What the NFL Can Learn from Olympic Curling

Last night the men’s curling teams of Scotland Great Britain and Norway battled in what was a positively enthralling one-game playoff for the medal round, after finding themselves tied for the fourth and final spot by going an identical 5-4 through the preliminary round robin. Seriously, enthralling. (Yes, it is curling, but get over it because curling is actually awesome.)

Observe:

What the hell is that? A snake?

What the hell is that? A snake?

Britain prepares to curl...

Britain prepares to curl. Suave. But they could use better pants.

UK Curling

And they add a fourth freeze shot just outside the house!

Norway's skip lines up the potential triple takeout.

Norway’s skip lines up the potential triple takeout.

Wait is this a billiards break or WHAT IS GOING ON?

There was a lot to love about that match, the first curling match I have watched live and in its entirety, though I had seen large swaths of some of the United States’ matches over the last week. Ed Lukowich, two-time Brier champion, former director of athlete development at the United States Curling Association, and current Olympic commentator is the Vin Scully of curling, making the game a thoroughly enjoyable experience for newcomers and old-timers alike. And of course curling is “chess on ice”, a remarkable blend of finely tuned skill and the ever-present strategy of an alternating move game. Yet like all good sports, curling is quite capable of coming down to a few key plays, perhaps even to the final shot. Such was the case last night (/ morning in Sochi).

Appreciating the finer skills of a good curl are still quite beyond me, but this article from The Independent provides a good sense of what went down. Real generally, down 5-4 on the final shot of the game, Scotland Great Britain had two options: tie it up, but give Norway the advantage of shooting last going into overtime1, or end the game right there by going for the win, either immediately advancing out of the tie-breaking game to the medal round or seeing their Olympic dreams end in the span of 105 feet of ice. The advantage of going last is a non-trivial, serious deal (again, just get over that it is curling we are talking about), and rather than face Norway with such an advantage, Scotland’s Great Britain’s captain (or rather, “skip”) David Murdoch opted to go for it.

Sound familiar? Throughout an NFL season, teams punt and kick field goals on fourth and short rather than increase their chances of winning by going for the first down or touchdown conversion. The conservative strategy of losing more slowly is the convention, and can occasionally still produce a win, but on net loses far more often than an aggressive strategy built on maximizing a team’s chances of winning. For whatever reason (see: “inertia“), the NFL’s conservative mentality continues, by and large. Riverboat Ron was a welcome addition this past season, but just glance at the “Thank You For Not Coaching” features in Bill Barnwell’s weekly regular season Grantland columns to see a laundry list of poor decisions.

Murdoch, with his fellow Scotsmen on Great Britain’s curling team, went for the win, knowing it would not work every time, but equally confident that it would work often enough to justify it. Well, maybe. Per that Independent article by Ronnie Esplin, Scotland’s Great Britain’s vice-skip Greg Drummond remarked afterwards:

You’re going to make that shot once in 50, maybe two in 50. … We called time out and weighed up our chances to steal the extra end. Against Norway I would rate the chances at 35 per cent on a good day so it was probably worth going for a win.

Said in the intensity of a post-match interview, the remarks are still perplexing: one in 50 amounts to two percent, two in 50 amounts to four percent, and both are far less than the 35 percent chances Drummond gave them in overtime. But unlike Brian Burke, Drummond does not have hard data to back up his claims; those are just his opinions on their chances.

It is hard to say what the true probabilities were. Data of the nature of the NBA’s recent foray with STATS LLC would make it possible, but unfortunately it does not yet exist for curling. Assuming going for the win was the right decision just because it worked is just as wrong as applying outcome bias in football.

Regardless, what the NFL can learn here is the philosophy of risk-taking in general. Do not let the worst-case possible outcome dictate all strategic decisions. The fact that calculated risks are inherently risky does not mean they are worthless. Just ask Scotland’s Great Britain’s curling team.


  1. Curling overtime works like baseball, in that there are “ends”, or innings, rounds, whatever you want to call them, and as soon as there is a score difference after a completed extra end the game is over. In each end there is a huge advantage to going last, sort of like batting last in a baseball inning, but in this context really more akin in magnitude to getting the ball first in overtime in the NFL. By tying the game and extending it to an extra end, Scotland Great Britain would give that advantage of shooting last to Norway. Just FYI, this advantage is known as “the hammer”. Scotland Great Britain did not like their chances against a Norway armed with the hammer. The vernacular may be the best thing about curling. 
Advertisements
1 comment
  1. Ryan said:

    As a Canadian, I feel obligated to explain curling to foreigners. So, the shot GB took was probably a significantly higher percentage chance than 4%; my dad, who is a very experienced curler, figured that as a 15-25% chance, given the skill levels involved. Additionally, if you look at the angles, there is a significant potential for a miss to the outside causing the British to still end up with 1 point, which would send them to extra ends the same as a draw shot would. Perhaps 20% of outcomes would cause that result. Also, the option to take 1 point with a draw shot isn’t a guarantee either. These curlers are very good, but they don’t draw the 4 foot 100% of the time. There’s a non-zero possibility of blowing that shot. Finally, the 35% estimate for winning in extra ends is quite generous. Given equal skill level, the side with the hammer wins extra ends more than 75% of the time.

    Using the numbers I just listed, the win probability taking a draw shot would be slightly less than 25%, while attempting the raise double-takeout they went with is (0.2+0.2*0.25)=0.25; in other words, the difference between going for the win and playing for the tie is very low. Their decision was likely motivated more by their emotional state, than by cool logic. The skip believed he could make the shot, and he turned out to be correct.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: