posted on Grantland.com by bill barnwell
Let's skip to the fourth quarter and take on the game's three big coaching decisions in chronological order. We'll start with Bill Belichick's decision to throw the red challenge flag out on Mario Manningham's enormous catch for 38 yards with 3:46 left, a play that finished virtually right in front of Belichick on the Patriots' sideline. Neither Belichick nor his video people likely got a chance to review a replay before throwing the challenge flag, which is normally a sign of a terrible challenge, but this was a situation where throwing the flag made total sense. It was a low-risk, high-reward challenge.
Why is that? Well, the reward is obvious: If Manningham happened to step out of bounds, the Patriots wipe the game's biggest play off the books and send the Giants back to their 12-yard line. It's an enormous shift in field position, particularly in a game where big plays had been so hard to find. If you believe in the power of momentum, a replay review would calm down the Giants-friendly crowd and give the Patriots a chance to recover from the shock of the play, even if the challenge ended up unsuccessful. The risk of losing a challenge is basically nil, since the Patriots had two challenges left with just under two minutes of game time to go before they lost them. The risk here is that you lose one of your three timeouts, and as it turned out, the Patriots ended up wishing they had that timeout when they were trying to stop the Giants near the goal line. Had the Patriots been able to stop the Giants short on three consecutive plays after the two-minute warning, they could have held Big Blue to a field goal attempt and still had plenty of time to try a drive for the win with Brady. That's why it's low-risk and not no-risk. Every decision like this in a close game carries a certain amount of risk and reward; a good coach considers risk without being unnecessarily averse to it. Bill Belichick, as you might suspect, is a good coach.
Next, Belichick sullied all that is right about the game of football by allowing the Giants to score on an Ahmad Bradshaw run with 1:04 left in the fourth quarter, giving the Giants a four-point lead while allowing his team to get the ball back in an attempt to drive for the winning touchdown with some reasonable amount of time. Bradshaw now-famously realized what was going on mid-play and tried to delay himself from scoring, but let's review the decision-making heading into the play. What should each team have done?
Win Probability charts aren't perfect because they don't adjust for the teams involved, but they're the best tool for answering a question like this. Here, the Giants-Patriots WP chart on advancednflstats.com notes that the Giants had an 89 percent chance of winning the game when Hakeem Nicks picked up a first down on the New England 7-yard line with 1:09 left. From there, the Giants could have chosen to kneel three times, force the Patriots to use their final timeout, and then attempt a game-winning field goal with seconds on the clock without ever giving the ball back to the Patriots. The model might even be underestimating their chances; history suggests that an average field goal kicker will convert a 24-yard field goal about 96 percent of the time, and the Giants were playing on turf with the options to both move the ball onto Lawrence Tynes' desired hash mark while falling on the ball and trying again in the case of a bad snap. And if you think Tynes is a terrible kicker, note that he's 56-of-57 on kicks from 20 to 29 yards during his career.
Instead, when Bradshaw scored the most mournful game-winning Super Bowl touchdown in history, the Win Probability analysis suggests that the Giants' odds of winning decreased to 85 percent. That's right: Bill Belichick was likely correct to allow the Giants to score, and the Giants should have taken a knee and decided to kick the chip shot field goal instead.3 If you use the 96 percent win expectancy that we're suggesting instead of the model's 89 percent, it's patently obvious that the Giants should have kneeled and kicked.
this still won't convince OP he's wrong because he's never seen it work.
Let's skip to the fourth quarter and take on the game's three big coaching decisions in chronological order. We'll start with Bill Belichick's decision to throw the red challenge flag out on Mario Manningham's enormous catch for 38 yards with 3:46 left, a play that finished virtually right in front of Belichick on the Patriots' sideline. Neither Belichick nor his video people likely got a chance to review a replay before throwing the challenge flag, which is normally a sign of a terrible challenge, but this was a situation where throwing the flag made total sense. It was a low-risk, high-reward challenge.
Why is that? Well, the reward is obvious: If Manningham happened to step out of bounds, the Patriots wipe the game's biggest play off the books and send the Giants back to their 12-yard line. It's an enormous shift in field position, particularly in a game where big plays had been so hard to find. If you believe in the power of momentum, a replay review would calm down the Giants-friendly crowd and give the Patriots a chance to recover from the shock of the play, even if the challenge ended up unsuccessful. The risk of losing a challenge is basically nil, since the Patriots had two challenges left with just under two minutes of game time to go before they lost them. The risk here is that you lose one of your three timeouts, and as it turned out, the Patriots ended up wishing they had that timeout when they were trying to stop the Giants near the goal line. Had the Patriots been able to stop the Giants short on three consecutive plays after the two-minute warning, they could have held Big Blue to a field goal attempt and still had plenty of time to try a drive for the win with Brady. That's why it's low-risk and not no-risk. Every decision like this in a close game carries a certain amount of risk and reward; a good coach considers risk without being unnecessarily averse to it. Bill Belichick, as you might suspect, is a good coach.
Next, Belichick sullied all that is right about the game of football by allowing the Giants to score on an Ahmad Bradshaw run with 1:04 left in the fourth quarter, giving the Giants a four-point lead while allowing his team to get the ball back in an attempt to drive for the winning touchdown with some reasonable amount of time. Bradshaw now-famously realized what was going on mid-play and tried to delay himself from scoring, but let's review the decision-making heading into the play. What should each team have done?
Win Probability charts aren't perfect because they don't adjust for the teams involved, but they're the best tool for answering a question like this. Here, the Giants-Patriots WP chart on advancednflstats.com notes that the Giants had an 89 percent chance of winning the game when Hakeem Nicks picked up a first down on the New England 7-yard line with 1:09 left. From there, the Giants could have chosen to kneel three times, force the Patriots to use their final timeout, and then attempt a game-winning field goal with seconds on the clock without ever giving the ball back to the Patriots. The model might even be underestimating their chances; history suggests that an average field goal kicker will convert a 24-yard field goal about 96 percent of the time, and the Giants were playing on turf with the options to both move the ball onto Lawrence Tynes' desired hash mark while falling on the ball and trying again in the case of a bad snap. And if you think Tynes is a terrible kicker, note that he's 56-of-57 on kicks from 20 to 29 yards during his career.
Instead, when Bradshaw scored the most mournful game-winning Super Bowl touchdown in history, the Win Probability analysis suggests that the Giants' odds of winning decreased to 85 percent. That's right: Bill Belichick was likely correct to allow the Giants to score, and the Giants should have taken a knee and decided to kick the chip shot field goal instead.3 If you use the 96 percent win expectancy that we're suggesting instead of the model's 89 percent, it's patently obvious that the Giants should have kneeled and kicked.
this still won't convince OP he's wrong because he's never seen it work.