I think this is a record for posting the first version of the Excel spreadsheet. For Insiders, it’s available under the Tips+ section. This version is missing three key stats: the RPI and SOS data, as well as the All-American selections from the US Basketball Writers Association. Those come tomorrow. RPI and SOS data cranks overnight, and USBWA doesn’t announce until the afternoon.
-
Peter Tiernan has been using stats to analyze March Madness for 22 years. He writes for CBSSports.com and has also contributed to ESPN.com and SI.com. His insights into the NCAA basketball tournament can help you build a better bracket.Sign-up questions? Click here.
Follow Us
Make the region by region forum… I need some help!!!
Thank you Peter! This is awesome. It looks like the stats you use are only from regular season, not including the conference tournaments. Is that right?
They include conference tourneys.
Thanks Peter – the excel sheet is the best. I’m doing the Final Four/Champ model by myself just because its fun to do and I’m wondering what the margin percentage is? Thanks
AVG MAR/POINT ALLOWED>
Awesome stuff Pete . . . I am cranking out the seed matchups as we speak
What section of the website will all of these models be posted?
Tips+. May do in two waves.
Thanks Pete. Look forward to all the good stuff. Sorry your Wolverines got such a bad matchup.
I’m okay with it. Hey–you gotta win tough games if you’re going to go deep. South Dakota is a test…VCU will be really hard. Games played in Michigan…but MSU fans will root against Blue. And vice versa.
For the seed by seed model, If the 4 seed criteria met the over .500 likelihood benchmark for selection, Michigan would be advancing to the elite 8 . . . following the rules though, I gotta advance Kansas for the model . . . found that interesting though, haha (Michigan fits the profile for victory, Kansas fits profile for upset, but the 4 seed win% is only .483)
Seed by Seed model is complete! Hopefully I can release the full report soon (note, I went strickly by the rules for EVERY PICK, so even though many lower seeds met criteria for upsets, if they didn’t meet the benchmarks made by Pete for the model, they were not rewarded the “W”) . . . quick hits, it has Indiana beating Duke in the Final . . . no true upsets . . . lower seed victories include:
round 1: 9-Missouri, 9-Wichita State, 10-Iowa State
round 2: 6-Butler
Sweet 16: 3-Florida
Elite 8: 2-Duke, 2-Ohio State
The Final Four pitted Duke against Ohio State, which Duke wins for proximity and Indiana vs Kansas, which Indiana won due to a scoring margin over 15.
This is the quoted rules from last year directly from Pete in which I followed:
“I go matchup-by-matchup, only taking the lower seed when the guidance yields better than even odds—and a higher success rate than the high-seed winning rate. In situations where there were countervailing factors—in other words, conditions were ripe for a Cinderella but not for a victim—I advanced the higher seed”
You gotta think that there are going to be way more upsets than that
agree completely
near misses for lower seeded victories which I observed that stood out included:
SOUTH
1st round:
-Akron, South Dakota State, and Northwestern St all met criteria for upsets, though none of their opponents meet criteria for being taken down
2nd round
-VCU meets criteria for 5 seed over 4, but Michigan does not
-Georgetown meets criteria for losing 2 seed, but SDSU does not meet criteria for upset
Sweet 16
-Michigan and Kansas both met the criteria for a lower seeded victory, however, the matchup did not meet the over .500 success rate criteria for selection (Kansas goes on to make the final 4)
Elite 8
-Florida meets criteria for 3 beating 1, but Kansas does not
near misses for lower seeded victories which I observed that stood out included:
EAST
1st round:
-California and Montana meet criteria to pull off upsets, however their opponents do not
-Butler and Marquette meet the criteria for being upset, but Bucknell and Davidson do not make the cut for the upset
2nd round
-Miami fits the profile for losing to a 7 seed, but Illinois does not fit the criteria for pulling off the win
Sweet 16
-Syracuse fits the profile for 4 over 1, but Indiana does not fit profile of losing 1 seed
near misses for lower seeded victories which I observed that stood out included:
MIDWEST
1st round
-gave Missouri meets criteria for 9 seed win but so does Colorado State for 8 seed win . . . Missouri has the better Pomeroy ranking (I gave the game to Missouri)
-New Mexico State and St. Mary’s fit the profile for upsets, however Saint Louis and Memphis do not
-Creighton fits profile for losing to 10, but Cincy doesn’t make the cut
2nd round
-Creighton fits profile of 7 seed over 2, but Duke does not, so Duke prevails
-Michigan St fits profile for 3 losing to a 6, but Memphis doesn’t have the profile for the win
Sweet 16
-Duke fits profile for 2 seed loser to a 3, but MSU does not fit profile of winner
near misses for lower seeded victories which I observed that stood out included:
WEST
1st round
-Both Wichita St and Pittsburgh fit the profile for a 9 beating an 8, so Wichita St advanced
-Belmont fits 11 seed victor profile, but Arizona does not fit the losing profile of a 6 seed
-Wisconsin and Kansas State both fit profiles of vulnerable 5 and 4 seed (respectively), but their opponents do not make the grade
2nd round
-Iowa State meets the criteria to advance as a ten seed, but Ohio State is not a vulnerable 2 vs a 10
-New Mexico, on the other hand, is a vulnerable 3 seed vs a 6, but Arizona does not meet the criteria for the win
Sweet 16
-Kansas State meets the criteria for a 4 beating a 1, but Gonzaga is not vulnerable to a 4 seed upset
one final note, after the first round (in which I looked at every match-up in every region), I only completed this analysis for the match-ups that had advanced in the model, therefor, there might be many more teams that meet criteria for upsets in later rounds that I did not check due to the fact they did not advance to the particular round in the model
Ryan – Check out my email to you. Can I get you to put these picks into the same format as the other models I just posted? A Word doc should be in your inbox.
I will check that now and get back to you