Lovie Smith was right to coach conservatively against the Saints

The Tampa Bay Buccaneers beat the New Orleans Saints in a nailbiter on Sunday. It was their first win since November of last year and it was deservedly celebrated, but the win was unreasonably close given the team's dominance through three quarters: they led 23-7 at the end of the third quarter, but Drew Brees came awfully close to tying or even winning the game in the final seconds.

That didn't happen -- the Bucs left with a 26-19 win -- but they made it much more difficult on themselves than needed. Obviously, two unnecessary fumbles contributed -- but the team's conservative playcalling has been blamed as well. In essence, the Bucs started milking the clock at the start of the fourth quarter when they could have been trying to at least add a few more points to the board -- and if it wasn't for a few lucky bounces, they could have easily lost this game.

In the fourth quarter, the Bucs had 11 first- and second-down place, and ran the ball on all but two of them. Those runs into nine-man fronts resulted in a predictably low total of nine yards, no first downs and a lost fumble. That's an abysmal result, and the only positive to come out of those plays was that they kept the clock moving. In the mean time, they gave the Saints a whopping four opportunities to score -- that New Orleans only managed to come out of that with six points was as much a result of their own ineptitude as it was the consequences of the Bucs' defensive performance.

To make matters worse, the Bucs didn't even try to let their offense win the game when they had a third-down immediately after the two-minute warning: Doug Martin was given the ball on third-and-11, the Bucs preferring to take four seconds and a timeout off the clock, rather than allowing Jameis Winston to do what he was drafted to do.

I thought Lovie Smith had made a clear mistake in playing so conservatively, but after checking the data, my gut reaction was probably wrong.

Lovie Smith is very conservative at the end of games

Lovie Smith has a history of throwing away games with conservative playcalling -- which is odd, because Lovie Smith really isn't all that conservative when it comes to fourth down decisions.

Last year, Lovie Smith ran the ball on 12 of 17 fourth-quarter first- and second down plays when the Bucs were leading by two scores or less, excluding kneels. In his career with the Chicago Bears, he did so on 367 of 505 plays, or 72.7% of the time. Over that same time period, the NFL as a whole ran the ball 70.2% of the time in those situations. And that percentage has been declining steadily: since 2013, it's dropped to just 67.9%.

Compare Lovie Smith to the most successful coaches out there, and it looks even worse. Bill Belichick has run the ball 60.6% of the time in that situation since 2010, Mike McCarthy 62.7% of the time over that period, Tom Coughlin 69.3% of the time, John Harbaugh 71.8% of the time, Sean Payton (and temporary replacements Aaron Kromer and Joe Vitt) 56.1% of the time, Pete Carroll 70.7% of the time (2011-now) Bruce Arians 71.7% of the time (2013-now) and Jim Harbaugh 74.8% of the time (2011-14). Only Jim Harbaugh is more conservative than Lovie Smith's record, and he had dominant defenses to fall back on as well as a quarterback with a tendency to run successfully -- neither of which the Bucs have right now.

Simply put: Lovie Smith is far more conservative than both the average and the most successful coaches in the NFL.

But evidence says being conservative is not a bad thing

Does this really correlate with losses? A quick survey says not really: teams rushed the ball on first or second down while leading by a single score in the final five minutes of the game over the past three years (excluding kneels) in 215 games. They won 183 of those games, or 85% of the time. That's pretty good -- and teams won 94 of the 114 times that they passed the ball in the same situation or 82% of the time. That's probably too small of a sample to conclude that passes are worse than runs, but it is enough to say that running probably isn't worse than passing there -- taking time off the clock in the final stages of the game is definitely valuable in and of itself. If we look at total plays instead of games, we come to similar numbers: 88% wins when rushing, 82% when passing.

Similarly, the decision to run Doug Martin into a pile on third down may not have been wrong. Teams won 91 of 118 games where they gave their quarterback the opportunity to win the ball game on third-and-three or more in the final five minutes while holding a one-score lead, or 77%of the time. Compared to the 88% of the time teams won when they ran the ball in that situation (89 of 102 games). Oncee again, there's no real evidence that running is better than passing, and there's even a small bit of evidence that the reverse may be true. In this case, the old-school, conservative playcalling may actually have been right.

There's one argument left: that the Bucs were right to be so conservative over the last few minutes, but that they should have been more aggressive early in the fourth quarter. But even there the evidence is mixed. Certainly, they ran the ball on three of their four plays prior to regaining the ball at 7:15, when they milked four minutes off the clock. But they were only four plays -- and the one time they dropped back to pass, Winston made the rookie mistake of scrambling into defenders and losing the ball.

And looking at the NFL history of wins when running the ball in the fourth quarter, the evidence in favor of running the ball is fairly strong, too. On 1478 first- or second down runs while leading by two scores or less with more than five minutes remaining in the game, teams won 1678 times or 88% of the time (356 of 429 games or 83%). When passing, they won 912 of 1087 plays or 84%  (317 of 388 games or 82%). Once again, it seems that rushing is correlated with winning more than passing is.

Now it should be said that none of this is definitive proof of anything. There are several biases here, including play selection (teams with weak defenses are probably more likely to try to win the game on offense than eat clock) and these are perhaps not the very best measures to be answering this question. But they do give us a clear indication that the general distaste for running out the clock isn't really supported by data.

Lovie Smith probably did the right thing

This data tells us that in general, there's nothing wrong with playing conservatively when leading at the end of a game -- but that doesn't mean it's the right call in every specific situation. In football, statistics can give us general guidelines but can't tell us exactly what to do -- every situation is unique and individual variables can alter the correct path to take.

That said, while I didn't like the conservative playcalling -- perhaps as a result of seeing the Bucs give away game after game doing so -- upon reflection, there's actually very little reason to believe it was the wrong way to go about winning this game. The fact that the Bucs' defense had been playing well, that they had a rookie quarterback known for taking a few too many risks, that that rookie quarterback had fumbled the ball thanks to a decision to run unnecessarily -- all of those factors only point in the direction of being more conservative, not less so.

In other words: Lovie Smith's conservative playcalling probably didn't cost the Bucs the game.