It's the playoffs.
Who cares that the Cowboys beat the hapless Washington Redskins 44-17 in the season finale? Who cares that the Lions won 34-17 against the Tampa Bay Buccaneers, proud holders of the No. 1 overall pick in the 2015 draft?
When the 11-5 Lions meet the 12-4 Cowboys on Sunday, they'll be doing something they haven't done much of in the regular season: play against a team with a winning record. The Lions have played five teams with a winning record this season and came up short with a 1-4 record in those games. The Cowboys faced six such opponents and have a 4-2 record to show for it.
There's been a lot of talk about how the Cowboys 5th-ranked scoring offense will meet the Lions 3rd-ranked scoring defense, which is all well and nice, but how much were those stats driven by games against inferior competition like the 4-12 Redskins or the 2-14 Buccaneers?
Here's a little table that shows points per game on offense and defense against winning teams and non-winning teams for both the Cowboys and Lions:
|Non-Winning Opponents||Winning Opponents|
|Points For||Points Against||Points For||Points Against|
The Cowboys have done much better against winning opponents (SEA, ARI, IND, HOU, 2 x PHI) this year. And even if their offensive output has declined by five points per game in those games, the defensive performance has remained largely stable. The Lions drop by nine points per game on offense, and even their otherwise stout defense gives up four more points per game versus quality opponents (ARI, BUF, NE, 2x GB).
But points for and against only part of the picture. What's interesting is that if you add yards to the mix, you get a metric called Yards Per Point (YPP) which gives you a better feel for how easily both teams scored against quality opponents, and how hard they made it for those opponents to score.
YPP (Totals yards divided by points scored) measures how many yards an offense needs to gain to score a point. On the defensive side, YPP measures 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).
Looking just at the stats against winning teams, the Cowboys' offensive YPP is of 13.6, which is a pretty good value and would rank the Cowboys fifth in the league in the regular season ranking. Unfortunately, the Cowboys’ defensive number doesn't look quite as good. Their YPP of 13.8 means that quality opponents can score against the Cowboys with relative ease, and would rank them 27th in the defensive regular season ranking. Compare that to the league-leading Cardinals (19.7) or Chiefs (18.8) and you'll understand that there's a lot of room for improvement for the Cowboys defense.
For the Lions, the numbers look different against quality opponents. They have a very tough time scoring with a YPP on offense of 22.6 (which would rank last), but their YPP on defense is solid at 17.0 (9th).
Some in the stat community get all high and mighty when talk turns to YPP, and like to point out its inadequacies as a metric. Despite that, there are many different ways in which YPP can and is being used, and what we’re going to do today is look at one of the more common uses for YPP: a quick-and-dirty way to predict scores.
The way this can work is by taking the average yards per game allowed by a given team and dividing 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 Lions @ Cowboys game, and we’ll do it using the stats from each team against winning opponents. Against these quality opponents, the Lions have allowed 346.8 yards per game. The Cowboys’ YPP value on offense is 13.6. Divide Detroit's yards allowed per game by Dallas’ YPP and you get a predicted score of 25.6 points.
And here's how much the Lions are projected to score by this method: The Cowboys have given up 310.3 yards per game against quality opponents, while the Lions offense has a YPP of 22.6. Do the math and you get a predicted score of 13.7 points.
So there you have it. The Cowboys beat the Lions 26-14.
If you do the same exercise using all 16 regular season games (because somebody is bound to cry foul at using only the records against winning teams), you'll get a much closer predicted score of 22.9 to 20.9 for the Cowboys.
The Over/Under for the game is set at 48 and the projections above are for 40 and 44 points respectively. Based on this admittedly academic exercise, the Cowboys offense will do its thing. The question is whether the Lions offense can take advantage of the Cowboys' sub-par defense.
And given the state of their defense, the Cowboys may just decide that the best defense is a high-scoring offense - as they've shown when they scored 30-or-more points a league-high 10 times this season en route to a 10-0 record in those games. The Lions have scored 30+ points just three times this season, and have also won all those games. But there's also the matter of the Lions road record: Matt Stafford has gone on the road to face a team with a winning record 17 times and all 17 times he's lost. That's 0-17.
First team to 30 points wins?