2013 College Football Modeling: Week 5 Game by Game Odds

• Model is updated with this week's game outcomes
• Stanford's drubbing of WSU cost me $10. Pullman - this is why you can't have nice things
• The UW-Arizona game was the rainiest game I've ever been to. Go Huskies :)
• I've expanded the report with two new items:
• On the right, your selected team's rank at the start of each week (according to my model). The rankings are at the start of the week, so week 1 shows the preseason ranking, week 6 (games to be played this upcoming saturday) shows the current ranking
• The graph at bottom shows each possible win total and how likely it is for your chosen team. For example, Washington has a 13% of getting exactly 10 regular season wins. 

2013 College Football Modeling: Week 5 Viewing Guide

• I will be at this week's UW game! Go Huskies!
• In a moment I expect regret bitterly, I bet that the Cougs would upset Stanford
• How much fun would it be to see Ole Miss upset Bama? Lots of fun, that's what I say.
• Apparently USC is still playing football?
• As always
• Watchability measures how good the teams are and how likely the game is to be close (Way more detail on what Watchability here)
• The more green a team is shaded, the better its chance to win; the more red, the worse
• Ranking (when listed) is this week's AP ranking
• All times Pacific

2013 College Football Modeling: Study break / math art

My next actuarial exam is coming up soon and many evenings I'm at the office late studying. I put this together during one of my study breaks.

We're looking at win probability (usual green/yellow/red shading) for every FBS team theoretically matched up against every other FBS team on a neutral site. Each team has a row. Bama's row is mostly green (Bama would be a strong favorite against most teams), UNLV's row is mostly red. 

I stripped out the numbers; it's more fun (for me anyway) to enjoy just the colors. If you click on it it'll blow up to full screen. If you right click then select "Open image in new tab" it'll open up full size roughly 2600x2600 pixels)

It's not meant to provide in depth analytic insight so much as just be fun to look at.

2013 College Football Modeling: Week 4 Game by Game Odds

• Game by game odds tool has been updated with outcomes from week 4
• Not the most exciting week, but a good win for Utah!
• Watchability took a grim view of last week. This week (and the weeks to come!) look much much better. More on that on Thursday.

2013 College Football Modeling: Week 4 Viewing Guide

• Looks like a slower week (particularly after the drama of ASU-Wisc & OSU-Utah) but the top few games should be good.
• UW is a 7 touchdown favorite
• It's Holy War in the state of Utah and I have BYU as a 70% favorite (sorry extended family!)
• Notre Dame/Michigan State and Rutgers/Arkansas forecast as very even matchups
• As always:
• Watchability is designed to capture how good the teams are and how likely the game is to be close (Way more detail on what Watchability here)
• The more green a team is shaded, the better its chance to win; the more red, the worse
• Ranking (when listed) is this week's AP ranking

2013 College Football Modeling: Week 4 Game by Game Odds

• Win probability tool is updated with week 3 outcomes. Texas, Michigan, and Nebraska all look much worse off. Maybe Michigan will think twice next time before overlooking the mighty Zips of Akron.
• Nice win, Huskies!
• The Watchability measure did well to find some close, exciting games, so I'm excited about that. I should probably give an assist to the refs for ASU/Wisconsin. I'll be publishing a week 4 viewing guide on Thursday.
• Something new that's coming: Looking at this game chart, it's difficult to surmise how likely Wisconsin is to win 4, 5, 6, etc. games. I'm working on a tool that'll let you input a team and see exactly how likely that team is to win 0, 1, 2, or 12 (and all the numbers in between!) regular season games. 

2013 College Football Modeling: Week 3 Viewing Guide

As we head into another glorious weekend of college football, I'd like to share a Week 3 Viewing Guide I've been working on. The guide includes a new metric I've been working on called Watchability. My purpose in inventing Watchability is to help people identify close, well-contested games they might not have otherwise thought to watch. Going forward I plan on posting one each week on Thursday mornings. I tried to include as much information as I could without being overwhelming. For each game I'm showing:

• Day and time (Pacific Time)
• Home team and away team
• Teams are shaded according to the model's estimation of their chance to win
• The more green a team is shaded, the better its chance to win; the more red, the worse
• Ranking (when listed) is this week's AP ranking
• Teams who have already played are italicized and shaded fully red or green as appropriate
• Watchability
• This is a measure (scaled from 0 to 100) which attempts to capture how likely the game is to be a good game
• Its purpose is to identify games that might be fun to watch
• It's based on how good the two teams are, and how likely the game is to be a close game
• The theoretical max of 100 would be two teams rated 1.000 (team ratings range from 0 to 1) playing each other on a neutral field
• The highest rating I've seen so far has been around 95 in some simulations of the BCS National Championship Game featuring undefeated teams from the Pac12/Big10/SEC
• Oregon @ Stanford at 80.9 is currently the highest rated regular season game
• There is a little more math on Watchability below

I calculate Watchability as follows:

I begin by calculating a Power Rating for each game. This involves taking the model's average rating for the two teams and scaling it from 0-100. A game's Power Rating is a measure of the combined strength of the two teams playing.

Power Rating treats Oregon (.892 rating) vs. Tennessee (.563) the same as Arizona State (.672) vs. Wisconsin (.727). The Power Rating for Oregon vs. Tennessee is 72.7, while the Power Rating for ASU vs. Wisconsin is 70.45.

I then translate Power Rating into Watchability by calculating on how far each team's chance to win the game deviates from 50% and reducing Power Rating accordingly. The reduction is based on the square root of the deviation, so 60/40 games are only slightly penalized relative to 50/50 games, but 90/10 games are penalized heavily because blowouts are so boring. Take the two example games: Oregon is a 90% favorite so that game's Watchability is 27.2 (reduced significantly from its power rating of 72.7). On the other hand, ASU and Wisconsin are nearly 50/50, so that game's Watchability is 70.43 (nearly identical to its Power Rating of 70.45).

A few last comments:

• Watchability is a context-neutral statistic. It doesn't know which team is your favorite, it doesn't know that this game has BCS implications, it doesn't know that this is a bowl game. It doesn't know that this is a rivalry game or the Rose Bowl or the BCS Championship Game. All it knows is how good the teams are, and how likely the game is to be close.
• For context, I added average Watchability for some different types of games as an output for the last round of 10,000 season simulations and calculated the following:

• Regular season games average around 30 Watchability, conference title games & non-BCS bowl games average 50, BCS bowls average around 70, and the BCS National Championship Game averages around 75.
• Watchability does not predict defied-the-odds kinds of upsets. Those are just surprises to be enjoyed by us all.

2013 College Football Modeling: Week 3 Game by Game Odds

Model is updated with week 2 game scores. 

• In honor of their amazing win @USC I'm featuring Washington State as the default loaded team
• On Sunday, I ran 10,000 simulations of the rest of the season. USC and Texas won the national championship one time between them. Anyone care to guess which one?
• At this rate ten Pac-12 teams will be bowl eligible. I wonder what the record is for bowl eligible teams from one conference
• USC didn't even have the worst week this week, their rating only dropped by 77 points. Texas (spanked by BYU) and Cincy (rolled by Illinois) each dropped over 100.

2013 College Football Modeling: Week 2 Game by Game Odds

Model is updated with this week's game scores.

I asked the model, and it says UW could not have looked better this week. Also the model thinks Oregon should be ashamed for scheduling a game against a Jr High School.

• I've added some scenarios and their different likelihoods at the bottom (looks like I calculated Oregon State's likelihood of going undefeated correctly)
• I've replaced week 1 win probability with week 1 results
• Type any FBS team in the gray box, hit enter, and watch the table populate
• As always please let me know if you find any scheduling or outcome errors

• Best week 1 prediction: WSU would lose to Auburn 7.2  (they lost by 7)
• Worst week 1 prediction: UW would only win by 1 (they won by um.. more)
• Had I bet on my model in any game where it differed from Vegas by 5 points or more* I'd be up $382. I don't expect this will even stay above $0 (beating Vegas is hard!) but it'll be fun to track through the season

* Based on betting increments of $22 at -110 ($22 wins you $20). 1 increment where model in 5 points different from Vegas, an additional unit for each point thereafter.