Dr. Green and White Sports Authority

It is officially the most wonderful (sports) time of the year!  As we all think about our Final Four and upset picks for our office pools, I thought that I would provide some unsolicited, stat-based advice on how to fill out your bracket.  Full disclosure right off the bat:  I am actually not that good at winning my office pool.  I think that last time I actually finished first was 2002.  But, I certainly have a lot of data that sheds some insight as to how likely certain events are.  I personally think that data, combined with the good old fashioned eye-test, is the best combination.  I will do my best to provide the first part. General Methodology Over the past few decades I have slowly accumulated data and developed more and more advanced tools to understand the role of variance and probability as it applies to college basketball.  I could go on about this for days, so I will try to highlight the key points here that underpin my entire philosophy: 1) Overall, the best predicto

Now that the Big Ten regular season is almost over, it is time to start thinking about the possible scenarios in the Big Ten Tournament next week. I have still not found a great way to do a thorough calculation of the various tie-breakers, but I managed to sort-of brute force the calculations and last night I finally got all the formulas figured out such that I am confident in the odds. It also helps that with only 7 total regular season Big Ten games left, the number of possible scenarios is more manageable (only 128, down from over 30,000 just a few days ago). So, here is what I have: Since Tuesday night, we have know the situation at the top. Once again, the winner of the MSU-UofM game will get the 1-seed. The loser will get the 3-seed unless Purdue blows their last game at Northwestern. All odds are derived for the most up-to-date Kenpom efficiency margins. MSU  will  be favored to beat Michigan, so MSU is the current favorite to get the 1-seed. As we move down the chart, it is cle