Five brackets with best being 95%.
Last year, I didn’t win, but learned a lot more. Still chasing the dream of almost winning over a decade ago. Still stings to.
First, the verdict is in. Seeding matters!
But before I dive into that, I have a few other worldly thoughts.
With Covid slowly resolving itself in the United States, it was my hope that we could focus on only one madness this year – the Madness that is March. After all, it is in this Madness that one can find an escape from so much of what is wrong with the world, even if it is but for a moment.
Alas, March Madness will have a new contender for its time and our obsession. For as I write this, another madness has taken over the world – the madness of war. However, I will not forgo the annual Madness that is March that in most years can give me a bit of joy and sanity. As to the war, I pray for peace. I extend my support as limited as it is. And I pray that the madness that is war will soon end.
So, this year we will have an NCAA Tournament in the middle of a war. Again, I pray for the Ukrainians and for the leaders of the world to make good decisions. I also hope and pray the 68 teams and all those involved can remain safe as they seek the elusive goal of a Championship Title.
Now back to the Madness that is March.
Personally, I continue to be on my personal quest to find the “perfect” bracket. It’s been around two decades since I almost won a national bracket challenge. The only problem was I did not trust the data and took my favorite team for two more wins than I knew I should. So as part of that quest, I thought I would revisit some of my favorite March Madness subjects. One of those subjects is tournament seeding. In fact, it is my opinion that seeding continues to be one of the most important aspects of the tournament for any team. To prove that point you just have to look at any team that has been “awarded” the ninth (9th) Seed. Then look at the history of their performance.
But more importantly, this fact is proven in this year’s update of Rusty’s data analysis. When you are trying to determine how many games a team will probably win, seeding is a player in three of the top four predictors. So yes, seeding matters.
Besides seeding, the other factors that are good predictors are Jeff Sagarin’s work, the old school RPI, and four calculated data factors we have developed. For now, back to seeding.
While studying seeding some years ago, I read an article online entitled “Bracket Science.” It contained a lot of specific data about seed wins. It also went through a lot of the criteria by which certain seed vs. seed games could be better analyzed. If you can find it on the web, it is a very interesting and informative read.
Not including the play-in and first four in games, there have been 2,268 games played over the last 36 years. After you have removed all of the games where the participating teams had the same seeding, there are 2,241 games to evaluate on seeding alone in the modern era. The following pages will offer up a numbers summary of the seeds and the results of those seed pairings.
In this 2022 edition of The Seed Report, we will leave in Rusty Haner’s special section “An Analysis of NCAA Basketball Statistics from the 2013 to 2017 Tournaments.” It is a good insight into some relevant statistics with an excellent explanation of how a data scientist got to his predictions. As stated before, his new analysis will be published in another book soon after the brackets are announced.
We hope you find this material helpful in your own pursuit of the perfect bracket.
Views: 4