This is related to some discussion we had below, but I don't really want it to be about Crean or where we are as a program, so I wanted to makes this post, for the benefit of those who find it interesting to discuss stats, and so forth. Basically, we have to start with the understanding that an actual W-L record doesn't necessarily accurately measure a team's performance. For example, teams that win all of their close games, but get blown out in all of their losses, will tend to have a better record than they should. Similarly, teams that lose a lot of squeakers, but blow weaker teams out of the water, will tend to have worse records than they should. This is about how to measure a team's performance. It is based only on results in the B1G.
In the first nine games, IU is scoring, on average, 0.44 points fewer than opponents. Now, due to the fact that IU has had a slightly tougher than average schedule, and had to play five games on the road, we can adjust for those factors. When we do, IU is actually scoring .11 points better than opponents, which, FWIW, equates to 6th in the B1G. Using Pythagorean expectation, we can turn this margin into an expected win percentage, which is .506. So you'd expect either 4 or 5 wins. In fact, if you compare our margin to our opponents, adjusted for home-court, we should have been favorites in three games only: Mary, PSU and RU, and very slight dogs against OSU at home and at Nebraska and Illinois. That sounds like a recipe for a 4-5 record. But we have done two wins better than that. Here are our games, margins of victory, predicted margins, and differences.
Opp IU MOV Pred O/U
Mary 19 3.9 15.1
OSU 3 -0.9 3.9
Ill 6 -1.3 7.3
Neb 5 -0.2 5.2
PSU 3 5.6 -2.6
RU 8 12.6 -4.6
osu -12 -9.0 -3.0
PU -16 -6.7 -9.3
MSU -20 -8.0 -12.0
There are two ways of looking at this. First, the majority of our poor performance came in games we should have lost anyway. Second, the three games we were very slight dogs in are all three games we slightly over-performed. In other words, we were able to use a little bit of good performance to great effect, while hiding a large amount of poor performance behind likely losses. The primary exceptions to this pattern are Maryland, whom we dominated, and Rutgers, whom we were terrible against, but it didn't matter because they are awful. In other words, again, both wins.
Now, this is all just a mental exercise unless you look at what this means for the second half of the season. Of our nine remaining games, we are favorites in all but two (Mary, Wiscy), although three of them are very close, virtually toss-ups (PU, NW, MSU).
Wiscy -15.2
UM 4.4
Mary -4.2
Minny 5.1
PU 1.4
RU 4.5
NW 0.1
Iowa 4.2
MSU 0.1
If you look at our win expectation for each game, the odds are that we'll go 5-4 to finish the season, but if we play up to potential in all of the close games, like we've done so far, we have a very real chance to finish 7-2, and 8-1 is not out of the realm of possibility. In other words, mathematically, our expectation should be to finish 11-7 when all is said and done, and 10-8 is almost as likely. But the difference between 10-8 and 14-4 may very well end up being just a couple of buckets in a small handful of very evenly matched games - exactly the type of situation we've been able to take advantage of so far.
goat
This post was edited on 2/1 5:10 PM by TheOriginalHappyGoat
In the first nine games, IU is scoring, on average, 0.44 points fewer than opponents. Now, due to the fact that IU has had a slightly tougher than average schedule, and had to play five games on the road, we can adjust for those factors. When we do, IU is actually scoring .11 points better than opponents, which, FWIW, equates to 6th in the B1G. Using Pythagorean expectation, we can turn this margin into an expected win percentage, which is .506. So you'd expect either 4 or 5 wins. In fact, if you compare our margin to our opponents, adjusted for home-court, we should have been favorites in three games only: Mary, PSU and RU, and very slight dogs against OSU at home and at Nebraska and Illinois. That sounds like a recipe for a 4-5 record. But we have done two wins better than that. Here are our games, margins of victory, predicted margins, and differences.
Opp IU MOV Pred O/U
Mary 19 3.9 15.1
OSU 3 -0.9 3.9
Ill 6 -1.3 7.3
Neb 5 -0.2 5.2
PSU 3 5.6 -2.6
RU 8 12.6 -4.6
osu -12 -9.0 -3.0
PU -16 -6.7 -9.3
MSU -20 -8.0 -12.0
There are two ways of looking at this. First, the majority of our poor performance came in games we should have lost anyway. Second, the three games we were very slight dogs in are all three games we slightly over-performed. In other words, we were able to use a little bit of good performance to great effect, while hiding a large amount of poor performance behind likely losses. The primary exceptions to this pattern are Maryland, whom we dominated, and Rutgers, whom we were terrible against, but it didn't matter because they are awful. In other words, again, both wins.
Now, this is all just a mental exercise unless you look at what this means for the second half of the season. Of our nine remaining games, we are favorites in all but two (Mary, Wiscy), although three of them are very close, virtually toss-ups (PU, NW, MSU).
Wiscy -15.2
UM 4.4
Mary -4.2
Minny 5.1
PU 1.4
RU 4.5
NW 0.1
Iowa 4.2
MSU 0.1
If you look at our win expectation for each game, the odds are that we'll go 5-4 to finish the season, but if we play up to potential in all of the close games, like we've done so far, we have a very real chance to finish 7-2, and 8-1 is not out of the realm of possibility. In other words, mathematically, our expectation should be to finish 11-7 when all is said and done, and 10-8 is almost as likely. But the difference between 10-8 and 14-4 may very well end up being just a couple of buckets in a small handful of very evenly matched games - exactly the type of situation we've been able to take advantage of so far.
goat
This post was edited on 2/1 5:10 PM by TheOriginalHappyGoat
