I wasn’t sure I could pull this off, but it seems like I can. Every year, I put out the probability of winning the pool, but I assume that every game is a coin flip. I know that isn’t true, but calculating the probability of winning based on weights and other statistics took too much computing power to calculate, but now it seems like I am able to do it.
Here’s what I did. I
have a spreadsheet that has all the possible scenarios plotted out. I needed a rating for each team left. I threw all that into ChatGPT and
told it to calculate the probability of each scenario happening. I then matched that up with who would win in
each of those scenarios and added up the total to figure out who has the best
weighted chance of winning.
I needed to have a ranking for each team relative to
each other, so that the odds for any game could be calculated. I used the Baysian system I used during the
regular season. It overrates some teams
and probably underrates others, but what I used was:
Duke 106.9
Illinois 104.97
Arizona 104.65
Michigan 103.96
Iowa St. 102.39
St. John's 100.98
Arkansas 100.69
Houston 99.23
Michigan St. 98.29
Purdue 98.01
Tennessee 97.16
Alabama 96.89
Nebraska 96.83
Connecticut 95.48
Texas 94.6
Iowa 92.82
Obviously, this ranking has a big effect on the outcome, but
this is what I had. I think Connecticut
is low and Illinois is probably high. I
used 11 as the standard variance for each game.
Based on this, here is the weighted probability of winning:
|
1.:Howard S. #1 (9.025%) |
|
2.:Pete J. #4 (5.9514%) |
|
3.:Josiah Tru (4.974%) |
|
4.:Irene 8 (3.9664%) |
|
5.:Brent E. #7 (3.9084%) |
|
6.:Andy Mil. (3.6043%) |
|
7.:Pete J. #3 (3.3943%) |
|
8.:Andrew D. #2 (3.2326%) |
|
9.:Dani A. #1 (3.0994%) |
|
10.:Wayne Der. #1 (3.0165%) |
|
11.:Brian Kerb. (2.3602%) |
|
12.:Andrew Simm. (1.9062%) |
|
13.:Eric V. (1.8826%) |
|
14.:Jared Bock (1.789%) |
|
15.:Bryce Sel. (1.7808%) |
|
16.:Addie H (1.7486%) |
|
17.:Pete J. #1 (1.5234%) |
|
18.:Brent E. #2 (1.3045%) |
|
19.:Irene 3 (1.2791%) |
|
20.:Billy Egn. (1.2402%) |
|
21.:Brandon H (1.2319%) |
|
22.:Rock Chalk Not Jayhawk (1.1988%) |
|
23.:Rich Fron. (1.0996%) |
|
24.:Finn D. (1.0852%) |
|
25.:Steve Mucc #6 (1.0603%) |
|
26.:Don’t Coog It (1.0317%) |
|
27.:Cicely Kohn. (1.0312%) |
|
28.:Jacob Able (1.0047%) |
|
29.:Scott Kohn. #1 (0.863%) |
|
30.:Steve Mucc #5 (0.8517%) |
|
31.:Bill Pr. #1 (0.7902%) |
|
32.:CAAT #2 (0.7641%) |
|
33.:Irene 5 (0.7455%) |
|
34.:Irene 7 (0.7449%) |
|
35.:Ryan Wiech. (0.7363%) |
|
36.:DAZE 4 (0.6314%) |
|
37.:Jackson I. (0.6054%) |
|
38.:Szecho L. #1 (0.5927%) |
|
39.:Mike Lu #4 (0.5745%) |
|
40.:Got Cluffed (0.558%) |
|
41.:Brent E. #1 (0.5414%) |
|
42.:Zach V. (0.5408%) |
|
43.:Matt Meis. (0.5294%) |
|
44.:Elizabeth N. #2 (0.5203%) |
|
45.:Brent E. #9 (0.5093%) |
|
46.:Jay Bilas (0.4917%) |
|
46.:Leuder 2 (0.4917%) |
|
46.:Mike Lu #3 (0.4917%) |
|
46.:Pete J. #5 (0.4917%) |
|
50.:Jim Del. (0.4859%) |
|
51.:Chris Min. (0.4682%) |
|
52.:Irene 2 (0.4655%) |
|
53.:Brent E. #8 (0.4522%) |
|
54.:Steve Mucc #4 (0.4405%) |
|
55.:Muccmars-2 (0.424%) |
|
56.:Mike Niem. (0.4215%) |
|
57.:Stefon K. (0.419%) |
|
58.:Mark Sp. #2 (0.4108%) |
|
59.:Becky Her. #1 (0.3909%) |
|
60.:Paul W. (0.3781%) |
|
61.:Luis B. #6 (0.3654%) |
|
62.:Luis B. #3 (0.3614%) |
|
63.:Harrell Rob. #3 (0.3561%) |
|
64.:Tim Talb. (0.3477%) |
|
65.:Brent Sw. #2 (0.3371%) |
|
66.:DAZE 6 (0.3344%) |
|
