Monday, January 5, 2009

The Terrible NFL AND College Overtime Rules

This is yet another area (along with "playoffs") where people confuse fairness with excitement. Many people think that the NFL overtime is unfair because there is no guaranty that both teams will get the ball. Take Peter King:

The overtime rule continues to be the dumbest, stupidest, most indefensible rule the NFL has on its books. Giving a coin flip more power than Tony Soprano has now deprived us of a satisfactory ending to two pivotal games this year -- Jets-Patriots in Week 11, when the Patriots and 401-yard passer Matt Cassel never saw the ball in overtime after a heroic fourth-quarter comeback, and Colts-Chargers, when we didn't get to see the NFL MVP even play in the fifth quarter because it was a one-possession overtime.

It is true that winning the coin toss gives the receiving team an advantage. Since kickoffs were moved back to the 30-yard-line, receiving teams win overtime games about 60% of the time. Prior to the move it was about 51%. As you can see, winning the coin flip is a pretty big deal.

The problem is that people are idiots about college overtime where winning the coin toss matters nearly as much. (Note: One of the consequences of the focus on the NFL's overtime is that it is very difficult to find stats about how the coin toss winner fares in college overtime. If anyone finds a reliable stat, please pass it along). The team that wins the coin toss in college always elects (properly) to go on defense first. The reason for doing so is that it gives your offense an informational advantage, as they will know exactly how much they need to score. If they only need a FG, they can play conservatively and minimize the risk of a turnover. If they need a TD they can repeatedly go for it on 4th down, something that you would never see if they had no such advantage.

The college overtime is exciting, but it's incredibly unfair.

The best possible overtime has been put forth repeatedly by our friend Michael David Smith first here, and then again today:

The solution is this: Let one team determine where the overtime kickoff will take place, and then let the other team choose whether to kick or receive.

I've been arguing this for years, and my idea hasn't gone anywhere, but I'm going to keep at it. They don't even need to have a coin flip. Just say that the road team picks a yard line, and the home team picks whether to kick or receive. Right now, the overtime kickoff is at the 30-yard line, which benefits the receiving team. But what if they moved the kickoff by 20 yards, to the 50-yard line? Or by 40 yards, all the way to the other 30-yard line? All of a sudden, receiving wouldn't be an advantage anymore.

The NFL, of course, will never implement this rule. But it should. A coin toss followed by the words "game over" is no way to conclude a playoff game.

This idea should be implemented tomorrow. No one can claim that it's unfair. If you want the ball first, you can get a price.

Just remember, the next time you here someone claim that the college overtime is better because it's not "decided on a coinflip" that it is, in fact, decided on a coinflip. The auction is the way to go.

Update: More here:

The win probability is almost the same in each period: a little over 0.52. To put this in perspective, it's a smaller advantage than a team would have if they could start a game with a 1-point lead (0.53 according to the Dynamic Programming Model), and is a much smaller advantage than the coin-toss winner has in the NFL's current sudden-death format, which we estimate to be 0.57. Still, a coach who mistakenly selects the first possession after winning the coin toss lowers his team's win probability by more than 0.04, which qualifies as a major coaching blunder. For those whose distaste for sudden death derives mainly from the importance of the coin toss, the NCAA format would not appear to be a good substitute, particularly when there are alternatives that truly render the coin toss irrelevant.


Jon said...

Why call people idiots for thinking college football OT is pretty fair when you don't even know if its fair or not (ie % of times coin toss winner wins game)?? I happen to like the college OT format from an enjoyability format, and HATE the NFL format. Anyway from a little research i've found that statistical models estimate the coin toss winner will win 52% of the time under the college system, assuming they choose to play defense first. To me thats close enough to 50% to pass any fairness threshold (assuming their math is correct). Not opposed to an auction type system but anything besides the NFL's current system - including the college system - would be a vast improvement.

PaulNoonan said...

The NFL system had the exact same percentage when kickoffs took place from the 35 yard line. And if the winner of the coin toss in college overtime did not have an advantage, we would see more variety in what teams decide to do. As it is, they all play defense first. I explained as much above. At any rate, both are unfair, and both are determined by coin toss.

And people ARE idiots about the college method. I just spent far too long googling college OT stats for reliable numbers and half of the articles that came up suggested going with the college method because the NFL was decided by the coin toss. More specifically, they are idiots for assuming that just because both teams get the ball on offense that it will be fair. It's simply not true.

The coin toss is not the problem. Coin tosses ARE fair. They are perfectly random and provide a 50/50 shot. The problem is the game that they have structured following the coin toss.

tracker said...

While the auction system may remove the coin flip in determination of the winner, it seems to me nothing more than added gimmickry that does little to determine which is the better football team.

The best solution: Play an entire 5th 15-minute period.

PaulNoonan said...

What? The auction system is PERFECTLY FAIR. That is its gimmick. The coin flip isn't the point. This game is fair.

Eric said...

Perfectly fair would be flipping a coin at the end of regulation to determine the winner. Why bother playing the game.

Scott H said...

I like the auction system. Teams would definitely find the most fair starting yardline to use; however, they could adjust for percieved strengths or weaknesses to date the opponent to let them have the ball first.
Of course, people bitched about the NFL rule back when the kickoff was from the 35 despite that the analyses I saw back then concluded that the 35 yard line NFL rule was closer to 50% than the college system.

steveegg said...

Counter-point - if a team can't stop the opposing team from marching 40 yards on the opening OT drive, that team probably wouldn't have won in the college system because that would've been 6-8 instead of 3. In fact, if the opening OT kickoff is returned for a TD, the team that got scored upon probably didn't deserve to win.

For those that bleat that both teams must have a chance, let's do a variant of the college system - each team starts at the opposing 25, no field goals, and only either 1-point or 2-point conversions after touchdowns. If you want to keep an element of sudden death (and ignore the fact that there aren't enough zebras for this), put both offenses on each end of the field, and the first team that scores a TD wins.

steveegg said...

For those that say that going back to the 35-yard line would bring back the balance, there is something fundamentally different in the game between that era and today - teams have become rather proficient at the 2-minute drill. Given that, and the higher-than-average "big play" that happens on the opening OT drive, the extra 5-8 yards gained by moving the kickoff back to the 35 won't appreciably change the bias toward the receiving team.

tracker said...

Point Eric. Who cares if it's perfectly fair, whether it be a coin flip or an auction system, which may be exciting, might produce an outcome in which receiver/kicker win 50% of the time, but does little to determine which team is better at football and much to determine which team is better at crapshoot plays, such as onside kicks from their opponents' 30-yard line.

I'd rather the elements of a football game determine which team is better at football. If at the end of an additional 15-minute period the teams are still tied, so be it: the game has fairly determined that on that day, the teams were equal. Coming up with some crappery just so one team can be called the winner and one the loser is what would be UNfair.

Jon said...

Meh, to me 52%/48% is "fair enough". of course teams will always take the strategy that gives them the 52%, and i'm not arguing this method is preferable to an auction method - but a 4% spread (52-48) is so much better than 20% (60-40) that i dont think its right to paint them as equally bad. college system is WAYYYYYY better than NFL system right now.

PaulNoonan said...

52% is only what William Krasker's dynamic programmign model predicts. He also predicts that NFL OT should produce a 57/43 split, not the 60/40 split that we see in real life. Someone has to have the actual percentage out there, I just can't find it.

With regard to Eric's comments, flipping a coin is fair, but invloves no skill. The auction still involves a skilled game, and removes the chance aspect from the equation.

Jib said...

I like Tracker's full fifth quarter, but with a twist. You play the full fifth quarter or until you get up on your opponent by 8 or more. It doesn't remove the one possession OT, but it would add a strategic element and make for some exciting two point plays.

Anonymous said...

Awesome! Wonderful topic, but will this really work?

Anonymous said...

Nice fill someone in on and this mail helped me alot in my college assignement. Say thank you you as your information.

Anonymous said...

Approvingly your article helped me truly much in my college assignment. Hats afar to you send, choice look forward for more cognate articles soon as its one of my choice subject-matter to read.

Anonymous said...

Auction system is not "perfectly" fair. Because you have to pick one team to set the yard line.

Assuming that statistics would eventually show which yard line is correct to set (and they would), then the team not setting the yard line would always get a very slight advantage - but an advantage nonetheless.

Assume the right yard line is the 35.6 yard line. The deciding team will offer the 36 yard line and the other team will decline or the 35 and the other team will accept. Either way they get a slight advantage and you would rather be the team NOT picking the yard line.