The Titans and success rate at midseason

One of my favorite simple measures for getting a perspective on how well the Titans performed on offense and defense in a given game is success rate. I've mentioned success rate before, typically in the context of evaluating running backs, but here's a brief overview.

A play is considered a success if it gains 40% of the needed yardage on first down, 60% on second down, and 100% on third or fourth down, with adjustments for leading in the fourth quarter (lower baseline on first and second downs) or trailing by 8 or more in the fourth quarter (higher baselines on first and second down). A successful play gets a 1, while an unsuccessful play gets a 0. This method has a number of limitations, which I'll get to later in the post, but it's a good initial gloss.

Here's what the success rate has looked like for the Titans and their opponents in each game this season:

Game Titans Opponent
1-vs Patriots 39% 61%
2-at Chargers 36% 51%
3-vs Lions 40% 60%
4-at Texans 43% 57%
5-at Vikings 41% 49%
6-vs Steelers 44% 37%
7-at Bills 48% 59%
8-vs Colts 44% 60%

Now, let's caveat the heck out of those numbers:

1. Success rate is a very unsophisticated, binary metric. It's only 1 or 0. The difference between a 3-yard gain and a 4-yard gain on first-and-10 isn't that big. The difference between an 8-yard loss and an 83-yard touchdown is much bigger. Yet, under success rate, the former play is a 0 and the latter play is a 1 in both cases. This is absolutely true. If I had to choose between success rate and DVOA, I'd use DVOA every time. I can't sit down on Sunday evening and calculate DVOA off the top of my head, though.

2. These are full game numbers, and full game numbers can be very misleading. Take, for instance, the Chargers game. In the first half, the Titans and the Chargers posted nearly equivalent success rates. In the second half, the Chargers annihilated the Titans. The full game total you see above is a blended rate from those two halves.

3. All plays are counted equally, even those when a team is down 20 points in the fourth quarter. Doesn't the Vikings game look better than you thought it would? Through three quarters, the Titans had a success rate of 25%, while for the Vikings it was 48%. Holding a 23-point lead in the fourth quarter, the Vikings played softer defense, and the Titans had a much higher percentage of successful plays. As noted above, the baseline for success rate does change, but I'd say it still overrates the Titans' offensive performance that game.

4. Special teams and turnovers are not included. They're important. I know they're important. An interception is an incompletion. A fumble is an actual offensive play that does (or doesn't) gain yardage. Blocked punts, long kickoff returns, missed field goals, they all matter in actually winning or losing games. You don't find them here.

5. Opponent adjustments. More sophisticated measures like DVOA include them. They matter a lot when looking at other games and how good a team is overall. On gameday, though, you're playing who you're playing. The scoreboard reads 0-0 at the start of the game whether you're playing the Patriots or the Jaguars. On just a per game, "who played better that day" basis, I'm fine with not having opponent adjustments.

6. Results in particular situations matter a lot in real life, but are not included here. In this case, picking up four yards on first-and-10 is good. Getting stuffed on third-and-1 is much worse. They're weighted equally. I came up with a 59% rushing success rate for the Texans in the Week 4 game. DVOA says the Titans did a good job in run defense that game (-16.1%). I agree with DVOA. The Texans had a number of minimally successful gains, are a good run offense that played a bad run defense, and the Titans did a good job of stopping them in third-and-short.

7. Consistent success in particularly situations can skew numbers. Two topics to highlight here. First, I've already noted the Titans success in the fourth quarter against the Vikings. They had a couple good drives that helped cover up some tremendous ineffectiveness in the first three quarters of the game. Second, particularly good quarterbacks tend to be particularly adept in third-and-long situations. The Chargers led 17-0 at halftime despite a poor success rate because Philip Rivers kept converting third downs. Ben Roethlisberger did the same thing, and has all year.

I could go on with the caveats for a while, but by and large DVOA agrees with the basic table you see above: the opposing offense has done better than the Titans offense in every or almost every game the Titans have played this season. The Titans could have .500 right now if they'd seized any of the myriad opportunities they had to beat the Colts, unless they start playing a lot better, they may be a bigger threat to go 0-8 the rest of the way than they are to make the playoffs. And, yes, in saying that I do realize they have two games left against the Jaguars, who are pretty much equally as bad.

I said before the season I expected the Titans to go 7-9, give or take a couple. They're 3-5 right now, which is about where I thought they'd be, but overall they've been worse than I expected. Even with a softer schedule coming up in the second half, my revised expectation is for a 6-10 finish, and I'm still taking the under on that.

Quantcast