Yes, Jason Garrett deserves our "Thank You for Not Coaching" recognition for an awful decision he was involved in during the fourth quarter of Dallas's crushing loss to Arizona. The only problem is that he's getting criticized for the wrong mistake.
At the moment, Garrett is being pilloried for "icing" his own kicker, rookie Dan Bailey, as he attempted a 49-yard field goal to break a 13-13 tie with six seconds left. Bailey's initial kick went through the uprights, but after the icing, his second field goal attempt went short and wide of the uprights. The Cardinals promptly won the game in overtime without the Cowboys ever receiving possession of the football.
With a stopped clock, it's hard to figure out what Garrett saw to justify a timeout. After the game, he said that his team was "still settling in" and that there were only six seconds left to go on the play clock, but the blockers didn't look particularly unsettled when Bailey attempted the first kick. Furthermore, Cardinals head coach Ken Whisenhunt said afterwards that he was about to call a timeout himself, noting, "I was glad they iced their kicker so I didn't have to."
Even if we think that Garrett was stupid to call the timeout, the question here then becomes whether icing the kicker really has any effect, regardless of which coach does the icing. It's easy to point to a situation where the kicker hits the first kick and misses the "iced" one and suggest that icing is a tangible effect, sure, but there are also situations where an iced kicker gets a second chance to win the game. In fact, this exact situation came up the very last time the Cowboys and Cardinals went to overtime in Arizona! With a 24-21 lead at the end of regulation in Week 6 of the 2008 season, Whisenhunt called timeout right as Cowboys kicker Nick Folk missed a 52-yard field goal. Given a second chance, the "iced" Folk promptly booted the field goal through, moving the game into overtime.1
By all accounts, the evidence in favor of icing the kicker is scant at best. Kickers like Adam Vinatieri have suggested that they actually appreciate getting a practice kick in a game situation. And even if you chalk that attitude up to the famed garrulous swagger of kickers, the statistical evidence suggests that icing really has little to no effect on kickers. A study conducted by Tobias Moskowitz in the recently published book Scorecasting compared the success rates of iced kicks and non-iced kicks in key fourth-quarter situations after adjusting for the distance of the field goal. With nine years of field goals to study, Moskowitz found that when an opposing coach iced the kicker with 15 seconds or fewer left to go in the game, those kickers actually got more accurate:
Despite the evidence, coaches will continue to ice the opposing kicker at the end of games because it's a play with no downside for them. If a team ices a kicker and that kicker hits the field goal, no one ever comes out to suggest that the team was wrong to ice the kicker. Heck, even if the kicker misses the first one and then hits the second kick, like Folk did against Whisenhunt, the coach rarely gets blamed. If the opposing coach ices the kicker and that kicker misses the ensuing field goal, though, he's lauded as some sort of coaching savant. It's nonsense. Garrett's timeout was unnecessary, but it didn't cause Bailey to miss the field goal. A bad kick caused Bailey to miss the field goal.
So if Garrett deserves a pass for his timeout, why should we be excoriating him? Well, because of what Garrett did before the timeout. On Dallas's final offensive play, Tony Romo hit Dez Bryant over the middle for a 15-yard gain to pick up an essential first down on third-and-11. When Bryant hit the ground, the Cowboys still had 23 seconds left on the clock and two timeouts to work with. The ball was on the Arizona 31-yard line, which is within makeable range, but far from a chip shot; the average kicker will boot that through less than 65 percent of the time.
The clear play for Garrett (or Romo) is to call a timeout immediately after the first down conversion. No one is suggesting that the Cowboys needed to try to score a touchdown, but with two timeouts left and a difficult field goal ahead of them, the Cowboys clearly would have benefited from getting at least a couple of yards with one or two extra plays. Even picking up five yards and turning the field goal into a 44-yarder would have increased Bailey's chances of making the kick, based on how kickers have performed over the past several seasons, from under 65 percent to greater than 77 percent.
Garrett claimed after the game that he was inside the " yard lines that we use as guidelines before the game," which doesn't pass the smell test; just because you're inside a range doesn't mean that Bailey's chances of success are equal from any spot within that range. Garrett also added, "You see so many situations where you have negative plays in those situations." Sure, negative plays can happen if you try to move the ball forward, but a missed field goal is a pretty negative play, too. Did Garrett really have such little faith in his offense that he couldn't trust them to get a few yards without screwing up? If he puts the ball in Romo's hands and rolls the quarterback out, Romo has the option of throwing the ball away or running for a few yards and then using Dallas's final timeout to set up an easier field goal. A conservative fullback belly handoff would have probably netted a yard or two with little chance of making the field goal significantly tougher. Garrett's quote speaks to the risk involved in giving your offense a chance to improve your situation, but it ignores the risk associated with putting your kicker in a difficult spot.
