You Don’t Know What Tony Romo’s Got Till He’s Gone

UPDATE (Nov. 26, 10:30 p.m.): Tony Romo reinjured his collarbone on Thursday in a game against the Carolina Panthers. ESPN reported that he will be out for the rest of the season.

By beating the Miami Dolphins Sunday, the Dallas Cowboys preserved their undefeated record this season with quarterback Tony Romo under center. That’s the good news for Cowboys fans; the bad news is that Romo has only played three games. Matt Cassel and Brandon Weeden were both winless when filling in at QB while Romo was out with a broken clavicle. At 3-7, the Cowboys have little chance of making the playoffs unless they go on a serious tear against what isn’t an easy schedule down the stretch.

If nothing else, the Cowboys’ season has reinforced just how completely essential Romo is. His absence has made it clear that he belongs in any discussion of the league’s top signal-callers.

Romo ranks among the top QBs in the NFL over his career, whether you use data from the box score¹ or more advanced, play-by-play metrics.² But Romo also has a reputation, starting early in his career and growing with each passing year since the Cowboys’ last Super Bowl win, for not being “a winner.”

While it’s possible Romo is a stat-stuffer whose great individual numbers overstate his true worth, there are other ways to detect a player’s value. For instance, here at FiveThirtyEight we’re oddly thankful when players miss time because it gives us a chance to quantify their influence on the performance of their teams. And few quarterbacks in NFL history have been associated with a better with-or-without-you (WOWY) ledger than Romo.

To help measure this WOWY effect, I used Elo ratings, FiveThirtyEight’s pet power rating system. Elo gives us a snapshot of how good each team was expected to be at any given moment, as well as how much a team’s strength estimate changes over time. By looking at those changes in conjunction with who the team’s primary QB was during that game³, we can construct WOWY scores that measure how much having a given passer under center helps the team.

For instance, after Romo’s Cowboys beat Philadelphia in Week 2 (the game in which Romo was lost to injury), they achieved a 1618 Elo rating. Seven games later, when he took the field again on Sunday, Dallas’s rating had dropped to 1463, meaning it lost about 22 points of Elo per game in his absence. By comparison, the Cowboys gained about 19 points of rating per game when Romo does play, so his WOWY this season is +41: the team gains 41 more points of Elo per game when he plays than when he doesn’t.

Gaining 41 points of WOWY is incredibly impressive by itself. In seasons where a QB both played and missed at least three games, only Dan Fouts’ 1978 season ranks higher since 1970⁴. But it’s nothing new for Romo, who has consistently been good for the Cowboys. Over his career, his WOWY score of +16.7 ranks third among qualified⁵ passers since 1970:

That Rex Grossman and Joey Harrington are anywhere near Tom Brady and Peyton Manning on that table shows that this isn’t a completely foolproof method. (For one thing, WOWY can’t factor in the quality of a QB’s backups, beyond what Elo is able to detect from changes in a team’s performance.) But it does suggest that, in terms of making his team more likely to win games, Romo has done an impressive job with the supporting cast the Cowboys have given him. Like Brady⁶ and Manning, Romo is associated with elevating his teams’ chances of winning games in addition to gaudy stats.

That probably won’t be enough to change anyone’s mind in one of those endless arguments about Romo, but it does quantify what we’ve seen so far this season: The Cowboys are a very good team with Romo — and are totally lost without him.

Check out our NFL predictions for odds on every game.