Advertisement

Expanded study boosts case for FGs/punting, but coaches still gamble less than they should

andrew bucholtz

One of the most hotly-discussed parts of Canadian football is what teams should do on third down. Do you try and pick up the first down, attempt a field goal or punt and attempt to pin the other team deep? Rob Pettapiece has been looking at this from a statistical perspective since 2010, publishing his first results in February 2011 (which generally stated that teams didn't try for the first down enough) based on the 2009 CFL season, considering how different decisions might have swung the 2010 Grey Cup and examining how often each team made the best decision from a probability standpoint throughout the first part of the 2011 season. This offseason, Pettapiece decided to expand his study, examining all the data from the 2008-2012 seasons instead of just 2009, and the results he published Monday demonstrate some rather surprising things. Namely, kicking field goals and punting have generally been much more effective options from a statistical point of view than the 2009 data suggested. Here's the chart he developed of what decision's optimal at each spot on the field:

And here's the expected value of each position on the field based on the average number of points the next time either team scored, data Pettapiece used to produce the above chart:

In Pettapiece's conclusion, he writes that one of the main differences from the 2009 data is that kicking a field goal or punting on 3rd and 2 appears to be the best move in certain circumstances given the expanded data:

I used to be adamant about going for it all the time on 3rd and 2, but the (new, improved) numbers don’t always say that’s a no-brainer. It’s clearly the best call, by this analysis, only when you’re close, then beyond the 25 a field goal attempt is an equal or better bet while you’re still in range. After that you wouldn’t be wrong if you punted. It pains me to admit that coaches are not wrong to punt here. (My previous 2009 numbers were
either wrong or an aberration from five-year trends.) Then again, at no point is going for it on 3rd and 2 or 3 more than half a point worse than punting. (You would see this if I were to widen the “toss-up” range in the middle.) There is always flexibility based on the specifics.

However, even with that, CFL coaches from 2008 to 2012 were still more conservative than probability would suggest is optimal. From another part of the conclusion:

Coaches rarely go for it on 3rd and more than 2, even when they have advantageous field position, and they probably should more often. Inside the 20, and with between 3 and 5 yards to go, coaches attempt a field goal an incredible 83% of the time. In almost all of those cases, going for it is justifiable, and it’s
sometimes the superior option.

As discussed back in 2011, that's probably largely thanks to coaching conservatism; plenty of analysts and media, especially traditionalists, always encourage "taking the points" from a field goal, and coaches tend to take more flack for third-down gambles that fail than they do for playing it "safe", even if that's a worse option from a probability standpoint. Another area where that conservatism shows up is in conceding safeties, which field position and expected points show are almost never worth it. From an e-mail discussion with Pettapiece:

The basic theory is that you have to be really concerned about the punt distance to concede the safety. Two points = about 35 yards. Teams usually start with the ball around their 40 or 45 following a conceded safety. For that to be a better bet than punting, you'd have to assume they would return the punt to your 30 or 35. Since the average net punting distance is 35 yards, the conceded safety really only makes sense when you are hopelessly pinned or if the wind is a factor. ... If the line of scrimmage is the 10, for instance, then an average punt would go to the 45 or so. Those extra 20 yards aren't worth the two points.

So, while there is a stronger case for field goal attempts and punting on third downs than previously thought, especially a ways from the opponents' goal line, CFL coaches still tend to behave more conservatively than the data would suggest. That may change, though. Keep in mind that the CFL's new coaches this offseason are pretty young: that's 42-year-old Rick Campbell in Ottawa, 43-year-old Mike O'Shea in Winnipeg and 46-year old Chris Jones in Edmonton. They're all new to a head coaching job, too, and won't necessarily carry on doing things the way they've been done. (Of course, former Eskimos coach Kavis Reed is only 40, so it's not like young coaches are entirely new.) 45-year-old Mike Benevides in B.C. and 36-year-old Corey Chamblin in Saskatchewan aren't exactly dinosaurs, either. We'll see if the new breed of coaches gets more aggressive, or if they carry on with the traditional conservative patterns.