Tuesday, October 14, 2014

College Football Math: Week 8 Watchability, Dashboard Updates and a Cool Graph


Before we start, I've added an easy reference collection of everything college football-related. It has links to all my key college football-related material. Just follow the blue arrow.


For the main event:

Last week I was asked how likely we are to end the college football season with no undefeated teams, 1 undefeated team, 2 undefeated teams, and so forth. I think the question was from an Oregon fan grasping for hope. 

I started looking into it, and then I remembered: anything that's worth doing is worth overdoing. So instead of only looking for how many teams will go undefeated, I went a little further, and made this. It's all of the possible number of losses, by team, in one chart. It shows the likelihood we end up with each number of teams for each amount of losses. For some specific examples:
  • There's a 34% chance with end up with 0 teams that have 0 losses. In other words, there's a 34% chance we end up with no undefeated teams. 
  • There's only a 19% chance we end up with a winless team (SMU or Kent State)
  • 5 losses is the most likely number of losses across the FBS



Team Dashboards
    • I've updated the both the Team Dashboard and the Expanded Team Dashboard through week 7 
      • Who can guess which teams are a) most likely to win out and b) most likely to lose out? A couple of the teams totally surprised me.

    Watchability
    • Week 8 Watchability index is below
    • This seems like a good mind to restate that Watchability is a context neutral statistic. It doesn't know which team is your favorite, it doesn't if a game has CFP implications, it doesn't know if a game is a bowl game. It doesn't know if a game is a rivalry game, or the rose bowl, or the CFP championship game. All it knows is how good the teams are and how close the game is likely to be.

    Thursday, October 9, 2014

    College Football Math: Week 7

    Last week 11 of the top 25 teams lost, including 4 teams of the top 6. I don't know that we'll top the excitement of week 6 for a very long time. 

    But don't fret! This week looks great. Every time slot has at least one really fun looking game, and with all the chaos of the last week each game is critical for every team with designs on the CFP.

     

    Team Dashboards
      • I've updated the both the Team Dashboard and the Expanded Dashboard through week 6 
        • Georgia Southern is having a great first season in the FBS. They're a lock to go to a bowl and could end up 10-2
        • My former universities range from decent (tOSU, UW) to awful (UNLV)

      Watchability
      • Week 7 Watchability Schedule is below
      • Watchability is a measure of how good the two teams are, and how close the game is likely to be
      • This post has way more detail on everything. Quick refresh:
        • Watchability is a combination measure of how good the two teams playing are and how likely the game is to be close
        • Teams are shaded by their chance to win, the greener the better

      Wednesday, October 8, 2014

      College Football Math: Rankings

      This week’s AP poll has Arizona has #10. If you take that to mean they should be favored over every team ranked below them, it's an absurd ranking. 

      Arizona had a great game against one of the best teams in the country last week, and they’re unbeaten. Except they needed a hail mary to beat Cal at home, and their other wins came over UNLV, Nevada, and UTSA. Based on what we've seen thus far, Arizona is not the 10th best team in the country. They might become that team, they might get into the CFP and win a national championship. But odds are they won't. If #10 Arizona isn't the 10th best team in the country - it begs the question: what kind of rankings are the AP voters doing?
      There are two bases on which one could rank football teams: achievement and prediction.


      Achievement
      • Achievement rankings say something about the past. They order teams based on what those teams have done on the field - who they've beaten and by how much
      • As a fan, this feels more satisfying since teams can only control what they do on the field
      • But rankings based solely on achievement are bad at predicting the future
      • Achievement rankings might be:


      That looks satisfying. We could quibble over the specifics of my methodology and the placement of specific teams in specific spots, but I think they work well enough. Most of the undefeated teams are near the top, Arizona included. Utah gets a nice bump for their win over UCLA, and Ole Miss/Auburn are high. UCLA is high too. They did lose to Utah at home but they've also played a tough schedule and they walloped ASU.
      The rankings look nice, but unless you think Vegas oddsmakers are very wrong about a lot of games, these rankings are a disaster from a predictive standpoint. Some examples:
      • They rate Stanford a 5.0 favorite over WSU (Vegas has them 17.5).
      • They rate Notre Dame a 35 point favorite over UNC (Vegas: 16.5 points).
      • They rate Arkansas an 12.5 point favorite over Alabama (Vegas: 9 point underdogs)  

      Overall, the correlation between lines generated by these “Achievement” rankings and lines generated by bookies is 0.66. That's a terrible correlation.
      It looks to me like the AP voters are ranking teams on Achievement, with some level of recency bias. That’s fine, but if that’s what they’re trying to do, we should stop thinking of teams as necessarily better at football than teams ranked below them.
      Predictive

      If the AP voters are trying to do predictive rankings, they’re doing it wrong. Football is full of craziness, good days, bad days, and the predictive power of any single football game isn’t as strong as the AP voters seem to think.
      • Predictive rankings attempt to say something about the future. Each team would be an underdog against every higher-rated team, and a favorite against every lower-rated team.
      • This is what my models rankings are - an attempt to identify which teams are the best teams
      • This week’s predictive rankings look like:


      That certainly looks less satisfying; at least it does to me. Namely, it seems wrong that Alabama is first and that Oregon/Stanford/UCLA/Oklahoma are all in the top 10 despite losing this week. But these ranking jive much better with this week’s odds:
      • They rate Stanford a 19.3 favorite over WSU (Vegas has them 17.5).
      • They rate Notre Dame a 17.2 point favorite over UNC (Vegas: 16.5 points).
      • They rate Arkansas an 8.8 point underdog to Alabama (Vegas: 9 point underdogs)  

      Overall, the correlation between lines generated by these “Predictive” rankings and lines generated by bookies is 0.95. That's a very good correlation.

