That didn't go quite as planned. The Seahawks offense/Eagles Defense matchup went as projected, but the reverse did not. The Eagles struggled to produce points and obviously dropped the game. That gives us a loss on the line, but a win on the O/U. Mistake was likely due to underrating the Seahawks' defense rather than overrating the Eagles' offense. Fortunately, now we're back into divisional games, which means we should have a LOT of information to work with.
This week's game:
Huge for the Eagles. A win essentially clinched the division title and keeps the team in the race for a playoff bye. A loss isn't catastrophic, but the Eagles would then need the Cowboys to lose to the Colts next week to be able to reclaim the top spot in the division. Of course, the most interesting aspect to this game is that these teams played just two weeks ago. As everyone here knows, the Eagles won the game cleanly, by 23 points on the road.
This week, the line favors the Eagles by just 3.5 points. Thus, we're left with an interesting exercise in new information integration. First, let's review the Thanksgiving game breakdown:
The Cowboys were favored by 3.
However, our breakdown suggested a final score of 29-31 points for each team, so I had it as a toss-up. The Eagles scored 31, hitting the range nicely. The Cowboys, though, scored just 10 points. Tony Romo struggled mightily, and the big question for this week is whether that performance was due to his back and whether it will reappear this Sunday.
In any case, though, we're left with the following building blocks:
- We thought the Eagles-Cowboys in Dallas should be a toss-up.
- The Eagles then beat the Cowboys by 23.
- Since then, the Eagles lost to the Seahawks at home by 10. The Cowboys won in Chicago by 13.
Before I get to the breakdown, I'm going to say that with these pieces alone, we're looking at Eagles -6 or so. We'll check that against the breakdown, but the fact is two weeks ago we though these teams were close to even (Eagles slightly better). Since then, we've moved even farther towards the Eagles. Note that the Cowboys actually beat the Seahawks in Seattle, so that result favors them by a fair amount. However, it's outweighed by the head-to-head result. Lastly, a win against Chicago doesn't mean very much at this point, no matter where it takes place. Chicago ranks 25th by DVOA, and at this point in the season we pretty much know both Dallas and Philly are better than Chicago, so the informational value of a win is quite low.
Eagles Overall DVOA: 13.7% (7th)
Cowboys Overall: 3.6% (13th)
Eagles Offense: -2.6% (16th)
Cowboys Defense: 10.2% (28th)
Eagles Defense: -7.2% (7th)
Cowboys Offense: 13.3% (6th)
The Eagles are #1 in STs, the Cowboys are 13th, with a DVOA just above zero (0.6%).
The DVOA breakdown has moved in the Eagles favor since Thanksgiving, and it now suggests a fairly significant gap in team strength. The biggest highlight above is the Dallas defense: They're not good...at all. The Eagles looked terrible against Seattle, but the Seahawks rank 4th by DVOA (and improving fast). The Cowboys are on the other end of the spectrum, and the defense is the only unit from either team that's actually BAD.
Opponents in Common
This is very similar to the section from two weeks ago. We just have to add the head-to-head and the Seahawks.
Jacksonville - Eagles won at home by 17. Cowboys won on the road by 14. (Tie)
Washington - Eagles won at home by 3. Cowboys lost at home by 3. (Advantage Eagles)
49ers - Eagles lost on the road by 5. Cowboys lost at home by 11. (Eagles)
Rams - Eagles won at home by 6. Cowboys won on the road by 3. (Tie)
Giants - Eagles won at home by 27. Cowboys won at home by 10. (Eagles). The Cowboys also beat the Giants on the road by 3.
Arizona - Eagles lost on the road by 4. Cowboys lost at home by 11. (Eagles).
Houston - Eagles won on the road by 10. Cowboys won at home by 3 in OT. (Eagles).
Titans - Eagles won at home by 19. Cowboys won on the road by 16. (Tie).
Seahawks - Eagles lost by 10 at home. Cowboys won on the road by 7. (Cowboys).
Head-to-Head - Eagles won by 23 on the road. (Eagles).
