September 18, 2010
If things had turned out differently, Ken Miller might just have been channeling Herm Edwards in an angry post-game press conference last night. Miller, the head coach of the Saskatchewan Roughriders, made one of the most unusual coaching decisions you'll ever see in football yesterday. With the game tied at 37 late in the fourth quarter, Saskatchewan put together one last drive. Aided by two somewhat controversial calls that both Andy Fantuz and Darian Durant were down when they fumbled (on separate plays), the Riders drove the length of the field and appeared in excellent position to give Luca Congi a field-goal attempt from around 35 yards out to win it. Instead, punter Eddie Johnson (pictured above on the critical play) trotted out, throwing much of the American audience watching on the NFL Network into confusion about what was going on (and causing much of the Canadian audience to question Miller's sanity).
Basically, Miller elected to try and take advantage of one of Canadian football's more unique rules. Up here, a single point (or rouge) is awarded when the ball is kicked into or through the end zone and isn't brought back out by the defending team. Thus, if Johnson had been able to angle his punt properly and have it enter the end zone and then go out across the sideline, or if he'd hit it hard enough to go through the back of the end zone, Saskatchewan would have won 38-37. That wasn't what happened, though; Johnson's punt had enough power to hit the sideline but lacked the proper angle, and it was returned out of the end zone by a Stampeders' player before the Riders could get to him, sending the game to overtime.
Saskatchewan eventually won in overtime, scoring a touchdown on their first possession thanks to a Wes Cates run. They missed the two-point conversion this year's rule changes forced them to attempt, but were able to shut down the Stampeders on their drive, eventually forcing Henry Burris to throw an incompletion on third and goal from the three. That gave the Roughriders a 43-37 victory.
There were a multitude of potential talking points coming out of this one, such as Fantuz (pictured at right making a catch against Milt Collins) picking up 255 receiving yards, the second-highest total in Roughrider history, Durant throwing for 500 yards, or even the simple fact that Saskatchewan came off a 31-2 loss last week to beat the 9-2 Stampeders, who still hold the league's best record. Despite those, most of the post-game chatter so far has focused on Miller's decision to attempt a punt single instead of a field goal on the final play of the fourth quarter.
It was a bold call, and an unusual one. In most cases like that, you'll see the team try for the field goal. If they miss it, they still have a chance at the rouge if the kicker hits the ball hard enough to go through the end zone or if the cover team can bottle up the returner in his own end zone. A punt single attempt leaves little margin for error; either it goes through or it doesn't. In this case, it didn't, and that led to everyone piling on Miller for making the call.
There's been plenty of criticism of Miller's decision already, with everyone from legendary Vancouver Sun columnist Cam Cole to The Edmonton Journal's Mario Annicchiarico to The Calgary Herald's Allen Cameron to Toronto Argonauts' offensive lineman Rob Murphy weighing in against it. Even the CBC recap calls it "an ill-advised decision". Miller accepts that it was controversial, and told TSN's Sara Orlesky he was "about ready to figure out a way to get out of town before daylight". However, he said he'd make the same decision again.
Q: ‘Would you do it again?'
A: ‘Why?... Again because 99 times out of 100 he's going to kick the ball through the endzone for the point with no time on the clock and you don't run the risk of the field goal going off the pipe and that sort of thing. And there is only one person who handles the punt as opposed to snap and hold and kick and that procedure so I'd do it again.'"
This is definitely the minority opinion on the matter, but I think Miller actually made the right call. 42-yard field goals are anything but a gimme, particularly in a windy venue like Regina's Mosaic Stadium. Moreover, Congi has made 22 of his 27 field goal attempts this year (81.5 per cent). That isn't bad, and is actually above his career average of 80.2 per cent, but it still represents just above a four out of five chance of making any given field goal. This was more difficult than the typical field goal, considering the conditions, distance and situation. As Miller points out, field goals also involve the chance of something going wrong on the hold, or the ball going off the uprights. Keep in mind that the last Saskatchewan kicker to miss a crucial field goal had his house egged, his family threatened and a load of manure dumped on his neighbour's driveway (in a case of mistaken identity) before he was run out of town. That's a lot of pressure riding on Congi's shoulders, and Miller removed that by trying something different.
In my mind, Johnson's attempt at a punt single actually had a better chance at success. He certainly had the leg to get it into the end zone from about that distance, as his punt was caught about halfway through the end zone. Nye makes an interesting point that there was actually good reason to believe Johnson could boot it through the end zone and out the back: "When the Riders had the wind in the second quarter Johnson had punts of 53, 56 and 45 yards. In the fourth quarter, with the wind, his only punt was 40 yards, which was the fateful play".
However, even a 40-yard punt provided an excellent chance to score. In order to notch a single on a punt of that length, Johnson would have had to angle it more sharply towards the side of the end zone and have it go out of bounds, but good punters can do that pretty regularly. It's no different than the typical coffin-corner punt you usually try for from that kind of field position, except you just move the target back about five to 10 yards. Johnson had already recorded a punt single earlier in the game (except that was probably by accident in an attempt to hit the coffin corner), and he'd also recorded a kickoff single. He's been a very solid punter this year, averaging a gross of 43.4 yards per punt (below his career average, but that's partly thanks to Saskatchewan often punting deep in enemy territory) with an excellent net average of 38.6 yards per punt. That's above his career net average of 38.2 yards per punt and speaks to his increasing skill at directional punting.