Why don't teams go for a two-point conversion when they score a touchdown and go up seven points (before the extra point) in the fourth quarter? There were two teams on Sunday that each had a chance to make that decision and chose against it, and in each case, it's hard to understand why.
First, the logic. Of course, going for a two-point conversion in every situation doesn't really make sense unless you're a team like Carolina or Denver that is capable of being dominant in short yardage. In a vacuum, this play only really makes sense if you think your chances of picking up a two-point conversion are greater than the other team's chances of picking up a two-pointer later on. It's probably not worth the aggravation.
In a situation where time is limited, though, it becomes an extremely valuable gambit. Let's think about this in the prism of that amazing Giants-Packers game from Sunday afternoon. When the Packers scored on a Donald Driver catch with 3:34 left, they went up 34-27 before the extra point. Kicking would give them a 35-27 lead, forcing the Giants to drive down the field and pick up a two-point conversion just to tie the game. Tough, but as the Giants ended up exhibiting on that very drive, not exactly impossible.
Now, what if the Packers had gone for two in that situation? If they don't make the two-point conversion, they're still up by a touchdown with less than four minutes to go. The Giants still have to drive the length of the field, just like they do in an eight-point game, but they only have to kick an extra point to tie it up as opposed to going for two. That's easier, but the long drive is still more difficult than the two-point conversion, which an average team will pick up 47.9 percent of the time. Furthermore, while the other team could theoretically go for two and win the game with a two-point conversion, they never will; teams in a one-point game will kick here virtually every time and tie it up, even if they should go for two. So whether the Packers kick an extra point or go for a two-point conversion, the worst result for them is almost surely that the game will be tied.
The upside, though, is massive. A two-point conversion puts the Packers up nine with 3:30 left. That's a two-possession game, and it simply leaves the Giants with too much to do. First, they have to score a touchdown. They'll obviously tack on an extra point down three with virtually no time left, so it'll be a two-point game with whatever time is left after the touchdown drive. Now, the Giants have to either recover an expected onside kick (success rate: 21.1 percent) or stop the Packers from picking up a first down in order to get the ball back. And they're still not done! If they get the ball again, they have to drive down the field and set up for a game-winning field goal, and that field goal has to go through the uprights. All of this has to happen in 3:30 with one timeout and the two-minute warning for the Giants. The Packers' chances of winning with a successful two-pointer have to be in the high nineties, and their chances of getting a two-point conversion with Aaron freaking Rodgers on the move are surely greater than 48 percent.
Let's go back and envision that final 3:30 in a scenario where the Packers pick up a two-point conversion and go up nine. The Giants have to march downfield and score, which took them 2:36. Even if we assume that they'll go into a no-huddle and drive faster, they're unlikely to score before the two-minute warning. After that play, the Giants have to attempt an onside kick versus a team of freakish athletes at wide receiver. And if the Giants do somehow score before the two-minute warning, all Aaron Rodgers needs to do is get one first down — at most, two — to end the game. Instead, when the Giants scored eight points and tied up the game, Rodgers had to march the Packers 68 yards downfield to set up a game-winning field goal attempt.
Again, the only way the Packers increase their chances of losing by going for two in this scenario is if the Giants score and then decide to go for two and get it. The reality is that coaches aren't brave enough to do that. By our memory, no team has been bold enough to try that move since the Bucs successfully pulled it off against the Redskins in 2005. Obviously, it's easier to win against a team that has to pick up a two-point conversion to tie as opposed to a team that only needs an extra point, but the tradeoff of creating a two-possession game with limited time seems far more valuable.
The other time to pull this move out of mothballs would be against a team that should be extremely effective on two-point conversions. The Vikings faced just that sort of team on Sunday, and when Christian Ponder hit Percy Harvin for a 48-yard touchdown pass with 9:41 left, they led the Broncos, 28-21, pending the extra point.
You can probably see where this is going already. Obviously, the time situation isn't as big of a concern, since it's pretty clear that Denver was going to be able to get two drives in, even at the glacial speed normally required for long Broncos drives before the final three minutes of the game. The bigger quandary here is that if the Broncos do score and you're up eight, well, are you really going to be able to stop Tim Tebow from getting into the end zone on a two-point conversion?
We quoted research from Brian Burke earlier in noting that two-point conversions have made it through 47.9 percent of the time. That's the overall rate, but if you didn't click the link, there's a pretty mean split between running plays and passing plays. Teams only call running plays for the two-point conversion about 35 percent of the time, but those running plays succeed 61.7 percent of the time. Furthermore, the most successful type of running play on two-pointers has been runs from the quarterback, which were converted 35 of 47 times, a whopping 74.5 percent success rate. And those figures are from 2000 through 2009, so Tebow isn't even included!