      All this extra look into the model's ratings means I'm a couple of days behind on the regular steps of updating. I'm running a simulation now and should have the dashboards and Watchability ready to go tomorrow.

      Wednesday, October 1, 2014

      College Football Math Week 6

      The 90th percentile Watchability rating for the 2014 regular season is 63.7. What that means is that of all the college football games played in 2014:
      • 90% have Watchability ratings below 63.7
      • 10% have Watchability ratings above 63.7. 


      In week 1 there were 5 "90th percentile" games, games with Watchability higher than 63.7, and i
      n weeks 2-5 there were 3, 1, 2, and 4 such games respectively.

      What about this week? This week there are 10 such games. 10! That's more than any other week in the regular season! Let's look at those 10:



      It's an embarrassment of riches. Particularly at 12:30 PM on Saturday. Enjoy this weekend, it could be one of the best of the year.



      Team Dashboards
        • I've updated the both the Team Dashboard and the Expanded Dashboard through week 5. 
          • Go check out Michigan, Penn State, and Pittsburgh, they all look hideously depressing
          • In each and every game this season the following teams have performed worse than my model expected: Florida State, North Carolina, Utah State, Southern Methodist, UNLV, Kent State. The graphs for each of these teams go down down down down.

        Watchability
        • Week 6 Watchability Schedule is below
        • Watchability is a measure of how likely a game is to be a close, well-played game
        • What a great week. Thursday night, Friday night, Saturday morning, afternoon and evening, there's always an interesting game to watch.

        • This post has way more detail on everything. Quick refresh:
          • Watchability is a combination measure of how good the two teams playing are and how likely the game is to be close
          • Teams are shaded by their chance to win, the greener the better

        Tuesday, September 23, 2014

        College Football Math: Week 5!

        Team Dashboards


        Watchability
        • Week 5 Watchability Schedule is below
        • Watchability is a measure of how likely a game is to be a close, well-played game
        • UCLA@ASU, Stanford@UW, and OSU@USC??? The Pac-12 rules this week's Watchability Index
        •  Also I guess Missouri is playing South Carolina?


        • This post has way more detail on everything. Quick refresh:
          • Watchability is a measure of how likely a game is to be a close, well-played game
          • Teams are shaded by their chance to win, the greener the better

        College Football Expanded Team Dashboard

        Good morning good morning! Thanks for stopping by.

        As I used my College Football Team Dashboard, I often found myself wondering about things that weren't shown. The dashboard is a snapshot of a team's current state, I wanted to know where teams had been and where they were going.

        So I built an Expanded Team Dashboard. It's meant to complement the original team dashboard and there's very little information that overlaps between the two.

        It's too wide for me to do my traditional blue text & arrows, so bullets will have to suffice.

        Rice is chosen as my example team to remind a certain good friend how terrible his college is at football. Go to the bottom of the page to see the Rice example.
        • Team Rating and Rank from Preseason to the current week
          • Rating is on top and in blue
          • Rating goes from 0 to 1. Think of it as the likelihood that Rice would beat an average college football team
          • Rank is on bottom and in orange
        • Rest of the Regular Season
          • This is a summary of what the model expects for Rice. It doesn't look great.
          • Expected total wins (and losses) is an average outcome for the season. When I simulate the rest of Rice's season over and over again, the average number of wins I get for Rice is 3.9.
          • Chances of winning (or losing) out is the chance Rice wins (or loses) all its remaining games
        • Strength of Schedule
          • Here's a a look at the average strength of Rice's opponents thus far (Games Played), and average strengths of Rice opponents for rest of the season (Games Remaining). 
          • Rice's opponents so far (Notre Dame, Texas A&M, Old Dominion) have an average rating of 0.611.
          • Rice's opponents the rest of the year look to be a cast of clowns
          • Rice's easiest and toughest remaining opponents are also highlighted
        • Most Likely Bowl Games
          •  It looks like Rice still has an outside shot at a bowl
          • For some teams, like those even worse than Rice, fewer than 5 bowls will populate. Those teams are either terrible or have specific bowl tie-ins like BYU
          • The bowl logic was among the hardest things to code. Please don't hesitate to point out anything you think is a flaw
        Any other questions? Let me know. 

        Check out the Rice dashboard below then head here to make one for your team.

        I hope you find the new extended dashboard as interesting as I do!






        Tuesday, September 16, 2014

        College Football Math: Week 4 Watchability

        Team Dashboards
          • I've updated the College Football Team Dashboard through week 3. Go check it out
          • Come back tomorrow on Thursday some time! ve been working on an expanded team dashboard I'm going to publish tomorrow night. I'm excited about it; I think it's pretty awesome!


          Watchability
          • Week 4 Watchability Schedule is below
          • Watchability is a measure of how likely a game is to be a close, well-played game
          • The first game of the week is the most watchable! Auburn gets is first real test of the year as they travel to Kansas State. Maybe. I feel like Kansas State always starts the year #17 then goes 8-4 and loses the Liberty bowl
          • I'm super intrigued to watch Utah vs. Michigan.
          • I think New Mexico @ New Mexico state probably wins the crummy game of the week award.




          • This post has way more detail on everything. Quick refresh:
            • Watchability is a measure of how likely a game is to be a close, well-played game
            • Teams are shaded by their chance to win, the greener the better