Podcast: College football championship week, Atlanta Falcons and Washington Redskins

thefootballanalyticsshow_cover_landscapeCollege football is getting real, folks. In this week’s podcast, I discuss the scenarios for all the top teams.

Ohio State isn’t a lock. There just aren’t any locks in sports, although it is highly likely the Buckeyes make the playoff.

Michigan isn’t dead. But they do need help, and maybe more than you think.

Then I transition to the NFL to discuss the Atlanta Falcons, the surprising top team in my member NFL rankings. We shall see how long they last there.

Last but not least, I look at the Washington Redskins and their potent offense. They have an interesting game at Arizona this week.

To listen to the podcast, click on the play button.

To listen on iTunes, click here.

The probability of chaos on championship week, 2016

Alabama, Clemson and Washington enter the weekend as big favorites over their conference championship foes. By my member numbers, here are their win probabilities:

  • Alabama has a 92.3% chance to beat Florida.
  • Clemson has a 79.9% chance to beat Virginia Tech.
  • Washington has a 72.2% chance to beat Colorado.

Still, it’s not out of the question they all lose. Assuming these games are statistically independent, you can determine this probability by multiplying the chances for the three teams to lose.

By my numbers, there’s a 0.4% chance that Alabama, Clemson and Washington all lose. Improbable but not impossible. This was about the same chance my numbers gave Clemson, Michigan and Washington to all lose in week 11 before it happened.

0.4% is also about the chance of drawing two Aces from a random deck of cards (1 in 221). Getting pocket Aces in poker is unlikely. However, any poker player has most likely had a night with multiple pocket Aces.

Don’t rule out the impossible.

College football playoff probabilities before championship week, 2016

screen-shot-2016-11-30-at-10-17-34-amOver at Bleacher Report, I wrote about my final college football playoff odds for the season. Alabama, Clemson and Washington are in with wins in their conference championship games, and the Big 12 is out.

There remains some uncertainty for the remaining top teams, most of which reside in the Big Ten. My article discusses the following:

  • Ohio State is not a lock, but it would take a lot to move them out of the top 4.
  • Michigan isn’t dead just yet.
  • How in the world is the champion of the conference with 4 of the top 7 teams on the outside looking in?

To check out the article, click here.

MLB free agents salary projections through analytics

This is a guest post from Nick Ceraso and Julian Frenkel, students at the University of Michigan.

How much should your team pay a free agent in baseball? Will your team strike the jackpot on a young, undervalued player or overpay for an aging star?

MLB free agency is particularly interesting, as baseball is the only one of the four major sports without a salary cap. Baseball’s offseason is an open market, with only a relatively small luxury tax for teams with the biggest payrolls.

Here, we take a data driven approach to predicting free agent salaries based on WAR, or Wins Above Replacement. This article discusses the method and looks at how these predictions performed for the 2015-2016 off season.

Finally, we look at the most interesting predictions for the 2016-2017 off season. You can check out all the results on this Google spreadsheet.

Regression model based on WAR

Using regression, we developed a linear and quadratic model for free agent salaries based on a player’s WAR for his three seasons prior to free agency. Regression provides the optimal coefficients for weighting each of these seasons.

The most precise blend of the three WARs for both models weights WAR for the last year very heavily, and the third year almost not at all. While this makes sense conceptually, it can cause our model to miss on some players.

We also attempted to measure the impact of player availability at each position on the market. By dividing individual players WAR by the total WAR available to the market for their position, we were able to gauge their relative strength on the market.

We then multiplied our WAR weighted average by 1 + ((Player WAR) / (Total Position WAR)), a term that gives an extra boost to high WAR players. This reduced the sum of squared error significantly and improved the accuracy of both models as a result.

The figure shows the results for the quadratic (nonlinear) and linear model.


The model is simple, and it doesn’t consider important factors that will affect a free agent contract, such as:

  • a slow-developing market for a position
  • a glaring need by a large market team
  • an impatient owner who wants to win now
  • age, as older players are often unwilling to take short term deals, and teams are unwilling to sign long term ones

However, we’ll see the model’s accuracy in predicting free agent contracts.

