it might be another instance of this mistake. I’m pretty sure that there aren’t any more instances of it in the database at the moment, but this problem may come up again in the future.
  3. I merged and split a few people who were/weren’t the same thanks to tips we got from the community. Thanks guys! Keep it coming. I also fixed some Mike/Michael, Andy/Andrew, Dave/David, and Tony/Anthony mistakes. I didn’t really know where to look to see if people from Moscow were the same as other people — I didn’t realize when we started the project that we’d need to be knowledgeable in Russian transliteration conventions. I’m sure there are some entries that need to be combined/separated that so far have gone undetected. Before I add more tournaments I’m going to work on cleaning up what’s here a bit.
  4. We’re now rating unintentional draws. For information on what this entails, check out the entry about draws in the FAQ. I was originally hesitant to do this because I was afraid that I wouldn’t be able to tell the difference between intentional draws and unintentional ones. But sometimes it’s not hard to tell if it’s intentional:

    5. Most intentional draws are reported as “Draw D-D”, “Draw 0-0-3”, or “Draw 0-0-0”, though sometimes 0-0-1 or just 0-0. My general rule of thumb was to interpret any of these notations as representing an intentional draw if I could find any remotely logical reason why the players were incentivized to draw. This includes making top 8, nabbing an extra Pro Point at a PT, or even the rare round nine matchup of 6-0-2s who could ID to make day two. Any sort of 0-0-x draw that ended with one or both players in the money I treated as an ID. Now there were some random apparent IDs in early rounds scattered throughout the tournaments. Sometimes it’s easy to confirm that these were unintentional:

    But I’m afraid some of these were people convincing their opponent to skip a round and get lunch. There are 17119 draws in the database. Probably about 1000 of them should be intentional, and I’ve marked about 800 of them. My guess is that about 200 intentional draws are inaccurate. Draws don’t have a big effect on the rating, either, so this isn’t something to lose sleep over. But doing it the other way, with no draws rated, about 16000 matches were being tallied incorrectly. If the goal is to minimize wrong results, this does represent progress.

    Note that for rounds that I had to reconstruct, like Porto Alegre round 13 from pont #2 above, all results are styled as 0-0, whether it’s a win, loss, or draw. So if you want to point out a match to me that’s a potential ID, you need more evidence than what our site is displaying.

← previous post
next post →