Writers of Pro Football Prospectus 2008

03 Sep 2010

Study Says 'Icing' Causes NFL Kickers To Choke

Well, this is annoying. A study by the University of San Diego says that icing the kicker, the NFL's most irritating strategy, is actually pretty effective. They found that non-iced kickers converted their field goals 80.4 percent of the time, while iced kickers were successful on just 66.4 percent of kicks. The study will appear in the September issue of the journal The Sport Psychologist.

Now we're going to have to see this in every close game this season. Dammit.

Posted by: Vincent Verhei on 03 Sep 2010

37 comments, Last at 01 Oct 2012, 1:49pm by Pennlion


by Spielman :: Fri, 09/03/2010 - 7:28pm

Okay, raw numbers, that sounds pretty impressive. However, did they control for the length of the kicks involved? I would *assume* they wouldn't be that stupid, but the article doesn't make that clear.

I will always carry the image of Ken Whisenhunt looking sheepish on the sideline after trying to ice Dallas's kicker on a game-tying long FG in 2008... that the Cardinals blocked. Hilarious.

by dmb :: Fri, 09/03/2010 - 7:48pm

It was also my first thought that the study would need to control for distance. The linked article only cited raw percentages, and didn't include a link to the study itself.

by Noah Arkadia :: Sat, 09/04/2010 - 10:31am

And mine as well.

by Brendan Scolari :: Fri, 09/03/2010 - 7:50pm

At first I thought your point regarding the distance of the field goals was irrelevant, as I couldn't see any reason field goals after "icing" would be longer on average. On second thought though, I would bet they quite a bit further, simply because coaches almost never "ice" kickers except at the end of the half or the game, and at both of those points teams often run out of time so they have to end drives prematurely to try to get some points, if that makes any sense.

It doesn't say anything either way in the article, but that seems like it could be a possible explanation for the difference in FG%.

by PTORaven :: Sat, 09/04/2010 - 2:54pm

It's a shame this study didn't go into correlation to distance (or temperature, which I think is where the word icing comes from), because it'd be useful for coaches to see at what distance icing becomes effective. If I'm a coach down 3 with 2 minutes left and my opponent is lined up at the 27 for a FG, I'd wanna know if icing would decrease their probability 2% or 20% before I burned one of my last timeouts.

by Pennlion (not verified) :: Mon, 10/01/2012 - 1:49pm

In their book Scorecasting, Tobias J. Moskowitz and L. Jon Wertheim marshal the most compelling evidence to date on the subject, analyzing “pressure” kicks from 2001 through 2009 while controlling for distance of the field-goal attempt. They found that icing the kicker certainly doesn’t produce the desired effect, and in some cases might even backfire. 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.

by ChargerJeff :: Fri, 09/03/2010 - 7:48pm

The other variable is the time of the game. Icing is FAR more prevalent towards the ends of halves, and kickers are already more nervous in those situations. Did they control for that variable? I don't see how they could.

"Shut up," I explained.

by Anonymous Coward (not verified) :: Fri, 09/03/2010 - 7:58pm

A sample size of 160? Seriously?

by dmb :: Fri, 09/03/2010 - 8:06pm

Only for the control group; "total" n = 271.

by John (not verified) :: Fri, 09/03/2010 - 8:32pm

It was a sample size of 325. The no icing + the icing + the kicks on which the kicker's own team called the time out, right?

Over 6 seasons, that would average out to roughly .2-.25 kicks a game. So, obviously, they had to only include kicks in some sort of "pressure" construct. Just educated guesses until the whole study is publicized.

by dmb :: Sat, 09/04/2010 - 10:23am

I assumed that the kicks for which the offense called a timeout were a subset of the "non-iced" kicks, but upon re-reading the article doesn't really seem to make that clear.

by arias :: Wed, 02/01/2012 - 2:46am

It's the defense that calls the timeout to give the kicker a few extra minutes to "think about" the kick he absolutely has to make unless he self destructs and chokes.

What would be the point of the kicker's own team calling a timeout? Icing their own guy is defeatist.

by arias :: Wed, 02/01/2012 - 3:26am

It's the defense that calls the timeout to give the kicker a few extra minutes to "think about" the kick he absolutely has to make unless he self destructs and chokes.

