Mathematical analysis: Michigan has a 3.8 percent chance of winning the NCAA tournament

Submitted by lakeside on March 26th, 2013 at 12:41 PM

Not to undermine Butterfield (LINK) who's numbers appear both impeccable and comforting but some hack named Nate Silver published an article for the New York Times (tumblr site?) giving Michigan a somewhat, err, dimished likelihood of winning the NCAA tournament.  Via UMHoops (LINK):

Michigan (up to 3.8 percent from 2.5 percent) Michigan could easily enough have been a No. 1 seed had it played better down the stretch, and it was probably underseeded as a No. 4 even with the losses that it took. In general, however, we’ve found that late-season performance doesn’t tell you that much more than early-season performance when it comes to tournament play – and Michigan’s slump has not extended into the postseason. It was a break for the Wolverines to play in Auburn Hills, Mich., but their domination of a tough Virginia Commonwealth team on Saturday was nevertheless impressive, and they should be thought of as the equivalent of a strong No. 2 seed right now.

It is worth noting that if the 3.8% continues to increase it will someday reach 81.45% and all will be right with the world.




March 26th, 2013 at 12:55 PM ^

For the South and East Regions anyway, here are the estimated probabilities based on some hypothetical matchup on the Massey site. If you take this out to the championship and add up the estimated probabilities of all scenarios in which Michigan wins, it comes close to Silver's number (it also creates a hellish spreadsheet). Strangely, that estimate has held fairly steady throughout the postseason too. 

The favored teams in this algorithm would be Michigan (slightly), Florida, Indiana annd Miami (YTM). The two rightmost columns show the resultant Elite Eight Game. I also did this for the other regional finals as well, but this one is more pertinent to our future. 

Kansas / Michigan Florida/ FGCU Indiana/ Syracuse Marquette / Miami PROB.  SOUTH EAST
0.53 0.89 0.68 0.68 21.811% Michigan / Florida Indiana / Miami
0.53 0.89 0.68 0.32 10.264% Michigan / Florida Indiana / Marquette
0.53 0.89 0.32 0.68 10.264% Michigan / Florida Syracuse / Miami
0.53 0.89 0.32 0.32 4.830% Michigan / Florida Syracuse / Marquette
0.53 0.11 0.68 0.68 2.696% Michigan / FGCU Indiana / Miami
0.53 0.11 0.68 0.32 1.269% Michigan / FGCU Indiana / Marquette
0.53 0.11 0.32 0.68 1.269% Michigan / FGCU Syracuse / Miami
0.53 0.11 0.32 0.32 0.597% Michigan / FGCU Syracuse / Marquette
0.47 0.89 0.68 0.68 19.342% Kansas / Florida Indiana / Miami
0.47 0.89 0.68 0.32 9.102% Kansas / Florida Indiana / Marquette
0.47 0.89 0.32 0.68 9.102% Kansas / Florida Syracuse / Miami
0.47 0.89 0.32 0.32 4.283% Kansas / Florida Syracuse / Marquette
0.47 0.11 0.68 0.68 2.391% Kansas / FGCU Indiana / Miami
0.47 0.11 0.68 0.32 1.125% Kansas / FGCU Indiana / Marquette
0.47 0.11 0.32 0.68 1.125% Kansas / FGCU Syracuse / Miami
0.47 0.11 0.32 0.32 0.529% Kansas / FGCU Syracuse / Marquette


Everyone Murders

March 26th, 2013 at 12:59 PM ^

I have a lot of respect for Silver and his work.  I also think it's great that he's gotten a fair amount of popularity for advanced data-crunching skills.  Geek power and all that.

Plus 3.8% seems somewhat reasonable give that a simple 1/16 chance equals 6.25% and Michigan's road is likely to be tougher than most teams' road to the Championship Game.  So I don't really see that much awry in his assessment.  (I still like Michigan's chances, but I'm a tad biased.)

But it's interesting to note that he's the son of an MSU department head (Poli Sci) and was born and raised in E. Lansing.  Just thought I'd through those tidbits out for the tinfoil hat crowd to chew upon.


March 26th, 2013 at 1:00 PM ^

He's completely ignoring the outcomes in other dimensions as they factor nowhere into his analysis. Somewhere out there we have a 100% chance of winning.


March 26th, 2013 at 1:03 PM ^

Here is the article you're referring to. 

He's got Louisville around a 30%, Florida around 20%, Indiana aournd 10%, and a whole lot of teams in the same boat with Michigan.  In Louisville vs the field, I'm still taking the field at higher than 70%...I dont see them getting there.


March 26th, 2013 at 1:06 PM ^

Kenpom gives us a 3.2% chance of winning it all. Compare that to Florida (29.2%), Louisville (25.3%), Indiana (13%), and Ohio State (8.5%). The probability of an all Big Ten final four is .64% according to Kenpom.

French West Indian

March 26th, 2013 at 1:07 PM ^

...numbers is that they only make sense some times but not all.  Sure if Michigan played in one thousand Sweet Sixteens then they would prevail thirty-eight times.  But the fact is that they are only playing in one Sweet Sixteen and only need to win it one time.

Trying to give people expectations of 3.8 wins when only 1 is even possible is really just unnecessary and probably an abuse of mathematics.  If Michigan is going to win it this year they say that there is a 100 percent chance or if Michigan isn't going to win it this year then say that there is a 0 percent chance.  There's really on two possiblities and it's not that difficult to figure out...if they'd stop confusing people.

eamus_caeruli (not verified)

March 26th, 2013 at 1:50 PM ^



March 26th, 2013 at 1:21 PM ^

as he's rather well respected statistician.  He's actually most notably known for projecting all 50 states correctly in the last presidential election.

Lastly, he alludes to using kenpom, so it's not like his numbers are coming from nowhere.  Full breakdown of his pre-tournament analysis here.

1 percent

March 26th, 2013 at 2:37 PM ^

After hours of number crunching I've concluded that if Michigan were to make the final 4 they would likely have around a 25% chance of winning it all.


March 26th, 2013 at 3:13 PM ^

I think there is an easy way in which you could reach a 4% chance for Michigan winning it all - Chances of victory against Kansas (around 45%), chances of a victory in a possible matchup against Florida (around 55%), chances of a victory in a possible matchup against Indiana (around 40%), chances of a victory in a possible matchup against Louisville (around 40%). The odds could obviously improve if we get more favorable matchups.

Mr. Rager

March 26th, 2013 at 4:12 PM ^

KenPom Odds for Michigan:

- vs. KU: 47%

- vs. UF: 40%

- vs. IU: 42%

- vs. UofL: 40%

- vs. Duke: 46%

- vs. Sparty: 49%

- vs. Ohio: 46%

- vs. Miami: 53%

Clearly UF would be a matchup issue for us.  Since it's basically our worst possible matchup of the remaining teams (outside of UofL).  

[Edit - these are all taken from the analysis someone here posted before the tournament started.  the odds now, I would guess, are more in Michigan's favor.]