You likely heard the same thing in every NFL broadcast you watched during Week 2. There was a coordinated effort, it seemed, to add drama and intrigue to a team’s second game by mentioning one simple thing on repeat: teams that start 0-2 are doomed.
Depending on how far you decide to go back, only about 11 to 12 percent of teams that start a season 0-2 came back to make the playoffs. At first glance, you can discern that this is because of the hole it has dug itself. You now probably have to go 9-5 at worst to put the odds back in your favor, and that’s hard to do in a parity-heavy league.
There’s a simpler reason why most 0-2 teams miss the playoffs: a team that starts 0-2 is quite possibly bad.
Not every 0-2 team is created equally, however. Let’s judge the NFL’s seven 0-2 teams based on a few key advanced statistical categories.
So what are we looking at here?
When writing about college football, I have for years referred to what I call the sport’s Five Factors, the five statistical categories most directly tied to wins and losses. They are efficiency, explosiveness, finishing drives, field position, and turnovers. While some of the expectations and definitions may change as you go from college to pro, the same things win ballgames.
There are a few different ways to measure these things, obviously, but for the purposes above, I stuck mostly with plain old margins — yours minus your opponent’s. There’s one primary exception.
Here are the stats listed above. I’ll define the first one last.
- Yards per play margin: This is probably the best measure you can cull from a basic box score — nothing “advanced” about it. If you out-gain your opponent on a per-play basis, they usually have to do something a lot more creative or unsustainable to beat you. Houston has out-gained its first two opponents (New England and Tennessee, both playoff teams in 2017) by a healthy margin.
- Marginal efficiency margin: I wrote about this in the NFL season preview. Marginal efficiency takes the success rate concept that is a massive part of everything I do and adjusts it for down, distance, and field position. If your marginal efficiency is positive, that means your success rate is higher than those factors would suggest it should be. And if you are doing better than your opponent in this regard (as Houston and Oakland are), that means you are staying in more favorable downs and distances and giving yourselves more reliable chances to score. You are also succeeding in the most sustainable aspect of the game.
- Marginal explosiveness margin: Same concept. I take my go-to explosiveness rating (in this case, IsoPPP, a measure of the magnitude of your successful plays, which you can learn more about by searching for it here) and adjust for down, distance, and field position. Big plays are random and volatile, but it’s pretty hard to lose if you’re making bigger big plays than your opponent. Seattle is doing that by quite a bit.
- Scoring opps margin: For college football, I define a scoring opportunity as a first down inside your opponent’s 40-yard line; for NFL data, I have added any drive that results in a field goal opportunity (simply to account for the fact that NFL kickers are better ... most weeks, anyway). At this point, the field begins to shrink on you, and your play-calling begins to shift. Houston has created more opps than its opponent, and thanks to those (unsustainable) big plays, Seattle’s created as many.
- Points per opportunity margin: Great, you’ve created a scoring chance — how well do you finish it? In games with such evenly-matched talent, a category like this takes on even more importance compared to the college ranks. This is a much more telling measure than something like red zone scoring percentage.
- Field position margin: This one’s easy: where are you starting drives compared to your opponents? Good field position margin is created by both good special teams, a happy turnover margin, and good efficiency — if you move the chains once or twice before punting, you’re flipping the field — and Houston has done a particularly good job of leveraging the field so far.
- Turnover margin: In the professional environment, where talent differentials are far smaller on average than they are in college football, turnovers can make an even bigger difference. It’s not a surprise that no 0-2 team currently has a positive turnover margin and that only two have broken even.
- Turnovers luck: There are turnovers, and there are expected turnovers. You have a little bit of control over fumbles or when your passes hit opponents’ hands. But when a fumble hits the ground, randomness more or less takes over — over time, your fumble recovery rate will trend toward 50 percent. And when a defender gets his hand on a pass, there’s a randomness element, too. Over time, 22 percent of a team’s passes defensed (interceptions and breakups) will be INTs. In the short-term, however, those numbers can skew drastically. Just ask Oakland.
Based on those overall averages, the Raiders’ turnover margin should be about plus-1.3 right now. It’s minus-2. Being that a one-turnover shift in turnover margin has been worth about four to five points so far this year, you could say that the god of randomness has cost the Raiders six to eight points per game so far this year. And every 0-2 team but Seattle (the exception to every rule, apparently) has been unlucky in this regard.
- Second-order wins: At the college level, I take all of a game’s key stats (the ones above, plus a few other stragglers), toss them into the air, and basically say “based on these key stats, you could have expected to win this game X percent of the time.” This postgame win expectancy gives us a pretty good impression of whether what should have happened — and what would happen over the course of, say, a seven-game series — actually did happen.
Basically, treat, say, a 60 percent postgame win expectancy as 0.6 second-order wins and keep a running total throughout a given season, and you’ll get a pretty good impression for what a team’s record should be. (At the college level, there are certain coaches who consistently over- or underachieve in this regard, but they’re rare. And the overachievers often run the triple option.)
From the perspective of second-order wins, two teams have stood out when it comes to comparing what should have happened to what did: Houston and Oakland.
Houston is basically doing everything right except finishing drives. In the Texans’ 20-17 loss to Tennessee, they won nearly every category. They enjoyed a 39 percent success rate to the Titans’ 32. Their successful plays were much bigger. They created two more scoring opportunities. Their field position margin was plus 2.9 yards per drive. Expected turnover margin was nearly 0. Postgame win expectancy*: 93 percent.
One of those Houston scoring opportunities, however, ended when Adoree’ Jackson intercepted a bomb attempt to DeAndre Hopkins. Throw in a pretty damn amazing fake punt touchdown (why were you letting the gunner run free???), and that was enough to flip the win. In a seven-game series with these stats, the Texans win in five.
Add that 0.93 second-order wins to 0.24 from a reasonably nip-and-tuck loss to the Patriots in week one, and on paper, you’re looking at a team that should have about 1.2 wins instead of zero — there was a 22 percent chance that these stats would have produced a 2-0 record and only a five percent chance it would have produced 0-2. Finish drives a little better, and the Texans still have a healthy shot at the playoffs.
* This figure is based on college averages, not pro. When I’ve got enough pro data to work with in this regard, I’ll create a new calculation, but it likely won’t change much — pro and college stats are much more similar than you might anticipate.
If you’ve watched either or both of Jon Gruden’s first two games back as Oakland’s head coach, you know this is a pretty good team. They had the indomitable Rams tied at 13-13 until the last play of the third quarter and only lost a grip on the game when quarterback Derek Carr briefly forgot how to throw a football. In week two, they had Denver dead to rights, leading 12-0 with 25 minutes remaining, having held the Broncos to two first-half first downs.
The Rams did pull away late, and the Broncos won 20-19 on a Brandon McManus field goal with six seconds left. But statistically, both games were tossups. Oakland enjoyed a better success rate in both contests enjoyed a 38 percent postgame win probability against the Rams and 58 percent against the Broncos. There were basically equal odds of a 2-0 record (22 percent) and 0-2 (26 percent). Oakland’s 0-2.
Neither Houston nor Oakland are great teams, obviously. In Football Outsiders’ 2018-only DVOA rankings, the Texans are 11th, the Raiders 20th. But they’re easily a few steps ahead of the rest of the 0-2 pack. I’d be willing to wager their odds of making the playoffs are far better than 11 or 12 percent.