What would be the point of the kicker's own team calling a timeout? Icing their own guy is defeatist.

by Raiderjoe :: Fri, 09/03/2010 - 8:21pm

Never choked on icing but one time when kid alsmot choked on cake part. Was really happy to have cake one day and started to eat fast and put too m uch cake on fork and almost choekd on it.

by Zheng :: Fri, 09/03/2010 - 9:15pm

Hey, that was actually funny.

Not a big icing fan myself either. It's just too sweet.

by Andrew Potter :: Sat, 09/04/2010 - 2:16am

You're obviously doing it wrong. Get the guy across from you to call a time out just as you start to swallow it.

by Still Alive (not verified) :: Mon, 09/06/2010 - 2:03pm

I don't generally like the Raiderjoe comments (he goes after too much low hanging fruit), but this one is priceless.

by The Ninjalectual :: Fri, 09/03/2010 - 9:05pm

Didn't FO publish/link to a study a couple years ago that concluded the opposite? That after an "icing" timeout FG% actually improved slightly?

EDIT: I found the XP to the study, but its conclusion agrees that icing works...

"Just look at that pumpkin."
-John Madden, looking at the moon.

by Jim A (not verified) :: Fri, 09/03/2010 - 10:40pm

Dr. Z debunked that study by using a larger sample size.


by Noah Arkadia :: Sat, 09/04/2010 - 10:27am

I was thinking about this whole icing matter, so I sent my query along to the good Doctor Stamms of Stats, Inc. He came up with the following: Pressure kicks (using the same guidelines as described above, except that they're in the last two minutes, not three) since 1991, regular season -- 457 of 637 (71.7 percent) made, without icing on them. After icing, the number is 152 of 211 (72 percent). So it's a push. Next week we'll discuss kicks with frosting instead of icing.

by elhondo :: Fri, 09/03/2010 - 10:10pm

Not far from the USD campus, another study tragically proved conclusively that playoffs also cause kickers to choke.

by Nathan :: Fri, 09/03/2010 - 10:49pm

Icing before the offense lines up (or just after) doesn't really bother me. It's gamesmanship and using the last bullet in your gun as a HC, especially at the half or game.

Icing right before the snap causing a kick that looks like it counts but doesn't is annoying. I'm surprised they didn't tweak the rules in the offseason after Shanahan (IIRC) first busted it out. Maybe they justified it as you might give a kicker who missed a second chance but it just looks sloppy.

Maybe the HC should be able to roll the dice and attempt to call a timeout at the last second but if you wait so long that the ball gets snapped before the refs can whistle the play dead (it's a judgement call but let's be honest, you'd be able to tell if the longsnapper tried to frame it by watching if the rest of the line reacted) you get a 5 yard unsportsmanlike conduct penalty. Just enough to make you think twice about taking the risk.

by Bright Blue Shorts :: Sat, 09/04/2010 - 5:55am

In principle you're right ... but it's just adding another layer of complexity to the rules that the game really doesn't need. The last thing the game needs is another judgement call from refs to argue over ...

by tuluse :: Tue, 09/07/2010 - 12:43am

Agreed, I don't understand why is bothers some people so much.

You just watched 3 hours of football, does another 2 minutes before the resolution really affect you that much? Sometimes I even enjoy the extra anticipation.

by Harmy G (not verified) :: Fri, 09/03/2010 - 11:39pm

When I first read this headline in my RSS feed, I thought it was coming from The Onion...

by Mike Y :: Sat, 09/04/2010 - 1:05am

Does this study say that kicks occurring after a time out are compared to all other kicks? Because no one would call a time out for a field goal in the 1st or 3rd quarter, and those kicks are done with less pressure on the kicker. All "iced" kicks should be compared with all other kicks in pressure situations (last two minutes of each half in a close game). What would the results say then? Another factor to consider would be whether the kicker actually kicked the ball after the other team called a timeout. I think it is an advantage if the kicker kicks the ball after a timeout, it is like a practice kick. But the study is useless if it is comparing pressurized "iced" kicks to all other kicks, whether they were "pressurized" or not.

by Bright Blue Shorts :: Sat, 09/04/2010 - 5:52am

From the summary given the study is valid. There are about 1,000 field goal attempts per season; yet the non-icing data only has 160 kicks from the 2002-2008 period. That suggests it's only looking at pressured kicks (i.e. last 2 mins of half )

I'd like to see two further breakdowns:
1) are some kickers better than others at not being affected by icing.
2) is there a difference between a kick to break a tie, and a kick to not lose (e.g. the Super Bowl kicks of Adam Vinatieri vs Scott Norwood). To me the former has less pressure than the latter.