I don't have stats on hand to back this up, as this kind of play really doesn't happen very often, but I'd imagine having Johnson go for the punt single on this play actually would succeed more frequently than a field goal attempt. As I wrote above, Congi's chances of hitting any given field goal this year are just above four out of every five attempts, so almost 20 per cent of his attempts miss, and this was a much tougher situation than the typical field goal, increasing the difficulty. A missed field goal from that distance isn't likely to result in a single. By contrast, if you had Johnson try to execute this punt 100 times or even just 10 times, I firmly believe he'd pick up the single more than 80 per cent of the time. That makes calling for the punt the percentage-based move. It isn't the usual strategy, so it gets more criticism when it fails, but you have to keep in mind that a missed field goal would have produced exactly the same results, just without the criticism thanks to its more typical nature.
This has happened at all levels of football with different plays. Gregg Easterbrook of ESPN and many other analysts have argued that American teams should go for it on fourth down more thanks to the probabilities of success. However, failed fourth-down gambles have more of a tendency to stick in the popular memory than the safer move of punting the ball away. For all the time they spend around football, NFL coaches actually spend a surprisingly small amount of time in game situations, much less than hardcore Madden players. Many of the decisions they make in games aren't necessarily based on experience or evaluation of the appropriate probabilities, but rather convention and the desire to avoid criticism for trying something different.
As a case in point, one of the biggest storylines of last year's NFL season was Bill Belichick's call to go for it on fourth-and-two against the Colts instead of punting. The play failed, and Belichick was roundly roasted for defying convention. However, several prominent analysts such as Shutdown Corner's Matthew J. Darnell, Advanced NFL Stats' Brian Burke and Smart Football's Chris Brown looked at the numbers and suggested that Belichick made the right move from a probability standpoint despite its negative outcome. Still, Belichick's move made news because so few coaches are willing to try something like that despite the chances of success, and the wake of criticism he received suggests that isn't likely to change any time soon.
There's no dispute that Miller's decision also produced a negative result, but that doesn't make it the wrong decision. For an example of what I'm talking about, let's look at baseball. One of the game's best hitters year in and year out is Ichiro Suzuki of the Seattle Mariners. He's having a bit of a down year by his standards, but still has a batting average of .310 and an on-base percentage of .357. By contrast, his team is on pace to have perhaps the worst offensive season since 1972. One of the chief regular culprits is first baseman Casey Kotchman, who is batting .221 and gets on base at a .284 clip. Ichiro is obviously the better hitter if you look at the whole season, but that isn't necessarily apparent in smaller sample sizes.
For example, consider the Mariners' August 29 game against Minnesota this year. Ichiro batted leadoff, as usual, and went an uncharacteristic 0-for-4 at the plate, giving him a .000 batting average on the day. Kotchman batted fifth and collected one hit in three at-bats, giving him a .333 batting average. If you were drawing conclusions based solely on this game, you might think that Kotchman is a better hitter than Ichiro and should bat higher in the lineup to get more at-bats. In fact, you'd conclude that Ichiro is one of the team's worst hitters and should either be benched or move to the bottom of the order. The correct percentage-based decision is to have Ichiro hitting first, but that isn't apparent from that game.
It's the same thing with Miller's decision to try for the punt single. I don't have clear percentages to back it up (and I'd feel much more comfortable arguing this if I had hard data either way), but I bet if you simulated it a sufficient number of times, you'd find that going for the punt single produced more wins than simply trying the field goal. I could certainly be wrong on that, but that's also difficult to prove due to a lack of data. However, the idea that trying the field goal is the correct move isn't based on data at all as far as I can tell, but is rather based on it being the expected move.
Speaking of expected moves, there was yet a third option that might have had even a higher chance of success. The surest way to score a punt single is to kick the ball straight through the back of the end zone. It's worth remembering that any player can punt the ball at any time in the CFL. Saskatchewan could have lined up for a field goal as everyone would have expected, but have either Fantuz or Rob Bagg at one end of the formation. They release as on a typical fake field goal and go about 10 yards downfield, Durant fakes the hold and hits them with a pass, and then they run as close as they can get to the end zone before punting the ball through it. A fake would have been completely unexpected, as the field goal was the move everyone anticipated, and I'm pretty sure an athletic receiver can punt the ball through an end zone with ease from 20 or 25 yards out. It's unusual enough that I doubt anyone would ever try it, and it does have complicating factors (the route has to work, as does the pass and the eventual punt, and the element of surprise is crucial), but it's worth keeping in mind as yet another alternative to the traditional field-goal attempt.
As humans, we have a tendency to stick within our comfort zone and to go with the familiar. That doesn't mean the familiar is best, though. In this situation, I think Miller made the right call based on the expected chances of success. That didn't work out, though, and so he'll face plenty of questions and criticism for challenging the status quo of how things are done in football. In my mind, that's unfortunate; praise or criticism for a football decision should be doled out not solely based on the result of the play, but also on the chances of its success.