67.:Andrew D. #3 (0.3301%) |
|
68.:LinVinSanity #2 (0.3286%) |
|
69.:Leuder 1 (0.321%) |
|
70.:Luis B. #2 (0.3195%) |
|
71.:Fly Eagles Fly (0.3174%) |
|
72.:Brent E. #4 (0.3145%) |
|
73.:LinVinSanity #1 (0.3127%) |
|
74.:DAZE 8 (0.3061%) |
|
75.:Missy Wyt. 2 (0.302%) |
|
76.:Mike Lu #5 (0.3006%) |
|
77.:Max70 #1 (0.2998%) |
|
78.:Wayne Der. #4 (0.2883%) |
|
79.:Dani A. #2 (0.2731%) |
|
80.:Craig Gerul. (0.2636%) |
|
81.:Marlee Talb. (0.2611%) |
|
82.:Beau H (0.244%) |
|
83.:Michelle L. #3 (0.238%) |
|
84.:Howard S. #2 (0.2278%) |
|
85.:Michelle L. #1 (0.2192%) |
|
86.:Mason A. #1 (0.2045%) |
|
87.:Mitchell Slu. (0.1961%) |
|
88.:Mike Lu #1 (0.1956%) |
|
89.:CAAT #3 (0.1927%) |
|
90.:Dan Dziub. (0.1926%) |
|
91.:Sarah Chao (0.1843%) |
|
92.:Mason A. #2 (0.1701%) |
|
93.:Jordan Hu. (0.1699%) |
|
94.:Jason St. #2 (0.1695%) |
|
95.:DAZE 2 (0.1649%) |
|
96.:Chris Elm. (0.1611%) |
|
97.:DAZE 7 (0.1517%) |
|
98.:Scott Kohn. #2 (0.1464%) |
|
99.:Chrissy S. (0.1368%) |
|
100.:Tina H (0.1322%) |
|
101.:Ray Min. (0.1187%) |
|
102.:Leuder 3 (0.1165%) |
|
102.:Leuder 4 (0.1165%) |
|
104.:Luis B. #1 (0.1072%) |
|
105.:Patrick McC. #2 (0.1054%) |
|
106.:Muccmars 1 (0.1012%) |
|
107.:Brandon Br. #2 (0.1001%) |
|
108.:Harrell Rob. #1 (0.091%) |
|
109.:Shaun Turn. (0.0881%) |
|
110.:Tyler Chr. (0.0861%) |
|
111.:Matt Niem. (0.0855%) |
|
112.:Kirt R. #1 (0.0825%) |
|
113.:Michelle L. #5 (0.0781%) |
|
114.:Matt H. (0.0728%) |
|
115.:Dianna Alv. (0.0671%) |
|
116.:Gutless Bracket (0.0659%) |
|
117.:Patrick McC. #4 (0.0658%) |
|
118.:Ally Her. #2 (0.0601%) |
|
119.:Mary Lou H (0.0577%) |
|
120.:Tim Dziub. #2 (0.0569%) |
|
121.:Noah N. (0.055%) |
|
122.:DAZE 3 (0.0521%) |
|
123.:Church On Time (0.0505%) |
|
124.:Klei R. #2 (0.0464%) |
|
125.:Klei R. #1 (0.0446%) |
|
126.:Elizabeth N. #1 (0.0445%) |
|
127.:Bret Torvik (0.044%) |
|
128.:Marc R. #2 (0.034%) |
|
129.:Chris T. (0.0305%) |
|
130.:Jacob Able Chalk (0.0241%) |
|
131.:Steve Mucc #3 (0.0167%) |
|
132.:Rob I. #1 (0.0162%) |
|
133.:Brandon Br. #3 (0.0155%) |
|
134.:Becky Her. #2 (0.0139%) |
|
135.:Mike Wiech. (0.0138%) |
|
136.:Charlotte Niem. (0.0122%) |
|
137.:Scarlett H (0.0093%) |
|
138.:DAZE 1 (0.0085%) |
|
139.:Jaxon S. (0.0078%) |
|
140.:Jason St. #1 (0.0076%) |
|
141.:Leighton Niem. (0.0074%) |
|
142.:COOL GUY (0.0063%) |
|
143.:Brad Rich. (0.0057%) |
|
144.:Dawn Tru (0.0043%) |
|
145.:Mike Lu #2 (0.0038%) |
|
146.:Dylan Bow. (0.0027%) |
|
147.:Tim Dziub. #1 (0.001%) |
|
148.:Steve Mucc #7 (0%) |
Howard S. #1 jumps to the top. This is because he has Illinois beating Arizona in the Final,
which accomplishes 2 things. One, not
many people have Illinois winning it all (only 4 people), and I have both
Arizona and Illinois pretty high in my rankings.
On the flip side, Brian Kerb is #1 if you think every game
is a 50/50 proposition, but he drops down to 11th in weighted
probability. That is because he has Iowa
State beating Connecticut in the finals. Both of which are ranked pretty low in
my standings, so even though Iowa State is a pretty rare pick, my system thinks
it is also very unlikely.
Hope this adds a little extra context to all the games. Illinois could lose tomorrow, and all of this
is out the window. As a matter of fact,
as soon as 1 game finishes tomorrow, all this work no longer matters. I’ll have to do it again after the next round
is over.
Enjoy and good luck!!!!!!!!
No comments:
Post a Comment