The simplicity of our model also contrasts it from the “value metric” of Fangraphs (pitchers and hitters). This method seems to place a lot of value on “market intangibles” or various factors that account for two players with equal productivity being paid differently.

Success and failure from 2015 free agency

After the 2015 season, we experienced great success in predicting some starting pitcher’s contracts. Let’s take a look at a few examples.

  • John Lackey signed a two year, $32,000,000 deal with the Cubs. Model prediction: $16,000,000 per year
  • Hisashi Iwakuma signed a 1 year, $12,000,000 deal with the Mariners. Model prediction: $11,925,600 per year
  • Rich Hill signed a 1 year $6,000,000 deal with Oakland. Model prediction: $6,043,300 per year.

Not only were these all starting pitchers, but they were starting pitchers who were not the best in their free agent class (David Price, Zack Grienke) thus they were not subject to as many market intangibles. These three starters all had an above average season in 2015, but they are not a franchise building block.

On the other hand, one of the largest misses last year was 2B Daniel Murphy, who signed a three year, $37,500,000 contract with the Nationals. Our model predicted him to earn $4,510,000 based off of his performance.

However, above other market intangibles, Daniel Murphy changed his swing during the 2015 playoffs. This change helped him win the NLCS MVP and carry the Mets to the World Series.

Without accounting for his new swing (and therefore increased performance), our model vastly undershot his predicted salary on the open market. These cases seem few and far between, and we do not expect many cases like this in the future.

Predictions for free agency in 2016

Our model does well with two types of players: the late bloomers and the models of consistency. This section will look at examples of each as well as a player we don’t expect the model to predict that accurately.

You can find all the predictions on this Google spreadsheet.

Rich Hill

Rich Hill is the ultimate late bloomer. After bouncing around the majors, Hill found himself in the Red Sox organization as a reclamation project. Looking at his WAR from 2014-2016, it seems like it worked, as he had a WAR of 0.2, 1.6, and then 4.1 the past three seasons.

In a unique case like this, it appears that his salary will be driven by his performance this year more than past years. We believe our model prediction of $16,540,000 is right about what he’ll end up taking home.

Justin Turner

Another example of an ideal player for our model is third baseman Justin Turner, a model of consistency. Turner has been consistently good-to-great for the Dodgers, averaging a WAR of 4.33 since 2014. This past year, he fell right in line with that, being worth 4.9 wins.

With his 2016 performance being indicative of the type of player that he is, we believe his predicted salary of $20,000,000 will be an accurate prediction.

Yoenis Cespedes

After defecting from Cuba and signing with the Oakland Athletics, Cespedes has enjoyed success during his time in the Majors. Looking at his WAR from the past three seasons, he was worth 4.1 wins in 2014, 6.3 in 2015, and 2.9 this past season.

As our model places a heavy emphasis on past year’s performance, his 2.9 WAR is the driving force behind his predicted salary. However, his talent level exceeds his 2016 WAR figure, and he will most likely be paid a higher salary than our model projects.

After two great years of contributing 4+ wins to his team, he will not be valued as heavily on his 2016 performance as the model suggests. With that in mind, we believe the model prediction of $12,390,000 is on the low side for Cespedes.

All stats used in this article and the regression model are from Baseball Reference and contract details are from MLB Trade Rumors.

Podcast: Michigan at Ohio State

thefootballanalyticsshow_cover_landscapeThis week, I focus on Michigan at Ohio State, a critical game in the college football playoff picture. After looking at all the numbers, the game boils down to which team can run the ball, because both teams will find it extremely difficult to throw.

I end with a prediction for Michigan at Ohio State, and then jump into college football upset alert. You might laugh at one of the picks, but hey, I’m here to entertain.

I end with a college football total that just might make one of our picks in the prediction service.

To listen to the podcast on iTunes, click here.

To listen here, click on the play button.