Member for

8 years 9 months

Recent Comments

Date Title Body
07/28/2016 - 8:49pm Championship Game Probabilities

Taking Bill C's win probabilities and running them through a Monte Carlo analysis with the division champion determination rules yields the following estimates of which teams will appear in the B1G Championship:


So Michigan proably won't win the East, but nobody has a better shot.

07/27/2016 - 10:17am Sum Distribution

04/26/2016 - 9:42pm In red earlier?

Want to Aston Villa in red earlier? I re-scaled red to 30 points behind.

04/26/2016 - 12:21am Season Perspective

Some perspective on how the season has gone in the league. Plot shows points behind as a function of matchday for each team. You can see that Leicester City has never been too far out of the lead. Man City looked like the odds-on favorite for the first half of the year. And Arsenal was there in the middle. But in the final third, Leicester City has pulled away:

02/26/2015 - 5:39pm Some Answers
1. What school is officially located in two cities?  (That’s an easy one to start with)
2. What two schools are located in the same city as each other.
3. Only conference where the cities of two schools have the same name (as an example -- Maintown NY and Maintown CA)
Columbia SC, Columbia MO
4. Only conference where two schools and their cities each have the same names as each other (as an example -- Smith University in Smith, NY)
ACC (Syracuse & BC & Clemson)
5. What other conference has one school with the same name as its city
SEC (Auburn)
6. What other conference has one school with the same name as its city, except that the city has the word “City” appended to it (as an example – Smith University and Smith City, NY)
B1G (Iowa)
7. What conference has the most number of cities with 2 words in each city name (example of a city name with two words:  “Main City”), and how many are there in that conference.
8. What city has 3 words in its name
9. What conference with 3 city names that are traditionally first names of males
10. Of all 64 P5 cities, which city's name is first alphabetically:  Hint:  It's not Ann Arbor.
11. Of all 64 P5 cities, which city's name is last alphabetically
08/25/2014 - 9:47am How I See the Numbers

01/07/2014 - 1:41pm You've got the conditional evaluation backwards.

We don't want to know what fraction of good rushing offences are pro-style. Rather, we want to know what fraction of pro-style offense rush well. It's certainly harder to figure that latter one out, but the first one doesn't tell us much.

09/11/2013 - 12:33pm All the Better

If Gardner wins the Heisman in #98 this is all to the good. Then #98 will be associated with Harmon AND Gardner. And some up can coming QB will want to be a part of that -- to add to that tradition of excellence. This idea that a guy needs his own number to make an impact is misguided. Michigan football existed for a hundred years before they got here and will exist a hundred years after. The best they can hope for is to add to the legacy that’s been established. Nobody owns a number because nobody is bigger than the program. Case in point: just because Anthony Carter made a huge impact in #1 doesn't mean we don't remember Braylon. 

08/01/2013 - 10:53pm The right way to think about season win totals

So adding up his win probabilities gets you to 8.3. I also think the expectation is closer to 9 than 8. I would add a few tenths against Penn State, Michigan State, and Notre Dame. But I would subtract a tenth against Ohio State.

But I do think it's important to think about this from a Bayesian perspective. These aren't estimates some Platonic "true" win probabilities -- such a thing doesn't exist. These are summaries of the information we have up to this point.

06/26/2013 - 10:50pm No

I'm pretty sure it's 8.64.

06/26/2013 - 9:56pm Uncertainty Distribution

Yeah there is a lot of uncertainty, but specifying the outcomes using this formulation allows you to look at that too. Here is the probability of winning 0, 1, ..., 12 games assuming the win percentages I gave above:

06/26/2013 - 10:43am Great point

There is a 0% chance that the team wins exactly 8.6 games. And given that we have only one outcome of the season, we don't know which way it will turn out. But two things:

1. If we flip a coin 10 times we are much more likely to get 5 heads than we are to get 10 (and we're more likely to get 5 heads than we are any other number of them)

2. If we flip a coin 10 times, we are just as likely to have more than 5 heads as we are to have fewer, so the expected value is 5. It probably *won't* be 5, but there isn't any better guess.

So, sure, there is a chance of a 12-0 season. But it's a small chance, and if you want to say we're 12-0 is likey (> 50% chance), then that means that you are implicitly saying there's a 95% chance of winning each game. That's a tough sell to me.

06/26/2013 - 10:15am I See 8.6 Wins

I don't even understand people who aren't willing to accept that there is uncertainty in the outcome of *every* game. As such I don't think it's prudent to say something like "We'll split the OSU and ND games and probably lose one of the Neb., Northwestern, and MSU games, so that puts us at 10-2. Book it." Is there really *no* chance that M loses any other game? Isn't it best to account for that cumulative probability somehow?

So let's do something simple and just write down our estimates of the probability that M wins each of the games on their schedule. Here is how I figure it for each game. Where am I high or low? If you see 10 wins (or more) you've got to make some big changes from this.

Central Mich.   0.95
Notre Dame      0.45
Akron           0.99
at Connecticut  0.85

Minnesota       0.8
at Penn State   0.65
Indiana         0.9

at Mich. State  0.6
Nebraska        0.65
at Northwestern 0.6
at Iowa         0.8
Ohio State      0.4
SUM             8.64
05/03/2013 - 11:04pm Expected Value of Wins


The way I see it, the expected value of # of wins is around 8.4. For those with rosier outlooks, where should I put the extra probability? 
Game                      Probability of M win
Central Mich.          0.95
Notre Dame           0.45
Akron                       0.99
at Connecticut       0.9
Minnesota              0.9
at Penn State         0.65
Indiana                    0.9
at MSU                    0.6
Nebraska               0.65
at Northwestern    0.6
at Iowa                    0.8
Ohio State              0.4
Sum                        8.39
01/02/2013 - 4:33pm Estimated Win Probabilities


Here are the probabilities I'm thinking now:
Central Mich. 0.95
Notre Dame 0.45
Akron 0.95
at Connecticut 0.8
Minnesota 0.8
at Penn State 0.6
Indiana 0.8
at Michigan State 0.55
Nebraska 0.65
at Northwestern 0.7
at Iowa 0.75
Ohio State 0.4
Only 2 predicted losses, but it sums to 8.0. 
I think the total will be a bit higher than 8. I'd like to get it up to the 9.5 region, but I'm struggling with where to put the increases. Anyone with thoughts?
10/25/2012 - 11:56am Yeah that and I'm a sucker

Yeah that and I'm a sucker for a brand name like "Huskers Defense" no matter how tarnished it might be.

10/24/2012 - 6:53pm Ominous Warning

Brian is pretty down on the offense. We'll see what comes out in the UFR tomorrow. I can't say I disagree, but I'm still sticking with 55% chance we win this game. I think 24 points will do it.


EDIT: "This game" meaning the upcoming game against Nebraska

02/09/2010 - 10:22pm Sinking feeling?

It sounds like a great feeling to me.

I'm not sure what is so bad about a bunch of fast kids turning things around. And if they can do anything like the Fab Five did that would be great.

Unless you're worried that these new kids are on the take? Or maybe dreadlocks aren't your thing?

09/28/2009 - 11:26am OSU >? PSU

I'd say the Ohio State resume is more impressive than Penn State's. But I agree they ought to be close.