Proposed college football rule change: when the defense commits a penalty after the snap in the final two minutes of a half and the offense accepts that penalty, officials enforce the yardage as usual and add up to ten seconds– but not more than the amount of time remaining in the half when the ball was snapped– to the game clock.

This sounds like a convoluted scenario, but it has real relevance. The not-all-that-uncommon scenario that inspired this proposal occurred at the end of the first half of Saturday’s Michigan-Michigan State game:

With nine seconds left in the first half, Michigan had the ball on MSU’s 20 without any timeouts left.

What play do you call on first down? A fade to the corner of course, and then kick a FG if you don’t make it, right?

Wrong.

It isn’t really possible to complete a pass in this scenario if the defense knows what it is doing.

Michigan called that pass play, though. The receiver had a step on the defender and the defender committed fairly blatant pass interference to save a touchdown, which is fine. Result: automatic first down (irrelevant), fifteen yards (moving from the twenty to the five doesn’t change too much), and the clock is down to four seconds. Michigan then kicked a field goal. While the field-goal try is somewhat easier following the enforcement of the penalty, the net benefit remains on the defensive side.

Here’s a quick illustration:

Let’s say that the fifteen-yard penalty improves the probability of a made field goal from 60% to 85%. Let’s also assume that if the defender doesn’t commit pass interference, there is a 70% chance the offense completes the pass.

The net improvement in FG% is considerable. The penalty increases the expected number of points by 0.75 (0.6 x 3 = 1.8 to 0.85 x 3 = 2.55), which is quite a bit.

That doesn’t tell the whole story though, because the scenarios are more complicated than that. Let’s do an easy conditional probability example (statistics can be fun, I promise).

If the defender doesn’t commit pass interference, the offense gets 0.7 x 7 + 0.3 x 0.6 x 3 = 5.44 points for the offense. If the defender commits pass interference, the offense gets 0.85 x 3 = 2.55 points. Committing the penalty thus results in a net swing of 2.89 points for the defense, which feels unacceptably high to me. Even if you fiddle with the numbers a bit, you’ll get a similar result: a considerable number of expected points gained by committing pass interference (even including the yardage benefit of the penalty enforcement).

Moreover, assuming the post-penalty improvement in FG% referenced above (60% -> 85%), the pass play would have to have < 15% chance of success in order to break even, meaning that, if the defender thinks that there is even a 15% chance of the receiver coming down with the ball, then he should commit pass interference. We can reasonably expect most teams to do better than 1-7 on fade routes when the defense doesn’t commit any penalties.

The problem is that no matter how good the defender’s coverage, or how poorly the receiver ran his route, the defender should commit blatant pass interference 100% of the time in that scenario, and this is where we have a problem. There is no real way to complete a pass in this scenario unless the defense doesn’t realize what I have described or falls down. When it is always beneficial to commit a penalty, something is wrong with the rule book.

(There are other situations where it is common to intentionally commit a penalty, but they mostly fit within the game. The first is pass interference on a generic play. This situation is different for two important reasons. The first is that offense still ends up better than they were before the play started, and the second is that the defender doesn’t begin the play thinking that he is going to commit a penalty no matter what.

In fact, defensive guru Buddy Ryan, while serving as defensive coordinator for the Houston Oilers, actually drew up a play with too many men on the field for exactly this scenario. While it is a clever use of the rules, this loophole should be closed.

Another situation is delay of game on a short field punt, but that one is also different because defenses can always decline that penalty (although teams rarely do for some reason).

Something else that might need to be added to this list is holding on an intentional-safety-to-run-out-the-clock play. This apparently happened in the Colorado’s win over Stanford on Saturday, but I haven’t found any footage of it.)

I see two solutions to this problem. The first is to always put the ball at the spot of the foul (at the one yard line in the case of the endzone) instead of merely advancing it fifteen yards from the line of scrimmage. This gives the offense a more viable run option or a very easy field-goal try. Personally, I don’t like this rule at all, as it strongly favors the passing game at the expense of the running game, but that is a statement of personal taste and not one directed at closing loopholes in the rule book. Moreover, it still doesn’t give the offense a chance for a touchdown attempt followed by a field goal.

The other solution, and the one I prefer, is something similar to the ten-second runoff. Currently, in the last two minutes of a half, if there is a penalty on the offense that stops the clock when it otherwise wouldn’t have stopped, officials run ten seconds off the clock before allowing play to resume. In the scenario described here, the fair thing would be to let the offense attempt the play again following the play on which the penalty was committed. The proposed alternative, for plays in the last two minutes of a half, would be to put up to ten seconds– but no more than the amount of time on the clock when the ball was snapped– back on the clock after a defensive penalty. This ensures that the offense doesn’t end up in a worse position as a result of the defensive penalty.

The most likely criticism of this proposal I anticipate is that people will be opposed to rule changes that result in significant clock manipulation, but other rules already grant officials significant discretion in controlling the game clock near the end of halves, so my proposal is not as major as it may sound.

I realize that this proposal poses practical challenge for referees, and people probably will be concerned about extending a game (despite the fact that the last minute of halves frequently take forever anyways), but it seems to be the only fair way to mitigate the defensive incentive to cheat that doesn’t result in eliminating the offense’s ability to throw the ball in these possibly crucial situations.