 |
| Seahawks Jermaine Kearse catches this 35 yard TD pass from QB Russell Wilson |
Though the decision paid off, we want to look at the mathematics behind the decision to figure out if it was a correct one or not. Which option was the best in this situation?
The Situation and Options
At the start of the 4th quarter (13:52 remaining), the Seahawks trail 13-17. The ball is at the 49ers 35 yard line and it's now 4th and 7. The three options here are all seemingly reasonable:
In order to figure out which option is the best, we'll have to make some assumptions and use the Win Probability calculator from Advanced NFL Stats, to break down the expected outcome of each situation.
- Punting
- Attempting a Field Goal
- Going for it
In order to figure out which option is the best, we'll have to make some assumptions and use the Win Probability calculator from Advanced NFL Stats, to break down the expected outcome of each situation.
Punting
Coaches in this situation sometimes choose to punt, rationalizing a 4th-and-7 is difficult to make, and a 53 yard field goal attempt is not guaranteed either. In this situation, there's also nearly the entire 4th quarter left to be played, so coaches "trust their defense" to get a stop and get the ball back for a better shot at a touchdown. Trusting the defense makes more sense with a team like the Seahawks, who were the #1 team in the league this year for yards allowed (regular season).
A punt from the 35 yard line is likely to go into the end zone, but might also be able to pin the 49ers deep in their own territory. Using the calculators at Advanced NFL Stats.com, we can figure out both the Win Probability and the Expected Points for the Seahawks in either case.
Expected Points for Seahawks
EP (Pin on the 2 yard line) = 0.40 points
EP (Touchback) = -0.34 points
From these numbers, if the Seahawks successfully pin the 49ers deep, they are more likely to score next, whereas on a touchback, the 49ers are now more likely to score next. The break even point is about the 14 yard line meaning the team with the ball is more likely to score next if they have the ball in front of the 14 yard line, and the team on defense is more likely to score if the ball is behind the 14 yard line.
Win Probability for Seahawks
WP(2 yard line) = 0.32
WP(Touchback) = 0.30
Notice the Win Percentage for the Seahawks doesn't change much in either case. This is likely due to the fact they're down 4 points. 0.7 point swing doesn't do much to help their chances, even with a touchback.
This calculation is using average numbers. The Seahawks, being a great defensive team, might have their win percentages actually tick up a little bit in this case.
Attempting a Field Goal
A 53 yard field goal try will be successful on average 56% of the time, according to this broad analysis of kicking between 2002 and 2012, certainly not an automatic field goal. Carroll actually considered kicking a field goal, trotting out kicker Steven Hauschka before changing his mind and calling a timeout, though this decision might have been made easier because Hauschka was a little late to get onto the field.
Expected Points for Seahawks
A made field goal is obviously 3 points and a missed field goal 0 point, but we'll also need to take into account the 49ers next drive to see what the total Expected Points for this decision is.
On a missed field goal, the 49ers would take over on their own 43 yard line, with an EP of 1.63 points, owing to the good field position they will be getting.
On a made field goal, the 49ers would be receiving the ball after a kickoff. On average, they would be receiving the ball on their own 23 yard line, leading to an EP of 0.46
The total EP for kicking a field goal for the Seahawks is thus:
EP(FG) = (0.56)(3-0.46) + (0.44)(-1.63)
EP(FG) = (0.56)(2.54) + (0.44)(-1.63)
EP(FG) = 1.4224 - 0.7172
EP(FG) = 0.7052 points
The expected points for a field goal attempt is higher than both outcomes for punting, so it might make sense to kick the field goal instead of punt. However, we should look at how it affects win probability as well.
Win Probability for Seahawks
Looking a little closer at the Win Probabilities shows a different picture.
WP (Made FG) = 42%
WP (Missed FG) = 25%
Clearly, making a field goal should give the Seahawks a better chance of winning, though down by 1 point and giving the ball up to the 49ers, they are still likely to lose. A missed field goal would make it even MORE likely to lose, because their field position would now be worse than if they had punted, with no points to show for it.
WP (FG) = (0.56)(0.42) + (0.44)(0.25)
WP(FG) = 0.2352 + 0.11
WP (FG) = 0.3452
On average, a team kicking a field goal in this situation would win the game 34.52% of the time, greater than the chances of punting from the calculations we made. Even if we adjust for the Seahawks being a better team, BOTH win percentages of kicking and punting would go up, leading this to still be a close decision.
Going For It
Finally, what if the Seahawks go for it on this 4th down instead of kicking or punting? A 4th and 7 is not an easy conversion to make, with one estimate looking at a 35% chance of converting. As before, we need to look at the expected points and win probability of converting the 4th down and that of failure.
Expected Points for Seahawks
When converting the 4th and 7, we will assume the ball will be spotted at the 49ers 28 yard line after the requisite 7 yard gain. This will actually understate the Seahawks chance of winning, since there is a chance the play, if successful, will actually gain more yards. This is especially evident in the actual play call in this case.
If the Seahawks fail the 4th and 7 chance, will will assume an incomplete pass and the ball will be given to the 49ers at their own 35 yard line.
Going back to the Win Probability calculator from Advanced NFL Stats, we can see the following numbers for this situation:
EP (Success) = 3.47
EP (Failure) = -1.15
These numbers reflect the chance for scoring a touchdown on 4th down success, and giving the ball back in good field position in the case of 4th down failure. The total Expected Points if the Seahawks go for it is:
EP(Go For It) = (0.35)(3.47) + (0.65)(-1.15)
EP(Go For It) = 1.2145 - 0.7475
EP(Go For It) = 0.467
The Expected Points for this play is actually much lower than a field goal try, given the relative high failure rate. However, the real calculation lies in looking at the Win Probability if they go for it.
Win Probability for Seahawks
Since a touchdown takes the lead, we expect quite a wide swing in the Win Probability for the Seahawks if they can convert. Using the WP calculator, we find the following Win Probabilities:
WP (Success) = 0.45
WP (Failure) = 0.27
Which leads to the aggregate calculation:
WP (Go For It) = (0.35)(0.47) + (0.65)(0.27)
WP (Go For It) = 0.1645 + 0.1755
WP (Go For It) = 0.34
Thus, going for it, gives an average team a 34% chance to win.
Conclusions
Looking just at Expected Points gives us a much different picture than that of looking at the Win Probabilities for the Seahawks in this situation. Looking at the data in a table form:
| Punt (Pin on 2) | Punt (Touchback) | Field Goal | Go For It | |
|---|---|---|---|---|
| Expected Points | 0.40 | -0.34 | 0.7052 | 0.467 |
| Win Percentage | 32% | 30% | 34.52% | 34% |
Surprisingly, though the number of expected points varies from -0.34 to 0.7052, the Win Probabilty chances for all 3 choices runs close together. For an average team, this means any decision made here is going to give roughly the same percentage of winning.
Since these calculations are meant to look at an average team, and the choices run close together, it's now much more important to look at the team's strengths and weaknesses. Whereas if the different decisions produced wildly different win probabilities, a team's strengths and weaknesses matter less.
In this case, a better defense, like the Seahawks have, lead to kicking or punting as a better decision, since they would be more likely to stop the opponent's offense. With faith in Russell Wilson and your offense, it pushes the decision to go for it.
Chance of Success with the Actual Play
In the actual game, Pete Carroll called a play with 3 vertical pass routes on the right hand side, basically going for a touchdown on this play. Since the 35% number was the chance of gaining just 7 yards, it stands to reason the 35 yard pass play would have a lower success rate than that of a short 7 yard pass. However, because success would guarantee a touchdown, the chance of success on this one play doesn't have to be as high as 35%.
If a touchdown is scored here, the Seahawks' Win Probability jumps up to 0.62 (after an average kickoff). The Win Probability in the event of failure remains at the same 0.27 we found earlier. We can set up an equation to see what the success rate on this play needs to be in order to make it the same win probability as calculated earlier.
WP(Go for TD) = (x)(0.62) + (1-x)(0.27)
WP(Go for TD) = 0.62x + 0.27 - 0.27x
WP(Go for TD) = .35x + 0.27
Since the general Win Probability of going for it was 0.34, we can set the two sides equal and try to solve for x:
0.35x + 0.27 = 0.34
0.35x = 0.07
x = 0.20
The chance of success for taking a shot at the end zone only needs to be 20% in order to make it equal to punting, kicking a field goal, or making a 7 yard pass to gain a first down.
One thing was clear. In pulling the kicker off the field, Pete Carroll put the game in the hands of his offense, and was handsomely rewarded on this play. Though the defense would ultimately make the stops needed to preserve the victory, this play allowed the Seahawks to take the lead and dictate the pace of the remainder of the game.
I re-watched the play, and I'm not convinced that they weren't just trying to pull the defense offsides to get a free play. If the free play fails, it's now 4th-and-2, which is easier to convert, or they're 5 yards closer for the FG. If they fail to draw the defense offsides, they take a delay of game and punt. The center only snapped the ball when the defensive player jumped offsides, and all of the receivers were just running go routes to the end zone, much like a hail mary. I'd think the receivers would be running more high-percentage routes if all they were trying to do was convert the 4th-and-7.
ReplyDeletehttp://www.nfl.com/videos/nfl-cant-miss-plays/0ap2000000315401/NFC-Championship-Can-t-Miss-Play-A-gift-and-a-Kearse
That is certainly a real possibility with the way it went down. However, it probably does tell us about Carroll's psyche that he was never going to kick the field goal, but intended on going for it (on 4th and 2) or punting (from 4th and 12). Remember, they had just faced a 3rd and 22 the play before, and are now in a much better spot than they just had been.
ReplyDeleteEither way, I still think it's close, but even on a 4th and 2 or a 4th and 12, the WP that changes the least is the punting number. Going for it and a FG try are both quite a bit more likely (or not) depending on the outcome of the "draw them offsides" play.