Fair Is A Feeling: A Critical Look at the NFL's Overtime Rules

The NFL is in something of a no-win position when it comes to their overtime rules. For the longest time, they had very simple, straightforward rules to it, but fans complained. The noise got so bad that in 2012, the rules committee voted to change it. On the most part, the suddenly convoluted rules were well received. And yet here we are, again, with a team losing in overtime, in the post season, without ever possessing the ball, and people are complaining about the rules.

Frankly, the NFL is never going to pass overtime rules that please everyone. At the end of the day, sports fans are very emotionally-driven. And let's be real here, the Green Bay Packers wouldn't be saying anything about the perceived "fairness" of the rules, or arguing that they need to change if their defense had, I don't know, not given up an insane 75 yard catch and run by Larry Fitzgerald.

One thing that's really interesting about sports though is that at their core, they're just games. Most of these sports have a lot of crazy rules, but many of them - as complex as they sometimes are - actually do a lot to keep a balance. People don't often look at rules or rule changes in terms of game design.

For example, look at intentional grounding. A quarterback cannot throw the ball away when under pressure if he's in the pocket. If he gets out of the tackle box, however, he can. On the surface, this seems totally arbitrary. Why can't the quarterback throw the ball away when he's in the pocket, but if he moves a few yards to the left, he can? Well, balance. It may be tempting to view that as a rule designed simply to drive up sack stats and increase excitement, but it also creates balance between offense and defense, making it fair. Given the formations of offenses and defenses, all of the pressure comes in through the pocket. If the quarterback were allowed to simply throw it away without penalty while there, the system would lean in favor of one side, giving him an unfair advantage.  By outlawing throw aways within the pocket, it gives the defense a chance to succeed too, rather than giving the QB the unfair ability to just chuck the ball wherever at the last second to avoid losing yards in a sack.

Additionally, when the quarterback gets out of the pocket, he's essentially earned the right to throw it away. The defense has had a chance to get pressure, and didn't capitalize, thus now the quarterback can perform that action in a way that is fair. These seemingly pointless rules actually create a fairness wherein the defense has a chance to sack the quarterback, but the quarterback - once meeting a particular condition - is not forced to take a sack every time there's pressure. These kinds of rules help create a level systematic playing field. Of course, why the quarterback has to get the ball back to the line of scrimmage doesn't strike me as anything but arbitrary, but no one ever said all of these rules were well thought out game design elements.

Often times, people want changes when they feel they got screwed out of a rule, even though those changes can't change the past, and could very well hurt your own team later down the road. Here, I'm going to attempt to break down why, from a game design perspective, the current overtime rules are a hot mess that doesn't actually make the game more fair. Bear with me, this could get long-winded.

For those who are not sports fans, the former overtime rules in question were quite straightforward. If, at the end of regulation, the game is still tied, they functionally restart the game with sudden death rules. Meaning, first one to score wins. This essentially put the impetus on chance. It was often thought that whoever won the ensuing coin toss at the start of overtime would win the game. After all, field goal kickers have gotten so good that you have a great shot at winning if your offense can just drive to about the opponent's thirty yard line.

Fans got so annoyed with the idea that a field goal early in OT wins the game that the league was eventually confronted with the idea of altering it. The more positively received rules aren't the most difficult to grasp, but it takes a bit longer to explain. Follow me here: the game can only end in overtime on one possession if the offense that first receives the ball scores a touchdown on their first drive, or the defense manages to score a safety or returns a fumble or interception for a defensive touchdown. If the first offense scores a field goal, then the other team gets one possession to tie the game with a field goal of their own, or can win if they score a touchdown. If the first offense doesn't score, or they score a field goal and the second offense also scores a field goal on their possession, then the game defaults to sudden death rules and the first team to score any points in any fashion wins. Alternatively, if the first offense doesn't score at all, and on their ensuing possession, the second offense doesn't score either, then the game defaults to sudden death rules as well.






Did you follow that? If it sounds pretty complex, it's because it is. It's rather unnecessary too. Let's break down why these rules don't actually make overtime any more fair.

The problem with the original overtime rules, it's argued, is that it's skewed too much in favor of offenses and special teams. It's often posited that sudden death rules that include field goals puts the defenses at too much of a disadvantage. Ostensibly, it's unbalanced and "unfair" to the team that doesn't win the coin toss.

