Dak Prescott’s unexpected emergence this year has been one of the biggest stories in the NFL. This week, he landed on the cover of Sports Illustrated.
One of the main features of Prescott’s game is his care of the football. He has already set the rookie record of pass attempts without an interception at 155. He’s only seven passes from tying Tom Brady’s record of passes without an interception to start a career. Prescott also threw 50 passes in the preseason without being picked off. And he set a Mississippi State record of 288 passes without a pick. His only turnover as a Cowboy is a sack-fumble last week against Cincinnati that had no affect on the outcome of the game, but which likely caused Dak to beat himself up over his mistake.
Tony Romo has always been cut from a different cloth. It’s not as if he wants to turn the ball over, but he seemed to come into the NFL back in 2006 without being too concerned about it. His first pass as a starter in 2006 was picked off by the Giants, and quickly led to a touchdown. His reputation was that he was a gunslinger, in the image of his fellow-Wisconsin quarterback Brett Favre, who has thrown 59 more interceptions than any quarterback in history. Bill Parcells gave Romo this label in his book A Football Life.
“If I had put in Romo in his first year and just let him play, he would have been out of football in a year and a half," Parcells said in his book. "He was just a gunslinger. He was indiscriminate. And he would do [expletive] that you just can’t succeed doing. But after a year or two of practicing in the preseason, getting his [reps], you could see he had a real good chance to come along.”
Even as recently as the 2014 season, Tony Romo won “gunslinger of the year”.
Over time, Tony Romo has trended in the right direction when it comes to interceptions, with the notable exception of the shortened 2015 season, when his interception rate went through the roof. But he’s never come close to the streak of passes without a pick that Dak Prescott has already put up.
Do These Interceptions Matter? Yes.
There is a positive correlation between turnover differential and winning percentage. We can conclude that 44% of the variation in a team’s winning percentage can be explained by their turnover differential. This shows that turnovers are not overvalued when it comes to winning percentage.
Here’s a table showing the winning percentages of NFL teams with a positive turnover differential in games from 2008-2012.
|SEASON||1||2||+3 OR MORE||TOTAL||WIN PCT.|
Needless to say, the teams with negative turnover ratios are on the flip side of these numbers.
Quarterback Stats Are Affected By Interceptions
The stats gurus in the NFL understand this correlation and build it into the key statistics we use to evaluate quarterbacks. Thus, quarterback stats are dramatically affected by interceptions. Take the standard passer rating formula. It’s a calculation based on completion percentage, yards per attempt, touchdowns per attempt, and interceptions per attempt. The perfect score is 158.3. Dak Prescott had one of those games in pre-season this year, in the second game against Miami. He was 12 of 15, for 199 yards, two TDs, and no interceptions. If Dak had thrown one interception, his rating would have fallen to 130.56, still excellent. Two picks would have dropped it to 102.78, just above where he is now. Three picks - 75. With passer rating differential being a key variable in who wins football games, this just confirms that interceptions matter.
Adjusted net yards per attempt is another stat often used to rate the performance of quarterbacks. Yards per attempt is a very simple division of yards over attempts. The number is adjusted in this stat by four additional factors - touchdowns (each worth 20 additional yards), interceptions (each subtracts 45 yards), sacks, and yards lost by sacks. The quarterback who wins this race against the opposing quarterback also usually wins the game. The key point? An interception reduces yardage by more than the bonus yardage awarded for two touchdown passes. That’s how much interceptions hurt.
Interceptions (or the lack of them) are the reason that Dak Prescott is running ahead of Tony Romo’s stats from 2014.
Here are Dak’s stats.
Here are Tony’s stats.
As you can see, Dak is winning at passer rating and ANY/A after five games, despite having thrown five fewer touchdowns. If you took away even one of Romo’s interceptions, he would be tied with Dak in passer rating. If you took away all of them, he would be well ahead.
Some Tidbits From This Season
The Cowboys have only had three turnovers this year. Ezekiel Elliott’s fumble against Washington, Terrance Williams’ fumble against the Bears, and Dak Prescott’s sack-fumble against the Bengals. All led to points by Dallas’s opponent: two touchdowns and a field goal. Two of the three likely also took points off the board for Dallas. Fortunately, none of them cost the Cowboys the game.
My last tidbit is that Carson Wentz was neck and neck with Dak Prescott on the rookie “interceptionless” record race until he finally threw one this last week against Detroit. Guess what? It cost Philadelphia the game.
But let’s get back to Tony Romo and Dak Prescott. How have interceptions affected Tony Romo’s record with the Cowboys?
Tony Romo’s Interception History
To get at this, I decided to look at all of Tony Romo’s starts from 2006 to 2015, and see what Dallas’ won-loss record was in the games where Tony threw at least one interception versus the ones where he didn’t. No attempt was made to count the number of interceptions in each game, or calculate if Dallas won the turnover battle for that game even when Tony was picked off (which may explain some of why the Cowboys have a winning record in games where Tony was intercepted). I just wanted to see how often Tony Romo gets picked off, and how it generally correlates to wins and losses. Here’s the table.
|Year||Int - Win||Int - Loss||No Int - Win||No Int - Loss|
What can we make of this data? First, in games where Romo has not thrown an interception, he led the Cowboys to victory 69.8% of the time. That’s not Tom Brady or Russell Wilson turf, but it’s a better winning percentage than any other quarterback since 1981. In contrast, Tony has been able to get the Cowboys to victory 56.1% of the time when he throws a pick - perhaps one of the reasons he doesn’t seem obsessed to prevent them. But this is Alex Smith’s winning percentage.
Second, Tony Romo has thrown an interception in 57.9% of the regular season games he’s played in over his career. He’s done better in the playoffs, getting picked off twice in six games. At the beginning of Romo’s career, he seemed particularly careless. He was picked off in eight of eleven games in 2006, ten of thirteen games in 2007, and nine of twelve in 2008. After 2008, Tony cut his picks down to roughly half of the games and has remained fairly steady at that pace. So, if Tony Romo returns to the field, we can expect he will get picked off in about half of the games, and the team will have to fight to overcome those setbacks. How many of those games will the Cowboys lose? If history holds true, about 44% of them.
This is not intended as a scientific analysis, but it’s nevertheless an insightful one.
In comparing these two quarterbacks, many have talked about how Tony Romo throws a better deep ball, or reads the defense more skillfully, or can thread the needle with more precision. That may be true.
However, is that what wins and loses football games? Aren’t turnovers, or the lack of them, a much bigger key to the outcome of games?
Tony Romo may be a better passer than Dak Prescott, but if he comes back and turns the ball over in more games than Prescott is likely to, is he the better quarterback for this Dallas Cowboys team?