The Broncos had converted on 14 of 23 runs with two yards to go or less since they installed the option with Tebow, a 60.8 conversion percentage in a small sample. The Vikings had stopped the Broncos on two different third-and-short carries earlier in the game, so it's possible that they thought they might have been able to stop Tebow or Willis McGahee if the Broncos ran for a two-pointer, but it's Tim Tebow!
The logic for this one ended up being a lot less relevant than it would have been in the Packers game, since the Broncos drove down the field and almost instantly scored before converting the two-pointer to tie it up. The defenses on both sides were playing so poorly that a nine-point lead for the Vikings with 9:41 left probably wouldn't have had a huge impact on the final score. But if you think the odds of Tebow converting are in excess of 70 percent, we get to the point where an eight-point lead isn't really that meaningful, either, so why not try to get a possession ahead?
Of course, NFL coaches shouldn't do this in every situation. Players and situations aren't numbers or percentages, and criticism of aggressive play-calling like this tends to suggest that people like us are saying they are. It's not true; that's where the coaching comes in! A coach needs to know when the right time to make this sort of decision is and actually implement it. In Leslie Frazier's shoes, there wasn't really much to be gained by going for two with nine minutes left, although there also probably wasn't much to be lost. Mike McCarthy, though, could have sealed up a game that his team very well could have lost. No coach is going to be able to analyze every situation in his head and spit out an exact probability to three decimal points of every possible scenario from then on to make an optimal decision, but every coach should be trying to create scenarios with their decisions where their team's chances of winning increase. And if they're too busy to do that in the middle of a game, they need to hire someone and entrust them with the ability to make those decisions on their behalf.
The Dolphins are not to be trifled with, people! A laughingstock of the league after an 0-7 start, Miami has promptly gone 4-1 in their five most recent games, with a one-point loss against the Cowboys in Dallas on Thanksgiving Day offset by four wins. The Dolphins have held their opponents under ten points in each of those four victories, and they have prevailed in those games by an average of 21.5 points. Considering that four of the Dolphins' losses were by a total of eight points, it's pretty clear that Miami is a team that's underrated by their record. In fact, some are wondering whether the Dolphins are the best 4-8 team ever. Is that the case?
We don't have advanced stats like DVOA going back through the '70s, so we'll make do in comparing teams across different years by using point differential. As we've mentioned in the past, point differential does a better job of predicting a team's win-loss record in the future than that team's actual win-loss record does, so it stands that point differential is probably a more accurate indicator of how that team is playing.
As you might suspect, with four big wins and four very close losses in 12 games, the Dolphins have a pretty solid point differential. After the 20-point win over the Raiders, they have now outscored their opposition this year by 26 points. That produces an expected winning percentage of .571, or roughly equivalent to a 7-5 record. It leaves the Dolphins as just the fifth 4-8 team since the merger to have a positive point differential, and among those five, Miami has the third-best differential.
This is all irrelevant statistical trivia, of course, unless it means something for the Dolphins going forward. If we look at those four 4-8 teams who had a positive point differential, there's some reason for Miami fans to be hopeful. Those four teams had a total of 14 games to go in their season (since one, the 1971 Bengals, played a 14-game schedule) and went a combined 10-4. That's not bad.
Even more impressively, though, those teams were able to hold on to their gains in the following season. Each of them were .500 or better in the following year, and they combined to go 39-23. The last team to pull off the feat was the 2008 Chargers, who started 4-8 and still won the AFC West when the Josh McDaniels-led Broncos collapsed. A year later, they were 13-3 and picked up a first-round bye.
The Dolphins might not have that kind of turnaround in them next year, but with a crop of young talent on defense, they're not as far away from contention as it seems. Matt Moore's been an effective quarterback since taking over for the injured Chad Henne, but the Dolphins could choose to upgrade the position further in the offseason by drafting a quarterback (although they are now out of the hunt for Andrew Luck) or acquiring a veteran to compete with Moore. If they can get average play at quarterback and the veteran rosters of the Jets and Patriots slip some, they could win the AFC East next year. Certainly, they're going to be one of the playoff sleepers heading into 2012.
Bill Barnwell is a staff writer for Grantland.
Previously from Bill Barnwell:
A Requiem For The Dream Team From Philly
The Annual Pursuit of Quitting on Tom Coughlin
The Surreal World of Thanksgiving in Vegas
To comment on this story through Facebook, click here.
The Cowboys lost the game in overtime when the Cardinals blocked a punt, but the decision to ice Folk clearly didn't help Arizona's chances of winning the game.