NCAA is meeting with quants to make tournament selection process better

On Friday, January 20th, 2017, hell will freeze over, and the NCAA will meet with analytics guys like Ben Alamar, Jeff Sagarin and Ken Pomeroy. The conversation will revolve around making the tournament selection process better.

You can read about it here, but two points stand out for me.

First, they say the following about the RPI rankings the committee currently uses.

An even more powerful microscope to go with the time-honored RPI.

Time-honored my ass. The RPI is stupid for two reasons:

  • It lacks a solid mathematical basis (compare it with the least squares rankings that Pomeroy uses)
  • It uses wins and losses instead of margin of victory in its calculations

I discuss both of these issues in relation to college football here. Hence, RPI fails as a predictor for how teams fare in the tournament.

The NCAA should eliminate RPI from the selection process.

Second, Jim Schaus, the athletic director at Ohio State and committee member, said this:

I’m going to have to strap on in the meetings to stay up with all the calculus that’s going to be discussed, but I’m excited about it.

Calculus is so overrated in our society.

You want to hang with the quants, Schaus? Then let’s talk probability, or that no analytics ever says a team will beat another team with 100% certainty.

Want to get fancy, Schaus? Then let’s dig into linear algebra so you can understand the least squares method used in adjusting for strength of schedule.

I’m all for learning calculus. It’s just not as useful in sports analytics as probability and linear algebra.

Podcast: Ben Alamar on the ESPN Football Power Index and NFL Divisional Playoffs

On this week’s episode of the Football Analytics Show, I’m honored to have Dr. Ben Alamar, Director of Sports Analytics at ESPN. We dig into a host of topics, which includes:

  • The NBA executive that opened the conversation with “I don’t believe this analytics stuff”
  • How FPI (Football Power Index) makes adjustments for quarterback injuries
  • The surprising team to which FPI assigns the second highest Super Bowl win probability
  • The hidden factor that gives Dallas an extra point in the FPI prediction against Green Bay
  • Spread predictions for all 4 Divisional Playoff games for FPI (not available anywhere on the internet to my knowledge) and my member numbers
  • Ben’s 3 tips for breaking into sports analytics

You can check out the work of ESPN analytics team by clicking here.

To listen to the show on iTunes, click here.

To listen here, click on the play button:

How Andy Reid wins football games with interceptions

Andy Reid’s teams throw a low rate of interceptions. The visual shows how his teams in Philadelphia and Kansas City have had a lower than NFL average interception rate (interceptions divided by attempts) in all but 3 of 18 seasons.

Reid’s Eagles had a particularly good stretch of pick suppression from 2000 through 2004. Led by quarterback Donovan McNabb, Philadelphia never won fewer than 11 games in any of those 5 seasons.

Despite a decreasing interception rate across the NFL, Reid has continued to beat the NFL average over the past four years in Kansas City. Led by quarterback Alex Smith, the Chiefs have won 43 regular season games and never dipped below 9 wins in any one season.

The randomness of turnovers

The visual goes against the typical quant narrative that turnovers are random.

For example, I’ve shown this visual that shows the relation between interception rate the first 6 games of the college football season versus the remainder of the season.

The lack of correlation between these quantities shows that you can’t predict a team’s interception rate later in the season based on the same quantity during first 6 games.

This suggests interceptions are random, and a team has a 50% chance to have a better or worse than average interception rate. However, if you assume this for Andy Reid’s teams, there’s only a 0.37% chance his teams would have had 3 or fewer seasons with a below average interception rate.

Randomness certainly plays a role in interceptions. No one who has ever seen a tipped pass fall into the hands of a defender should doubt that.

However, the Reid visual suggests that some coaches can suppress interceptions over a very large sample of games.

Steelers at Chiefs

This has implications in predicting the outcome of the Steelers at the Chiefs playoff game this weekend.

Kansas City doesn’t seem like much of a Super Bowl threat with the 16th and 11th ranked pass offense and defense, respectively, by my adjusted yards per attempt. I use these pass efficiency numbers to evaluate teams for two reasons:

  • My research shows the importance of passing over rushing in the NFL.
  • Turnovers have little impact on yards per pass attempt.

However, if Kansas City is truly skilled at not throwing interceptions, then these pass efficiency numbers will underestimate their team strength.

Team rankings based on adjusted margin of victory might be a better way to evaluate Kansas City. Their low interception rate will impact margin of victory, as I’ve found that an interception is worth 5 points in the NFL.

My member numbers combine both pass efficiency and margin of victory to make Kansas City a 1.3 point favorite against Pittsburgh. However, my team rankings based on only points would make Kansas City a 3 point favorite.

I interviewed Ben Alamar on the Football Analytics Show this week, and his FPI (Football Power Index) makes the Chiefs nearly a 5 point favorite. They use an expected points added, a metric which accounts for interceptions but their own twist. To listen to that part of the discussion, go to 14:50 of my interview with Ben Alamar.

The top 5 insights into Alabama vs Clemson, Part II

How will Clemson match up with Alabama in the College Football Playoff championship game? Will the result be any different from last year, when Alabama claimed their 4th national title in the last 7 years?

Let’s count down the top 5 insights that will affect the outcome of this game. Yes, we’ll revive the Star Wars analogy from last year’s preview.