As you can see, the opponents in common points strongly towards the Eagles. The Cowboys have just one advantageous result, the Seahawks game. While that's the most recent result for the Eagles, the Cowboys comp took place back in October (Week 6).
The Cowboys rank 28th by DVOA on defense. The Eagles, of course, played them and scored 33 points on the road. The Eagles also played the Titans, who rank just below the Cowboys on defense. Against Tennessee, the Eagles scored 43 points (at home).
The Eagles offense now ranks 16th by DVOA, just ahead of Houston and a few spots beneath Chicago. Against those teams, the Cowboys allowed 20 (Houston at home) and 28 (Chicago on the road).
For the season, the Eagles are averaging 29.9 ppg. The Cowboys are allowing 23.2 ppg, a rate that doesn't look terrible but is skewed by the easy schedule Dallas has played (though you can say the same about the Eagles). I'm going to try not to overthink this one too much. The Cowboys defense is bad, the Eagles are at home, and we have a very recent direct comp. All together, I'm setting the Eagles projection at 33-35 points.
On the other side, the Cowboys' offense ranks 6th by DVOA. That's just one spot behind Seattle. Against these teams, the Eagles allowed 10 points (Dallas on the road), and 24 (Seattle at home).
The Eagles rank 7th by DVOA, behind Arizona, SF, and Seattle but a little ahead of St. Louis. Against those teams, the Cowboys scored 17 (ARI home), 17 (SEA home), and 30 (SEA road). Against St. Louis, the Cowboys scored 34 on the road.
We see a pretty big home/road split for the Cowboys, so we'll need to adjust our expectations upwards a bit. All told, those games point to a scoring range in the mid-high 20s. For the season, Dallas is averaging 26.4 ppg. The Eagles are allowing 23.8 ppg. Given the matchup, we should expect both teams to perform worse than average (Eagles have a good defense, Cowboys a good offense). That gives us a projection of 24-26 for the Cowboys. With the big home/road split, I'm inclined to raise the Cowboys high end a couple of points. It's tough to quantify this tip of thing accurately (is it real or coincidence?), but we should at least add a buffer just in case. That gives us a Cowboys range of 24-28 points.
Together, we get a projection of Eagles 33-35, Cowboys 24-28. Using the midpoints, we've got a spread of Eagles -8. That seems aggressive. Note, though, that the spread is just 3.5, so we have a fairly large buffer. Moreover, I didn't shift the projection at all to account for the Eagles big advantage on Special Teams. Checking that against our back-of-the-envelope spread up top, we've actually come out more bullish on the Eagles (thought Eagles -6 initially). That's a potential flag for confirmation bias, but that's why I use things like DVOA and comps. On the flip side, it's a good sign that the numbers back up what we thought initially.
In total, take the Eagles -3.5. We haven't seen a difference this big in a while, but I think there is a lot of opportunity here. The real question is: Why is our spread so much different? First, recency bias. The Eagles are coming off a loss, the Cowboys a win, so there might be some recency at work here. Two, narrative. This is a big one. The Eagles JUST BEAT THE COWBOYS cleanly, in Dallas. However, the reaction to that has been to point to the home/road splits for the Cowboys, talk about the short week hurting Dallas' prep (nevermind that the Eagles had the same short week), and to talk about Romo not taking a shot before the game. Those are all possible explanations for why we shouldn't put too much stock into the head-to-head result.
We need to think probabilistically, though, and in my opinion, odds are those explanations are bullshit. Maybe I'm wrong (always a significant chance of that), but it looks like the narrative is obscuring a pretty clear Eagles advantage here.
For the O/U, the line is 55. Or projection midpoints get us to an O/U of 60, so take the over again. However, keep an eye on the conditions. It's a night game, so the temperature will likely be close to freezing at game time. Right now there's a 10% chance of precipitation. Wait to play here until we're closer. If there's rain, stay off. If it's snow, go for it.
Lastly, for what it's worth, some other sources:
FiveThirtyEight's Elo system has the Eagles -4 (64% chance to win).
ESPN's PickCenter says take Dallas. (Eagles by 1-3 points).