According to a super interesting breakdown from Football Outsiders, the average starting field position off of kickoff for most of the early aughts was about one's own 28 or 29 yard line. In 2010, NFL kickers connected on 74% of kicks between 40 and 50 yards. Let's use these parameters for a moment. The argument goes that for the offense that gets the ball first in old rules has to drive about 45 yards to have a 75% chance of winning off a long field goal, given the improved overall quality of modern kickers. With the current rules in place, the first offense could drive the same 45 yards and kick a long field goal, but then they'll have to play defense and stop the opposing team from doing the same thing or better.

First off, that isn't really "nothing." The idea that this system is imbalanced in favor of the offense and special teams is kind of absurd. Sure, the offense only has to drive about half the distance that it would need for a touchdown, but that's to specifically set up a long field goal, which would mean that it's not skewed particularly favorably for special teams. Kickers have gotten a lot better over the years, but most teams don't play overtime to set up a 40+ field goal, especially when you recall another rule: if a kicker misses a field goal, the opposing team gets the ball at the spot of the kick, not the line of scrimmage like usual. Meaning, if a team does play for a long 45 yard field goal to win, and misses, then you're giving the ball to your opponent with great field position. The second offense would only have to drive 35 to 40 yards for the same opportunity to win the game.

And, of course, for shorter field goals, the offense still has to drive a good chunk of the field. If a team starts on their own 30 yard line and hopes for a 30 yard field goal, they'll have to drive about 50 yards from to get into range at their opponent's 20 yard line. It shouldn't seem so unfair that a defense has to stop an offense from driving down literally half the field.

Secondly, and more significantly, it is disingenuous to suggest that the other team - the team that lost the coin toss - "didn't get a chance." They did. In such field goal winning situations, the other team had chances to reduce field positions or block the kick with their special teams, and had the opportunity to keep the opposing offense out of field goal range with their defense. Football is not just offense. The original overtime rules were surprisingly balanced to not show any real favor to the offense or the defense or the special teams. It was balanced. The defense has an opportunity to make a stop or get a turnover. The special teams has an opportunity to score/contain the opponent. And the offense has an opportunity to drive to score the game winning points.





Here's what the new rules did, if anything: it actually skewed it so that the defense has an advantage. They are no longer on the line for giving up large chunks of yardage. If a defense played poorly in overtime and allowed an opposing offense to get into field goal range, that was their fault. They'd lose the game for playing poorly. Now though, if a defense gives up a lot of yards and a field goal, their offense has one more opportunity to bust them out of trouble.

See, here's the other thing about the new rule changes: they were completely unnecessary. Apart from the fact that a team can still win without the opposing offense ever seeing the field (as happened in the Green Bay/Arizona game), there was another rule introduced that actually fundamentally made things better for the original rules, that instead, with the new OT rules, makes the new OT rules even more imbalanced than the original rules.

That major change is, of course, the kickoff spot. Sports are a bit different than most games in the sense that these aren't pieces moving around a board. These are human beings. As such, there are going to have to be concessions for the sake of player safety. For example, the new kickoff spot, mixed with banning the wedge formation and more than two players linking are dumb rules from a gameplay perspective, but make total sense when considering player safety.

And here's what the new kickoff spot does for overtime rules: it pits the first offense in worse field position. Instead of starting near their own 30 yard line, teams are regularly starting on their own 20 due to the huge increase in touchbacks. Plus, if for some reason they do get a return, there's a strong probability they'll be starting inside their own 20. What this does it makes the distance the offense needs to go for a potentially game winning field goal farther by about ten yards. A team needs to drive about 55 to 60 yards to get into reasonable field goal range now. In effect what this does is makes it more challenging for the offenses. So the new kick off rule hurts the offense and helps the defense, and the new overtime rules hurt the offense and helps the defense. These two rules, however separate they may be, have ultimately skewed the balance in favor of the team that is on defense. They are presented with a better opportunity.





What's kind of funny though is that what people thought about the original overtime rules and what they assume about the new ones are, frankly, without statistical merit. Now, admittedly I'm using a rather limited sampling size. Given that the main argument is that the original rules were skewed too heavily in the favor of the team that wins the coin toss (read: the offense that gets the ball first), and the new rules are supposed to balance it out, then presumably, looking at the data should show us that the team that gets the ball first should have been winning most of their games in the old rules, and the winning has been subsequently more even with the new rules. To check this, I meticulously looked through the game logs of four years: 2010 and 2011 (the last two years of the old OT rules), and then 2014 and 2015 (the two most recent years of the new OT rules).  The data suggests what I've always felt to be true: the mechanic wasn't broke, and there was never a reason to change them in the first place. Here, check it out:

Between 2010 and 2011 - the final two years of the old rules with the simple sudden death even by field goal stipulation - teams that got the first possession had a losing record of 15-23. That's a winning percentage of .394.  More than that,  of the 38 games that went to overtime, only 6 were won on the first possession. Yes, that's right. Only 15.8% of the time under the old rules did a team that got the ball first win the game on their first possession in a two year span.