5. Clemson dominated the line of scrimmage in Part I

Last season, my numbers favored Alabama by 6.4, and the Crimson Tide won 45-40. Solid victory for analytics, right?

No. Clemson clearly won the line of scrimmage. I’ve never seen this type of domination over Alabama, as the Clemson defensive line in particular made a mockery of attempts to block them.

Despite this edge in the trenches, Clemson lost due to numerous blown coverages in a secondary that had excelled all season. Alabama also recovered an on side kick to provide an extra possession, and it was enough to hold onto a 5 point win.

Will Clemson dominate the line of scrimmage again in Part II against Alabama? Probably not.

However, consider that Clemson’s defense has put 4 and 7 starters in the 2015 and 2016 NFL drafts, respectively. Despite these losses, they continue to dominate, as they allowed 4.56 yards per play, 4th best in the nation this year.

4. Mike Williams vs Marlon Humphrey

Clemson QB Deshaun Watson has many solid receivers, but none more talented than Mike Williams. Watson has targeted Williams 129 times this year for 9.8 yards per target.

Many expect Williams to be a first round NFL draft pick. He can improve his draft stock in the championship game as he will face Alabama cornerback Marlon Humphrey, another projected first round pick.

We’ll put numbers behind the transcendence of this Alabama defense later. But for now, let’s focus on the secondary in the semi-final game against Washington.

Dante Pettis and John Ross had terrorized Pac-12 defenses all season, as Washington had a potent pass offense had gained over 8.5 yards per attempt heading into the Playoff. (Numbers include sacks unlike traditional college football statistics.) Against Alabama, these top receivers had difficulty getting open, as Washington threw for a meager 2.6 yards per attempt.

Clemson’s Williams might have more talent than any of Washington’s receivers, but we’ll see whether he can get open against the Alabama secondary.

3. Can Jalen Hurts lead a come from behind victory?

Before the start of this season, Nick Saban told ESPN “I don’t care who we start at quarterback. He ain’t going to be that good.”

True freshman Jalen Hurts won the job, and he has excelled beyond all expectations. Hurts has completed 64.7% of his passes (college football average is 60%) as well as gained 6.2 yards per carry as a dual threat quarterback.

But how will the freshman play if he must bring Alabama back from a late game deficit? He hasn’t face that situation yet this season.

My member numbers favor Alabama by 9.4 points. However, the clear favorite doesn’t always have a late game lead.

What happens when Alabama no longer gets points from the defense or special teams? They have scored 15 non-offensive touchdowns this season, although randomness has certainly played a role in this astounding number.

If Alabama faces a late game deficit, Hurts will have to throw the ball. This plays right into the strength of the Clemson defense, as they have the 3rd best pass defense by my adjusted yards per attempt.

Then if Clemson doesn’t make the same blunders in the secondary as last year, they could very well pull off the upset.

2. The Alabama defense

In my preview of last year’s game, I made a Star Wars analogy for Alabama.

But Alabama is college football’s empire, a finely oiled machine with infinite resources to destroy the opponent. Their defense is a Death Star aimed at Clemson and another national title.

This season, Alabama acquired more potent Kyber crystals to power their Death Star.

The defense remains the top ranked unit by my adjusted yards per play just like last season. But while they projected to allow 4.3 yards per play to an average FBS defense last season, that number has dropped to 3.9 this season.

To see how the Alabama defense match ups with Clemson’s potent offense, I use data visualization to plot opposing units on the same line. Better defense appear further to the right to facilitate comparisons with the opposing offense. The unit that appears further to the right is predicted to have the advantage.

The visual shows how offense and defense match up in Alabama vs Clemson, Part II.

The top chart shows the expected dominance of Alabama’s defense over Clemson’s offense.

The bottom visual shows an even match up between Alabama’s offense against Clemson’s defense.

1. Deshaun Watson, adolescent Jedi

In keeping with the Star Wars theme, I said the following about Clemson QB Deshaun Watson in last year’s preview.

However, Clemson has played exceptional this season, and QB Deshaun Watson has taken a starring role. He’s a young Jedi beginning to use his full powers, just like Luke Skywalker in the New Hope.

Watson was spectcular against Alabama last year, throwing for almost 8 yards per attempt, a number that accounts for sacks unlike traditional college football statistics. This prompted Nick Saban to call Watson the best college QB since Cam Newton.

If Watson was Luke Skywalker in the New Hope last season, he is now Luke in Return of the Jedi this season. He’s had his full training from Yoda, and he won’t surprise anyone with a big game against Alabama.

Can Deshaun Watson blow up Alabama’s death star? He must have a spectacular game for Clemson to convert their 24.5% chance for the upset.

Podcast: Cade Massey on the NFL Playoffs and College Football Championship Game

On this week’s episode of the Football Analytics Show, I’m joined by Cade Massey, professor at the Wharton School at the University of Pennsylvania. He studies judgment under uncertainty, and there’s no better example than his Massey-Peabody football predictions.

We cover a wide range of topics, which includes:

  • How Cade has learned humility in building a predictive football model
  • The playoff karma of the New York Giants
  • What data says about whether match ups matter in football predictions
  • The sneaky trick for breaking into the sports analytics world
  • The Massey-Peabody prediction for Alabama versus Clemson

For match ups, I discuss a similar study in college basketball.

To listen on iTunes, click here.

To listen here, click on the play button.