/cdn.vox-cdn.com/uploads/chorus_image/image/2747117/136102142.0.jpg)
The Cowboys have had a hard time scoring points this season. While the offense ranks sixth in the league with 382.5 yards per game, they rank 26th in points scored per game. It you divide the offensive yards gained by points scored, you get a metric called Yards Per Point.
YPP measures how easy it is for an offense to score points (how many yards the offense needs to gain to score a point) and how hard a defense makes it for its opponent to score (how many yards the defense gives up for each point scored). Ideally, the number on offense is as low as possible (your team scores with ease) while you want the defensive number to be as high as possible (your opponent has to work really hard to score points).
On offense, the Cowboys rank 30th in the league with a YPP of 20.4. Compare that to the league-leading Bears and Texans with values of 12.1 and 12.5 and you begin to see the enormity of the Cowboys’ offensive futility. Unfortunately, the Cowboys’ defensive number (in large part due to the many turnovers by the offense) is nothing to write home about either: their YPP of 13.8 ranks them 25th in the league.
There are many different ways in which YPP can and is being used, and we could go off on many different tangents with this topic, but what we’re going to do today is look at one of the more common uses for YPP, as a quick-and-dirty way to predict scores.
The way this can work is you take the average yards allowed by a given team and divide those by the opponent’s YPP to arrive at a predicted score. You can do this using the cumulative season statistics; you can do it using the stats from just the last couple of games; just the home games; whatever tickles your fancy – there is no one single approved way to do this.
Today we’ll use YPP to predict the score for Sunday’s Cowboys @ Eagles game, and we’ll do it once using the full season stats and once using the stats for the last four games. For the full season, the Cowboys’ YPP rating on offense is 20.4 and the Eagles have allowed 343.5 yards per game. Divide Philly’s yards allowed by Dallas’ YPP and you get a predicted score of 16.8 points. Will that be enough to win against the Eagles?
Let’s see how much the Eagles are projected to score: The Eagles’ YPP is even worse at 21.2, while the Cowboys have only given up an average of 312.5 yards per game. Do the math and you get a predicted score of 14.7 points.
So there you have it. In a low-scoring affair, the Cowboys beat the Eagles by two points.
And now for the same exercise using only the data from the last four games:
Cowboys predicted score: Eagles Y/G: 387.5 / Cowboys YPP: 18.9 = 20.5 points Eagles predicted score: Cowboys Y/G: 241.5 / Eagles YPP: 17.7 = 13.6 points |
Using the YPP of the last four games played to predict scores is a betting system known as the Dudley Method. By subtracting 1.5 points from the road team and giving the home team an additional 1.5 points, that method gives you a predicted point spread for the game. In our case, that would result in a 19-15 win for the Cowboys.
The two methods described only use the offensive YPP, but another method popular with some bettors is the "Total Dudley Method" which is a calculation of each team's last 4 games using offensive and defensive YPP to predict the score of a game. Usually the bettor is looking for large discrepancies (generally @ five or more points) in the predicted score from the Vegas spread in order to make his bet.
None of this is rocket science of course, and there are many far more complicated algorithms out there that try to predict scores. But the nice thing about this one is that it’s simple and it predicts a Cowboys win on Sunday.
Then again, this is a divisional game, and those tend not to adhere closely to any set of rules and predictions. In fact, if the prediction is for a low-scoring nailbiter, chances are it could turn into an offensive slugfest. In their last 10 regular season games against each other, each team has won five, and the combined score has oscillated between 16 (6-10 Cowboys loss in Week 15, 2007) and 78 points (41-37 Cowboys win Week 2, 2008).
How are you feeling about the Cowboys’ chances on Sunday?