Updated Big 12 Odds - KU on the verge

KU has a 99.6% probability of winning a share of the Big 12 title and a 93.3% probability of winning the Big 12 outright. I'll take those odds. Our probability of winning against Texas Tech this weekend is 88%. It is still likely (probabilistically speaking) that we will lose one of the three remaining games. I sure hope we head to the NCAA tourney 30-4. A 36-4 final record has a nice ring to it. Heading into next season with a 20 game winning streak also has a nice ring to it. Both West Virginia and Oklahoma have to play Baylor so there is a good chance Baylor will knock off one of them. Oklahoma also has Texas this weekend in Austin so that will be a tough test. West Virginia has Texas Tech and Oklahoma St. so Baylor is really the only likely loss. Let's hope Baylor runs the table and KU ends up winning the Big 12 by 3 full games. A guy can dream can't he?

A request that you can feel free to turn down:

I'd love to see a graph of the week to week probability of KU's chances for the title over the course of the conference season. Better yet a graph plotting all of the teams' weekly chance of winning. I'd imagine that they'd all start out at 10% and KU would slowly climb and then take a HUGE dip during our cold 5 game stretch, with WVU and OU getting ahead of us, only to see a reversal of that trend as one-by-one teams are eliminated and their chance becomes zero and KU rockets upward. Of course, I'm a nerd.

Exactly: was it a gradual decline with a steady recovery? A nosedive? A hard upward bounce?

But, yeah, no need to ignore the job or family for this. When I have time tonight, I might just go rifle through the old posts and do it myself.

Not a problem. I actually have all sorts of graphs that I use for various insights and trends. I do a trailing 5 game and trailing 10 game graph to get the trends then take the curve fit and the slope determines the trend. I do a graph of difference from projected margin to actual margin verses opponent rating to get mental toughness. In other words, some teams play much better (normalized against expectations) against weaker opponents and some teams play better against stronger opponents.. That one is incredibly interesting. Of course I have the graph of probabilities by round. To be totally fair I give a lot of credit to Asteroid for providing the great work he does that has been very edifying to me. I am his biggest fan. Probably the only thing I do that he doesn't (or hasn't posted) is the probability of winning. That is a complex calculation. There are 90 games in the Big 12 schedule. Each one has the possibility of two outcomes. So the total possible outcomes is 2 to the 90th power. That is a very big number. So you have to run all of those possibilities using the probabilistic outcome for each game (which I now get from my DPPI). So the probabilities I publish here are extremely robust. I would be happy to publish those graphs here. I never imagined anyone else would be interested in graphs. Another thing I am planning on doing next season is a more complete analysis of all 351 teams in terms of annual trends. I am in the process of compiling all the stats for all teams back to 2003-2004 season (for now). Last year I downloaded a great many stats for the NCAA tourny into a database so I will probably be publishing some fun facts on that this year. If you are a real nerd there are some incredibly interesting things in the seeding outcomes from the last 30 years of 64 team tourney's. Please feel free to request anything you would like in terms of Jayhawk stats.

You are truly the man, Corpus. Can't wait to see whatever you come up with.

Remember your audience here....we're total geeks for this stuff. Visual representations of data (i.e. graphs and charts) are awesome ways to bring the data to life. You'd be amazed at the crap I'll chart/graph if I have the data handy. Heck, my own "season tracker" spreadsheet (just a full conference season schedule that I update to show w/l, standings, home/away record, and "magic numbers" to 13 and 14 wins, along with the actual "magic number" would be seen as totally bizarre by most people, but is a drop in the bucket compared to what I'm sure you and Asteroid have going.

Thanks!

Attached is the graph requested by NotO. Thanks HawkErrant for your help in learning how to attach it here. After 7 games, KU's chances looked grim. Out probability was around 17% of a shared title and 6% for an outright title. We improved slightly the next two games but that was when West Virginia was surging. But after game 9 it has been all KU all the time. OU peaked after game 7 and has been headed down ever since. WVU peaked at game 9 and has crashed like a lead balloon. What an amazing Big 12 season for KU.

This is very cool. Thanks for posting it. Aside from the big 3, we now have close to half of the league eliminated, so just as you see OU and WVU getting dangerously close to hitting 0%, you can go back 2 or 3 weeks and you'd see TCU, KSU and other impact craters.

It's also interesting to see this if you look at it alongside the schedule. KU started strong...wins vs. Baylor, OU and @Tech. Loss @WVU and we're brought back down (imagine if we had won that one). TCU didn't help, and the losses @OSU @ISU got us to the bottom around game 7-8. Wins vs. KSU and @TCU started the climb out, but "the big week" where we beat WVU and won in Norman not only dinged WVU's and OU's chances (and their beating each other up), but rocketed us to the top.

Kudos and many thanks. I'm totally geeking out on this.