Big Ten Tournament Seeding Scenarios (updated after Wisc/MSU)
I was counting on the screen and came up with 32 distinct scenarios:
SEED | Indiana | Ohio St. | Wisconsin | Michigan | Michigan St. |
1 | 84.38% | 3.13% | 12.50% | 0.00% | 0.00% |
2 | 9.38% | 43.75% | 18.75% | 18.75% | 9.38% |
3 | 6.25% | 28.13% | 15.63% | 25.00% | 25.00% |
4 | 0.00% | 25.00% | 15.63% | 31.25% | 28.13% |
5 | 0.00% | 3.13% | 37.50% | 25.00% | 37.50% |
I wasn't quite sure how to approach the tie that takes place in one scenario, so the 4th and 5th seed numbers for Ohio State and Wisconsin are italicized to denote admitted approximation on my part (hence, they total more than 100%).
LATE EDIT: The table above is indeed the coin-flip scenario - thanks those who pointed it out. I meant to add more but hit "Save" prematurely and lost some stuff by compounding the problem and exiting Excel and lost the Massey-adjusted table.
Here are the Massey estimates -
Taking into account Massey's predictive probabilities, which give us an estimated 55% chance at beating Indiana, for starters, we can find the most probable scenario based on the OP's table.
Further, Michigan State has superb odds against Wisconsin and Northwestern (74% and 95% respectively), and Ohio State has an estimate 83% chance of beating Illinois. Wisconsin also has 83% against PSU per that algorithm. Taking that into consideration, the scenario in which Indiana beats us, MSU wins out, and Wisconsin and Ohio State both win is overall the most likely based on his estimates.
So, it would be, in that case, Indiana at #1, Ohio St. at #2, MSU at #3, Wisconsin at #4 and Michigan as the #5 seed under those assumptions, I believe.
It got cut off but the Wisc/OSU 4/5 tiebreaker goes to Wisconsin if Minnesota lost either one of their remaining two games, which happened last night.
Those % are kinda interesting but basically meaningless since they assume a 50/50 chance of each outcome. So MSU pops up at 37.5% chance of a #5 but in 2/3rd of those scenarios they will have lost at home to NW which is like 2% chance of happening. If you weight them for probable outcomes it does not look pretty for Michigan if we do happen to lose Sunday.
Of course, that assumes that all of the games are tossups. Since most of the teams that would help Michigan to a higher seed are underdogs, the chance of us ending up on the 5 line is probably significantly greater than 25%.
You can't do that, your data is assuming every game is a coin flip. A perfect example of GIGO.
Here is the table I meant to post which uses the Massey numbers. I basically switched out the winner with the estimated probability of winning, so the essential arrangement of the data did not change. I didn't include the seeding scenario side just because I didn't want to do the entire thing again, but here would be the probabilities based on the OP's table:
WISC. / MSU | MICH / IND. | MSU / NW | WISC. / PSU | OSU / ILL. | LIKELIHOOD OF SCENARIO IN OP'S TABLE |
0.26 | 0.55 | 0.05 | 0.17 | 0.17 | 0.021% |
0.26 | 0.55 | 0.05 | 0.17 | 0.83 | 0.101% |
0.26 | 0.55 | 0.05 | 0.83 | 0.17 | 0.101% |
0.26 | 0.55 | 0.05 | 0.83 | 0.83 | 0.493% |
0.26 | 0.55 | 0.95 | 0.17 | 0.17 | 0.393% |
0.26 | 0.55 | 0.95 | 0.17 | 0.83 | 1.917% |
0.26 | 0.55 | 0.95 | 0.83 | 0.17 | 1.917% |
0.26 | 0.55 | 0.95 | 0.83 | 0.83 | 9.359% |
0.26 | 0.45 | 0.05 | 0.17 | 0.17 | 0.017% |
0.26 | 0.45 | 0.05 | 0.17 | 0.83 | 0.083% |
0.26 | 0.45 | 0.05 | 0.83 | 0.17 | 0.083% |
0.26 | 0.45 | 0.05 | 0.83 | 0.83 | 0.403% |
0.26 | 0.45 | 0.95 | 0.17 | 0.17 | 0.321% |
0.26 | 0.45 | 0.95 | 0.17 | 0.83 | 1.568% |
0.26 | 0.45 | 0.95 | 0.83 | 0.17 | 1.568% |