Unfortunately even though Dr Z goes back to 1991 in his 2005 debunking; it's not valid because kickers in the 2000s are about 10% better than kickers in the 1990s. The nature of pressure kicks means that you will always have a small sample size.

by Jim A (not verified) :: Sat, 09/04/2010 - 8:51am

If there's a minute left and the clock is running on 4th down, the defense will usually call time out to stop the clock so it has more time to score on its final possession. I wouldn't really call this icing the kicker. You really need to look at only the last 15 seconds at most or overtime.

I am a bit skeptical that there are actually 18 instances of icing the kicker (110 over a six-year period) per season, as the data cited in the article implies.

by AlanSP :: Sat, 09/04/2010 - 10:09am

I'm not sure that kickers being better overall in the 2000's really invalidates Dr. Z's conclusion unless the rate at which coaches tried icing the kicker also changed.

As far as whether some kickers are better than others at not being affected by icing, there probably just isn't a large enough sample size to answer that. As you point out, sample sizes for this type of situation are small even when you're aggregating across the league, so trying to look at individual kickers in pressure situations is going to be basically impossible.

As far as the Norwood vs. Vinatieri comparison, the summary did say that "other factors such as experience, game location or game score were not associated with success," which would imply that it doesn't make a difference, but that could mean a number of things depending on their methodology.

by tuluse :: Tue, 09/07/2010 - 12:42am

Well if you have an equal number of attempts both iced and un-iced from each era, it wouldn't matter. However, lets say icing has become much more prevalent in the past 10 years, then most of your icing attempts would involve the more accurate kickers.

by arias :: Wed, 02/01/2012 - 5:16am

So what?

How does that invalidate the overall trend of a decrease in completion % after getting iced to not getting iced at all?

by Dean :: Sat, 09/04/2010 - 10:25am

I haven't had a chance to read the study in detail and haven't had a chance to read the comments yet (and won't until next week - I'm about to go tailgate, get drunk and watch Mizzou/Illinois - not that I care about either team)

I'm wondering about:

Sample size.
Length of kicks.
How did they establish what constitutes a clutch kick for non timeout purposes?

by AlanSP :: Sat, 09/04/2010 - 10:30am

I can't be sure until seeing the actual study, but from the description, it doesn't sound like they adjusted for things like weather conditions, and most importantly, FG distance. This is problematic if coaches are more likely to try icing hard kicks (e.g. a 50+ yd FG in windy conditions) as opposed to easy ones (e.g. a 25 yd FG in a dome). The Berry and Wood study in 2004 adjusted for these factors, but it's unclear whether the current study did.

Personally, I've always disliked the strategy mainly because it's often used in situations where the timeout could be put to much better use. Even if icing has an effect, I'm skeptical that its value is greater than the extra 40 seconds that a timeout buys (if used on defense, as it ideally should be. On offense, it's probably more like 15 seconds, which is still pretty valuable in close/late situations).

by PTORaven :: Sat, 09/04/2010 - 2:59pm

A new study has found that icing a kicker with Smirnoff (http://www.urbandictionary.com/define.php?term=icing) is far more effective than icing a kicker with a time out.

by Joseph :: Sat, 09/04/2010 - 8:07pm

Two things, although most of the points I see above are quite valid:
1. I think that if the AWAY team's kicker is kicking, it probably has more effect (fan's on the bottom near the away team sideline should be SCREAMING insults and what not at the kicker).
2. In OT, that timeout probably doesn't matter--nor at the end of a half/game when the clock is under 30 seconds. If you have two or three, it FOR SURE doesn't matter.

IMO, the Vikings icing Garrett Hartley at the end of the NFC championship game might have been counter-productive. I mean, that FG was good from the moment he kicked it.

by langsty :: Sun, 09/05/2010 - 9:01am

i'd think the alcohol would relax them

by Chocolate City (not verified) :: Tue, 09/07/2010 - 9:47am

Ever try to kick sh*#faced? The holder's in trouble.