Jason Garrett is getting a lot of crap for being conservative at the end of the game from everyone including the owner of the Cowboys. I don't necessarily agree with the particular play calls, but I think he actually employed the right strategy on that drive before Brady's game winner. Here are my reasonings:
1.) The Cowboys had the lead and had an opportunity to force the Pats to burn their last two time outs and one first down would have at best given the Pats less than one minute and no timeouts to tie or beat the Cowboys.
Couple of considerations:
- Could the Cowboys reasonably expect to just line up and gain 10 yards on three running plays? Obviously not...so expecting a first down that way is not really on the table.
- What would the time situation be if you run 3 times and punt? About 2:30 with one timeout...which is an eternity when you consider all the Pats needed was a FG.
2.) The Cowboys defense has been giving Brady and the offense fits all day. It wasn't like the Pats were scoring at will yesterday. The way the defense was playing, the safe bet was to play conservatively on offense avoiding the big turnover and letting the defense win the game.
The Cowboys were defending 10 yards and 3 downs up to that point. Adding another down alters the dynamic significantly.
As for the Cowboys giving the offense fits, the Pats had 8 drives before that last one. 5 of them ended with the Pats in FG position...which is all they needed to tie the game.
Another thing to realize is that Dallas got a 52 yard net on the punt exchange...which is more than they reasonably should have expected. So the plan should have assumed limited gains on 3 runs, Pats getting the ball back at their 30 with 2:30 left and 1 timeout (plus 2 min warning), only needing 40 yards for a reasonable FG attempt to tie. With those assumptions, time really isn't a factor. If time isn't a factor, playing to drain time and timeouts isn't a good or even sensible plan.
Dallas was going to have to turn over the Pats offense (downs or turnover). Even if Romo had thrown 3 incomplete quick passes, the Pats would have been in the same situation with 3:10 and 3 timeouts (plus 2 min). Is that really that big a difference?
It is all about measuring positive vs. negative outcomes in both situations. If you can trust Romo not to throw a pick, the negatives are pretty much identical. The positives for running the ball 3 times are pretty weak (very small chance of getting a 1st down). The positives for throwing 2 or 3 safe/short passes are significant (likely close to 50% chance of getting a first down based on the defense in front of them).
3.) It is these situations where Romo usually spits the bit.
If Romo is that fragile, then get another QB. There are a lot of QBs that could have put up 16 points on the Pats yesterday with that offense. You pay a QB big bucks to win the game on the final drive. The Cowboys were determined not to let Romo lose that game. But in trying to protect Romo, that is exactly what happened.