Pomeroy Ratings are Reset

[Henry Sugar]

http://kenpom.com/blog/index.php/weblog/game_on/

Interestingly, he decides not to start every team at zero.  (as explained in the BP book.  I got it as a Christmas present two years ago - meh).  We start the season ranked #32.

Here is the Marquette page.  He predicts 20-10 (11-7)

http://kenpom.com/team.php?team=Marquette

Also, the site looks different.

[bilsu]

Now that I have seen the scores for all the games, I do not have to waste my time going to them.

[rocky_warrior]

I don't know if I can handle losing to Vandy.  Luckily he only has 38% confidence in that.

[Lennys Tap]

Looks like a replay of last year with lots of games going down to the last possession.

[TallTitan34]

He has us seeing only one loss at the Bradley Center.  That loss is to Syracuse. 

Home: 17-1
Neutral:  0-1
Away: 4-7

[MUDPT]

Always interesting.  Swing games between 40 and 60% are:

Wisconsin (56% win)
West Virginia (60% win)
at ND (52% win)
Syracuse (46% loss)
at UConn (42% loss)
at Seton Hall (44% loss)

Overall we are 20-10, 11-7 in the Big East.  Also doesn't include the second (championship) CBE game.

[HoopsMalone]

20-10;  11-7 in conference.   Might as well fill that in for the next ten years because we are like the sun rising and the sun setting with that record.

[Henry Sugar]

20-10;  11-7 in conference; first or second round NCAA loss.   Might as well fill that in for the next ten years because we are like the sun rising and the sun setting with that record.

sigh

In truth, however, I expect MU to make "the leap" next year and contend for the BE championship.  <crosses fingers>

[Benny B]

20-10;  11-7 in conference.   Might as well fill that in for the next ten years because we are like the sun rising and the sun setting with that record.

Ned: Phil?
Phil: Ned?
[Punches Ned in the face]

[HoopsMalone]

In truth, however, I expect MU to make "the leap" next year and contend for the BE championship.  <crosses fingers>

Me too.  Honestly, I feel better about these next two years with DJO than I did during the Amigos years.  Hopefully we can get a Sweet 16 out of it.

[Mr. Nielsen]

I guess he doesn't pick day 2 in KC, since he only has MU playing 30 games.

[dsfire]

The opponent for that game is unknown, so it's not the in projection yet.  Same as postseason play.

[Brewtown Andy]

Ned: Phil?
Phil: Ned?
[Punches Ned in the face]

They say we're young and we don't know
We won't find out until we grow
Well I don't know if all that's true
'Cause you got me, and baby I got you

[Tugg Speedman]

Unless my eyes are crossing, Pomeroy's "Ws" and "Ls" shows us losing 9 games but lists our record 20-10.  I then clicked on Duke and he has a "W" on every game this season (suggesting an undefeated season) yet says their record will be 26-4.

Can we trust a guy that has a problem counting to 10 correctly?

[Benny B]

Unless my eyes are crossing, Pomeroy's "Ws" and "Ls" shows us losing 9 games but lists our record 20-10.  I then clicked on Duke and he has a "W" on every game this season (suggesting an undefeated season) yet says their record will be 26-4.

Can we trust a guy that has a problem counting to 10 correctly?

The discrepancy is explained right there on the page:

"Projected record is based on cumulative probabilities of winning each game and may not equal the individual game predictions."

[Tugg Speedman]

so my eyes did cross and I missed that ...

So the sum of the parts (nine losses) is greater than the whole (projected 10 losses).  Same with Duke (no losses versus 4).  Are all teams like this?  Or does the fact that our projected record is worse than the sum of all the the games suggests that Pomeroy thinks the "intangibles" will work against us (and Duke ... but it is hard for the intangibles to work for Duke when you they start with no losses).

Or, I'm I reading too much into this?

[Brewtown Andy]

so my eyes did cross and I missed that ...

So the sum of the parts (nine losses) is greater than the whole (projected 10 losses).  Same with Duke (no losses versus 4).  Are all teams like this?  Or does the fact that our projected record is worse than the sum of all the the games suggests that Pomeroy thinks the "intangibles" will work against us (and Duke ... but it is hard for the intangibles to work for Duke when you they start with no losses).

Or, I'm I reading too much into this?

It's probably just a by product of doing one prediction and then doing 30 much more focused predictions.  The schedule also doesn't have the 2nd CBE game, which will be against either the #4 or #12 team in the country, so there's the other projected loss anyway.

[Henry Sugar]

bump

http://kenpom.com/blog/index.php/weblog/prediction_tracker/

Pomeroy is tracking how the prediction percentages map to the actual results so far.  They're mostly as expected.

[MuMark]

Pomeroy doesn't deal with intagibles. He is a math guy.

Think of it this way....If you are a 51% favorite to win 2 games is it more likely that you will go 2-0 or 1-1?

That in a nutshell is why the season recrod doesn't match the individual game predictions.