Conversely, between 2014 and 2015 - the most recent two years of the new rules that are supposed to be more "fair," - teams that got the first possession had a winning record of 19-13. That's a winning percentage of .594. And again, more than that, of the 32 games that went to overtime, 14 were won by teams that scored a touchdown or scored a field goal on their opening possession, followed by their opponents failing to score (which is functionally the same result as the old overtime rules).  That means that in the new overtime rules, 43.8% of the time, the game is over because of scores on the opening possession.

Essentially, there is no statistical data that actually suggests that winning the coin toss and receiving the kick off for the first possession held any discernible competitive advantage under the old rules. Probably because it's football and teams have defenses and special teams. Contrary to popular opinion, it isn't only offense.






Of course, the most obvious counterargument here is that statistical correlation does not equal causation. Obviously, teams that get the ball first were not losing games because they got the ball first. The same argument, naturally, can be applied to the new overtime rules too. The Packers, for example, did not lose because they didn't get the ball first.

Bottom line: the new rules do virtually nothing to make the game "more fair." From a design perspective, the old rules were nearly perfect. The advantage of winning the coin toss was obviously not a problematically large advantage. The team that kicked off was given a chance to get the ball back (and at times, even win the game - there were at least three games I encountered in the old rules that were won on defensive touchdowns, and none of them were that Packers game when they beat the Seahawks through a pick six of Matt Hasselback way back when!). Functionally, it was fair to both teams. They were neutral rules. It didn't favor the offense. It didn't favor the special teams. And it didn't favor the defense. It provided equal significance to each phase of the game.

The new rules have made no difference, except that now, it gives a defense more leeway. Mixed with the new kickoff rules, it even gives the defense more of an advantage. In essence, the new rules are actually more unfair than they were before (even if I completely agree that it's not particularly "unfair" - however, the rules are slightly imbalanced, leaning more favorably on defenses now).

Yes, yes, I just made a rather confusing argument. While I believe the statistics prove that tales of the coin toss's significance in the original rules are dramatically exaggerated, I don't believe statistics are everything. I'm not a total idiot and know I just made an argument that the new rules should be helping the defense, but the statistics show that more overtime games are won on first possessions than ever, despite it being harder.

They're sort of separate arguments though. As I said, the statistical correlation does not mean causation. The statistics argument was used to highlight that the old rules were not giving a team any particular unfair advantage in terms of results. Examining how the mechanics work in detail, though, highlights why they actually do - as a system of rules - give a slight advantage to the defense. Obviously, one can quite simply look at the pro-offense statistics of 2014 and 2015 to say that the new rules clearly don't give defenses a big advantage, but I hope the fact that I also examined how the mechanics actually work on a game-level, and not just on a results basis, helps clear up that I'm arguing in defense of the old rules and suggesting the new rules are really only slightly more unfair on a mechanical level.



What was the real cause of concern to begin with? Is the problem really that it's "unfair" the other team doesn't get a possession? If that were true, the new overtime rules don't actually do much to change that, since a team can still be robbed of a possession if the first team scores a touchdown. If that were the argument, then by all means, the NFL should adopt college rules, where each team gets a possession from the 20 yard line and is given a chance to score - playing a bit more like penalty kicks, if penalty kicks weren't a load of nonsense that is extremely favorable to the kicker and extremely unfair to the goalkeeper.  (While I don't necessarily think college OT rules are "more fair," I do think that it is a more entertaining system at any rate.)

If the problem people had was that a field goal on an opening possession would win the game (something that didn't happen nearly as often as people seem to think it did), that might be a fair point. Kickers are more accurate than they've ever been. The league is currently altering PAT rules to make the most boring and unnecessary play in all of football even vaguely relevant, since kickers and their extra point attempts were converted at an absurdly high rate. Seriously, it was basically 100% of the time. Even more, accuracy at a distance has also jumped pretty substantially over the last decade and a half.

Admittedly, I'm one of the few football fans who likes kickers and punters. I've always found them an intriguing part of a football game. They're far more a part of a team's game strategy than people ever give them credit for, and while people may find kicks boring, I find the long kick with the game on the line one of the most exciting plays in football. Few football plays have the same kind of anticipation and anxiety-inducing build up as the field goal attempt. Adam Vinatieri lining up in the snow against the Raiders or against the Rams or Panthers with no time left was the football equivalent of a batter stepping up to the plate with a full count, bases loaded, bottom of the ninth down by two.

