#,*-$
- # = Rank among seed group using KenPom rankings. The 1-seed with the highest KenPom ranking gets 1, next highest gets 2, next highest gets 3, and lowest among the 1-seeds gets 4. For some of the 11- & 12-seeds involved in play-in games, the ranks go out to 5 and 6 among the seed group.
- * = Seed comparison between tourney seeding and KenPom efficiency seeding.
- O = Over-seeded: A team seeded as a 2, but efficiency rankings suggest 3-16 or worse.
- A = Accurately seeded: A team seeded as a 1 and efficiency rankings suggest the same.
- U = Under-seeded: A team seeded as a 7, but efficiency rankings suggest 1-6.
- $ = A team's suggested KenPom efficiency seeding. (A team ranked KenPom 1-4 would be a 1-seed, 5-8 would be a 2-seed, etc, etc, 41-46 would be an 11-seed assuming two 11-seed play-in games), 47-50 would be a 12-seed, etc, etc.
- The , and the - are separators, they do not represent anything.
- The colors of each region are also insignificant. They are just a way to tell the difference between each region for easier visual comparison, and to tell from which region the NC and NR came.
Here are the models since 2014 for comparative purposes.
As always, if there are any errors, please let me know in the comments. I will fix them. I'm good at typos, not as good at catching them in the process.
2014 Green Region: 3 seed says U-6 should say O-6 correct?
ReplyDeleteYou are 100% correct. I even double-checked to make sure the other two parts of the notation were correct, so it was a legitimate typo error on my part. Thank you very much for catching it and I have fixed it.
DeleteHi there, thanks so much for putting this together. I’m a long time fan of Pete Teirnan and am so glad I found your site. I’m totally geeking over the OS/US chart. But unless my eyes are crossing from looking at brackets, I think I found a typo error.
ReplyDeleteIt looks like the 6-11 and 3-14 pairs are numerically switched? It looks like it’s in all 4 regions. Those two pairings just look “reversed” to me, as if I was just reading the seed numbers top to bottom. Then the next question...is the data also swapped, or not? Let me know if this looks accurate. THANKS!
Jimmy
For your first question, if by "numerically switched" you mean "they look different than the blank templates" provided by ESPN/CBS/NCAASPORTS, then yes they are. I've done brackets since 1998 and I've always remembered them as 1-16-8-9-5-12-4-13-3-14-6-11-7-10-2-15 so that the 2-3 octet would look like a mirror reflection of the 1-4 octet. Sometime in the mid-2010s, 3-14-6-11 was flipped to 6-11-3-14 and that's the way it ran since then. I don't prefer the new way simply because it isn't logically consistent. If the 2-3 octet were isolated and re-ranked: Then, 6-11-3-14-7-10-2-15 would look like 3-6-2-7-4-5-1-8. My way of writing it keeps 2-7 on the outside and 3-6 on the inside, the same way the 1-4 octet follows. Having 6-11 then 3-14 just seems impure. After ten years of the new way, I've still never trained my brain to see the new way, so out of spite, I do it my way, not the cabal's way.
DeleteFor your second question: The data matches the correct seed, the 3-14-6-11 pod is just written my way. For example, Green 3-seed is an O-4 (4-seed quality over-seeded as a 3-seed). If this was a 6-seed, then a 6-seed that is a 4-seed quality would be a U-4 (under-seeded 4-seed).
Hope this answers your question. If there is a particular bracket that has the wrong winners running out of it, that's entirely possible. For example, I had two 1-seeds running out of the 2016 orange bracket because I missed the 0 on the 10-seed.
OK, thanks for that info, it makes sense now. Been a crazy tournament for sure, hope you were able to enjoy the games!
Delete