Tweet
RPIforecast gives the following percentages for an undefeated season :
Arizona 6.23%
Syracuse 0.14%
Wichita State 11.73%
Everyone else 0.00%
Arizona 6.23%
Syracuse 0.14%
Wichita State 11.73%
Everyone else 0.00%
-
-
jdmee, that "everyone else 0.000%" addition made me laugh. Not sure why, but I found it funny. Thumbs up.Comment
-
Arizona's percentage is surprisingly high.Originally posted by jdmee View PostThe mountains are calling, and I must go.Comment
-
Originally posted by Jamar Howard 4 PresidentStill and always will be a terrible methodology, but here's the update with 12 games left:Originally posted by Kung Wu
Chances of Shoxers going undefeated with 95% chance of winning each game -- now at 54%! At 95%, we now have a greater chance of going undefeated than not going undefeated, but barely.
Chances of Shoxers going undefeated with 80% chance of winning each game -- now nearly 6.9%
But here's the best part of all this ... kenpom puts us around a 0.96% chance of winning each game. So ...
The chance of going undefeated through 30 games using kenpom's game by game projections: 61.3%Kung Wu say, man who read woman like book, prefer braille!Comment
-
Is the RPIForecast going through 30 games, or the entire season and post season?Kung Wu say, man who read woman like book, prefer braille!Comment
-
kung, your numbers are wrong. kenpom.com does not have us at 96% each game. Where did you get that?
Also, why are you so against this methodology?Comment
-
rpiforecast currently only shows through the regular season. later in the year, they will add the option to look at projections through conference tourneys, but they aren't there yet.Originally posted by Kung Wu View PostComment
-
Just the regular season, no conference tournament or NCAA Tournament.Originally posted by Kung Wu View Post
ETA: Jamar beat me to it.Comment
-
For those wondering, Ken Pomeroy does an awesome job with numbers.
http://kenpom.com/team.php?team=Wichita+St.
He currently has WSU's chances of an undefeated season at 15.3%, and unlike our "pick a single % chance for every game this year and stick with it" approach, he is able to give each game an individual % chance of winning which makes his final number of 15.3% pretty useful. It is based on his rank of WSU and every other opponent on their schedule, which of course is up for debate, but the percentage that he has assigned (98% chance of winning a home game vs Loyola, 64% chance of winning a road game at Indiana State, etc) seem to be reasonable in my mind.Comment
-
Oregon is tanking and Colorado lost it's top scorer for the season, leaving Cal and UCLA as their only legitimate threats. Even so, that percentage does seem a point or two high.Originally posted by wsushox1 View PostComment
-
Yes he does, I'll explain later ... playing with the kids at the moment. But the hint is he gave us a 50% chance of running the tables a few games back.Originally posted by Jamar Howard 4 President View PostKung Wu say, man who read woman like book, prefer braille!Comment
-
Kung, I still have no idea what you are talking about. Very interested to hear your response.Comment
-
Been BUSY, but I'm sneaking onto Shockernet! I think you are correct, I think I misread one of his points or I misremembered the number I calculated. Instead of .96 he had us at .917 chance to win each game over an 8 game stretch. We were 12-0 and he had us at 50/50 to go 20-0. That means we had x^8 chance to go 20-0. Solving for x gives about .917 (double check my math, I'm in a hurry). So we could plug in that number for the last remaining 12 games, and we'd get a 35% chance of winning the next 12 games. Now obviously, the .917 is based on his match-ups over 8 games, and the match-ups we have left are very different so the .917 could be completely different. But in your original "analysis" you were just giving ranges based on nothing ... at least this gives us _something_ to go on. But if you pay Kenpom I guess you can get the actual chance based on exact projections (which is what I think you posted above, but I am not a subscriber). Sorry this is sloppy, but I got to run! Again, check my math if you want to, it could be all jacked up. Later!Originally posted by Jamar Howard 4 President View PostKung Wu say, man who read woman like book, prefer braille!Comment
-
OK, I think I understand what you were doing Kung Wu.
Anyway, I'm a $20/year subscriber. Here is what Ken has on a game by game basis:
Indiana State - 87%
@Illinois State - 84%
@Drake - 81%
Loyola - 97%
Evansville - 97%
@Indiana State - 64%
@Northern Iowa - 69%
Southern Illinois - 97%
@Evansville - 91%
@Loyola - 91%
Drake - 94%
Bradley - 92%
Missouri State - 95%
Put it all together and WSU has a 16.5% chance of an undefeated regular season. Doing some reverse calculations, I come up with 87% as the number we should be using if we are going to look at my crude "one number for every game" method. An 87% chance x 13 games = 16.4% chance of winning all 13.
Kung Wu, I'm still curious why you say my "one number method" is so terrible. If I list 95%, 90%, 85%, and 80% and let the reader pick the one they think is most accurate, how is this a bad starting point? Obviously now that the season is midway through, sites like Ken Pomeroy's have good rankings to help us fine tune the %'s, but especially at the beginning of the year when we didn't really know how good the teams on WSU's schedule would be, I think my "crude" method was very helpful as a starting point.
Just for fun... crude method updated as of today:
95% each game - 51% chance of undefeated regular season
90% each game - 25% chance of undefeated regular season
85% each game - 12% chance of undefeated regular season
80% each game - 5% chance of undefeated regular seasonComment
Comment