However, I can concede that kickers are too accurate now. When they're as great as they've been, even long field goals are a bit boring. If this is the problem though, then the solution for overtime isn't to change the overtime rules into this super complicated thing that takes two paragraphs to explain. Instead, the thing that should really have been addressed was increasing the difficulty of field goals. How about instead of the convoluted overtime rules, they make the field goal cross bar higher and the side posts narrower. Make it so kickers need to be even more accurate. If field goals aren't automatic, then no one can really complain about a field goal winning a game in sudden death overtime.

Or, alternatively, the league could examine the possibility of altering point values. Field goals could function a bit like shots in basketball where at a certain range, they're worth two points, at another range, they're worth three, and at 50+ yards, they're worth four. (It's really not that hard or more complicated to keep track of this stuff...) Or, we could abolish the extra point and only do two-point conversions, as starting from the five yard line to make it something that isn't automatic. (11 teams that attempted at least one two-point conversion in 2015 had a conversion rate of under 50%. 15 teams had 50% or higher. 4 teams didn't attempt one at all. Point being, successful conversions are still a bit skewed favorably for the offense. Setting it back to the five would presumably give the defense more of a shot.)

Point being, looking at adjusting point values would help reduce the chances of a game going into overtime entirely. Yes, yes, I know, I know. "But tradition!" "A field goal has always been worth three!" "You can't abolish the dumbest, most pointless, and least exciting play in all of sports! (the PAT)" But the thing is, in game design, you typically don't just accept a points system that allows for a bunch of ties to keep happening. This suggestion, by the bye, isn't entirely necessary (even if I think it's a good solution). All in all, 38 overtime games in 2010 and 2011 is still a rather small percentage. Given there are 267 games in an NFL season, 38 OT games in two seasons (534 games) means that only about 7% of those games went to overtime. Is that a large enough percentage to merit rule changes? Up for debate, really (I'd say yes, but others might disagree).






Look, Packers fans have been on the wrong side of overtime post season games two years in a row. I get it. It is kind of anticlimactic to have one score at the beginning of a new period conclude the game. Frankly, if we want complete fairness, the NFL post season overtime rules should be a full, ten minute period that plays out in its entirety. No sudden death at this point. If it it's still tied at the end, then you play another fifteen minute period that is completely sudden death. No one can complain about "fairness" when you've played basically five periods of football! At some point, you just have to accept that the other team made more plays when you lose.

It's just funny to me that people always want to change rules that cost them that year, without examining what its impact might be. We changed a system that wasn't broken to one that is slightly imbalanced, just because a few times at key moments in the season, a field goal won a game in sudden death. From a design perspective, the original rules were better, cleaner design. From an entertainment perspective, it was virtually the same.

All in all, there's no need to change the current rules unless we are aiming to simplify them. As it happens, I prefer my rules simple and straight forward, so I would support changing it back to simple sudden death, or just a full period followed by sudden death if still tied. But at the end of the day, the current rules - as unnecessary as they are - are every bit as fair as the originals.

And yes, the original rules were totally fair too.



UPDATE:

Done a little more digging. For the sake of transparency, the reason I went with two years on each end for the OT victory stats was that I had to go through all the game logs myself. There wasn't really a central source of the information. It was a tedious and time consuming process.

I've done a little more digging and included three additional years. It does appear, with a larger sampling size, that winning on the opening possession happens more frequently than it did in 2010 and 2011. Hard to say without doing far more intensive digging if those two years were just weird outliers, but even in 2007, 2008, and 2009, it still wasn't a common enough to feel like it was hugely unfair.

In those three years, the winning percentage does skyrocket compared to 2010 and 2011. Teams that got the first possession were 30-15-1 in those years.

But what's funny is, 21 of those 45 games were won on the opening possession. That's 45%, still less than half the time. While that paints a different picture than 2010 and 2011, it hardly gives a portrait of a horribly unfair system. If anything, it still comes off as rather fair.

Looking at individual years also shows that there isn't anything universal. It sometimes swings wildly. For example, if you were to look at 2009, teams that got the first possession were far more successful than than the same teams in 2007, a year that fits the norms found in 2010 and 2011. Point here being, there doesn't appear to much of a consistent picture being painted, and that the success of teams getting the first possession vary wildly by year. None of it really suggests that the rules were the problem.