0.26 | 0.45 | 0.95 | 0.83 | 0.83 | 7.657% |
0.74 | 0.55 | 0.05 | 0.17 | 0.17 | 0.059% |
0.74 | 0.55 | 0.05 | 0.17 | 0.83 | 0.287% |
0.74 | 0.55 | 0.05 | 0.83 | 0.17 | 0.287% |
0.74 | 0.55 | 0.05 | 0.83 | 0.83 | 1.402% |
0.74 | 0.55 | 0.95 | 0.17 | 0.17 | 1.117% |
0.74 | 0.55 | 0.95 | 0.17 | 0.83 | 5.456% |
0.74 | 0.55 | 0.95 | 0.83 | 0.17 | 5.456% |
0.74 | 0.55 | 0.95 | 0.83 | 0.83 | 26.636% |
0.74 | 0.45 | 0.05 | 0.17 | 0.17 | 0.048% |
0.74 | 0.45 | 0.05 | 0.17 | 0.83 | 0.235% |
0.74 | 0.45 | 0.05 | 0.83 | 0.17 | 0.235% |
0.74 | 0.45 | 0.05 | 0.83 | 0.83 | 1.147% |
0.74 | 0.45 | 0.95 | 0.17 | 0.17 | 0.914% |
0.74 | 0.45 | 0.95 | 0.17 | 0.83 | 4.464% |
0.74 | 0.45 | 0.95 | 0.83 | 0.17 | 4.464% |
0.74 | 0.45 | 0.95 | 0.83 | 0.83 | 21.793% |
So, there it is - this is the scenario where Indiana is #1, OSU is #2, MSU is #3, Wisconsin is #4 and we are #5. The next most likely, under these assumptions, would be Indiana as #1, OSU as #2, us as the #3 seed, MSU as #4 and Wisconsin at #5.
Just eyeballing those numbers it looks like about an 8% chance of both losing to Indiana and getting the first round bye. Most of that is from the scenario where WI loses to MSU and PSU. Otherwise, it is pretty much beat Indiana or else we have to play on Thursday against PSU.
Didn't check how Kenpom's % vary from the Massey #s you used but I know we are 44% against Indiana on kenpom and Massey has it as 45% so it sounds like perhaps they are pretty similar.
What could prove to be interesting is if Michigan St and Michigan both get placed at The Palace. I know there will be an afternoon session and evening session, but carry over fans from both teams will gladly cheer for whichever directional school we are playing.
So, I guess that what we want to happen is at least a 3 seed, which means if we win against IU, and MSU beats wisconsin, and everyone else wins like they are supposed to, then We grap a 3 seed.
If MSU drops out of the four-way tie due to a subsequent loss, we're still at least a 3 seed.
I don't really care that much about the B1G tournament seed or outcome (other than how it affects our NCAA seed). Not much difference in a 4 or 5 seed other than having to play an extra game.
Looks like we are going to be a #3 if we beat IU and a #5 if we don't.
Using Pomeroy's pythagorean number I come up with the following percentages for Michigan
1 Seed - 0% chance
2 Seed - 4% chance
3 seed - 17.5% chance
4 seed - 16.5% chance
5 seed - 62%
Appreciate the numbers but that looks a bit off. Kenpom has us with a 44% chance of beating Indiana and if we beat Indiana there is no scenario where we are the #5 seed. So we have at least a 44% chance of being higher than a #5 seed and no more than a 56% chance of being a #5 seed but you are listing us at 62%. Doesn't follow logically.
Perhaps you just took the pythag numbers and came up with a lower win % against Indiana based on that w/o accounting for home court? Not sure what else might be the issue.
1. Beat PSU on Friday by 20 points.
2. Beat Wisky on Saturday, with no possibility of late game heroics.
3. Beat Indiana for the second straight Sunday.
4. We get the #1 seed in the Midwest.
Sparty gets shifted to a different sub regional.
I'll live only numbers 3 and 4 happen!
Michigan (or any B1G team) can not be the #2 in the Midwest if IU is the #1. There is a rule that the first three teams from a conference must be placed in three different regions. Now that IU has slipped up a little and there are some projections with Kansas as the #1 in Midwest, you may start seeing more having a B1G as the #2.
First post updated after Wisconsin/MSU. Pretty straight forward now: