It’s a tie game with time running out. Your opponent has brought out the field goal unit to attempt the game winning kick. What do you do?
Obviously, you take on the kicker in an unseen but fiercely fought psychological battle, imposing your formidable psionic will upon his fragile pyche as you display your mastery of the metaphysical arts by . . . calling a timeout. Or maybe you just give him a practice kick and time to make adjustments.
To ice or not to ice. Just about every coach seems to do it at one time or another. Opinions vary among fans, but it often boils down to “it’s stupid . . . unless it works”. But what does “it works” look like?
Kickers miss about 1 in 5 field goals so how can you identify if a single miss was caused because of a timeout? The answer is you can’t. You already know that correlation is not necessarily causation because you have faithfully read all my articles, amiright? . . . am I right? Therefore, just because one event is followed by another does not mean that the latter event was caused by the former.
We’ll never know if a specific kick was due to a pysche-out, but if we look at all field goal attempts, we can at least determine if there is even a trend to begin with. So, on to the data!
Since 2000, kickers have had a success rate of about 82% on field goals. As a Colt, Vini has rarely been iced, but he is 83% when iced and 87% when not iced (Jesus, that just hit me when I typed it . . . 87%!).
If we split the results for all kickers into iced and not iced situations, it looks like this:
OK then. Icing the kicker works. Thanks for reading.
Of course you know it’s never that easy. Just because there is a 6% differential doesn’t mean that we have accounted for everything.
Think about this, often times the offense calls the timeout for clock management purposes. In those situations, kickers have a 76.9% success rate, which is almost identical to iced kicks. So did the offense ice their own kicker?
It’s more likely that something else is causing these results. The average line of scrimmage for kicks without timeouts is the 18 yard line (about a 35 yard attempt). However, iced kicks are longer attempts with an average LOS of the 21 yard line. And the LOS for kicks where the offense calls a timeout? Also the 21.
Distance is definitely a factor, so to account for it, let’s look at results for all attempted lengths:
The black dots are the un-iced kicks and they show a fairly stable trend with success rate falling off for the longer kicks. The blue dots are the iced kicks and while they vary, they are not statistically different from the un-iced results (p-value = 0.19).
So that doesn’t support that icing works, regardless of the decisions made by every coach ever.
The previous graph is a little hard to read and so collapsing the data into 5 yard bins will make it easier to digest visually (is that a thing?). Also, I’ll add a category for attempts with < 2 minutes left in the game where the offensive point differential is between -3 and 0. These are the double-secret-clutch-ice make it or lose situations that are sure to crush the spirit of every kicker.
Hmmm. Double-secret-clutch-ice situations don’t seem to faze kickers at all. As a Colt, Vini has seen 2 double-secret-clutch-ice situations: he’s 2 for 2.
The bottom line is that successfully icing the kicker has no empirical support, which in statistical terms is also known as “pseudo psychological crap.”