I'm not quite sure how or why it happened, but last week's examination of the NFL's overtime rules became one of the most viewed entries. Since then, I've been involved in a strangely high number of conversations - online and off - about the perceived "unfairness" of the current rules. After hearing a lot of arguments, I figured I would do some more digging and update my argument with a few thoughts that came to me later.
In the previous article, I mentioned that in 2010 and 2011 combined, teams that got the ball first in overtime had a game winning score on their opening possession just about 16% of the time. Some have accused me of being deceptive with my limited selection of years. I don't think I was being deceptive about it (in the original article, I did specifically state outright that I was citing a very limited sample size, so it's not like I made it a secret). So, naturally, the first thing I went and did was found data in more years.
For the sake of transparency: this information is not readily available. Turns out, no one else has really been collecting this data in one easily accessible spot. Googling, "how many OT games were won opening possession scores" doesn't yield anything particularly helpful. However, the information itself is out there. What I've had to do is painstakingly find all the games in a season that went to overtime (by looking at the scores week by week), then read the play-by-play logs to find out which OT games were won on an opening possession. If you've never tried collecting data like that, let me assure you it's extraordinarily tedious and time consuming. I'm even going to mention that you have to take these stats with an error margin of maybe +/- 2 games total. (I think, for example, I originally claimed that teams that got the ball first in 2014 and 2015 went 15-23. They didn't. They went 15-22. I apparently re-counted a game. Staring at a computer screen and reading tons of tiny text can make it kind of easy to mess up here and there.)
Point being, there's a reason my sampling sizes are not that large. First step, then, was double checking the four years I had already tracked, and then counting more. I've gathered the stats for all four years of the new OT era, and I went back seven years - to 2005 - with the old rules. Not because I'm trying to cherry-pick. Indeed, I fully plan to keep digging at another time. It's just that this has been such a time consuming process that I need a break. Still, seven years of data is enough to give us a bit of a clearer picture.
Here is what I've found:
Teams that got the ball first in overtime from 2005 to 2011 had a combined record of 49-52-1. However, out of those 102 overtime games, only 31 were won on an opening possession. That means that in the span of seven years, the game was over after a single possession 29.5% of the time.
Comparatively, from 2012 to 2015, teams that got the ball first in overtime had a combined record of 37-32-3. Out of 72 overtime games during that span, the game was won on the first possession score, or the team that got the ball first scored and the other team failed (functionally the same as sudden death) 23 times. That's about 31.9% of the time.
Basically, a team that gets the ball first and scores on their opening drive is functionally the same. In four years, there has only been one occasion in which a team scored on their opening drive, then wound up losing the game (the Colts lost to the Panthers in 2015 after getting the ball first and scoring a field goal). 21 out of 23 teams that scored on their opening possession in overtime win the game. (One time resulted in a tie - more on that in a moment.)
Well, ok, what about games that were won on an opening possession touchdown? From 2012 to 2015, that's happened 20 times, out of 72 overtime games. that's about 28% of the time. It makes sense that we'd see a jump in this stat given that the new rules fundamentally shift the strategy of coaches, offenses, and special teams now that a field goal doesn't guarantee anything (even though it kind of effectively does). One might expect to see a rather substantial dip when looking back at this stat when looking at the original rules. Yet from 2005 to 2011, overtime games were won an an opening possession touchdown 24 out of 102 times - or 23.5% of the time.
Yes, yes, I understand that results don't equate to fairness. But this further backs my argument that this idea that all you have to do, really, was win the coin toss was complete nonsense. Football is a little more complex than that. Games won on opening possessions has been almost identical in the first four years of new OT rules as they were in the last seven years of old ones. The frequency a game is won on an opening touchdown isn't particularly different either.
So what was the point of these overtime rules?
I suppose in defense of the anti-old rules and "win the coin toss, win the game" fighters, it's worthwhile to examine years independent of each other too. Certainly when you look at the years individually, you find that there is no real pattern. Below is the breakdown of percentage that games with the old rules were won an the opening possession, without the other team's offense seeing the ball:
2005 - (31%)
2006 - (41.6%)
2007 - (40%)
2008 - (50%)
2009 - (40%)
2010 - (11%)
2011 - (27%)
At first glance, one might say that there is something of a commonality in that three of the four years saw it occur 40% of the time. One year even saw half of all overtime games end on the first possession! And yet, there's also 2010 and 2011, and 2005 only saw this happen less than a third of the time. I'd guess that if we went back further, we'd see something similar in that the success rate fluctuates pretty wildly at times.
Regardless, none of those statistics indicates that this was something happening all the time, like everyone suggests. Even 2008, the year most favorable to the team that gets first possession, it only happened half of the time. Hardly enough to indict the rules. And oh yeah, the overwhelming majority of overtime games today are still won on field goals.
However, there's one other stat I'd like to call attention to. Did you notice that in the last seven years of the old rules, there was just one tie? Did you happen to catch that in the first four years of the new overtime rules, there have been three? 2015 is the first year of the new rules that didn't have a game end in a tie, but there were a few close calls. This leads me to a few additional thoughts about overtime rules from a design perspective.
Take a moment to consider what the function of overtime is. The entire point of overtime is to determine a winner. The whole "fair" argument is really strange to me since it throws out the fact that there was an entire sixty minutes of regulation in which both teams had plenty of opportunities and chances to legitimately win the game. When it's still tied, the game slightly alters one rule to help determine a winner when a full game wasn't enough. All it changes is that the first team to score wins. That's it! The game plays out almost exactly the way it would in regulation. The only strategical difference is that coaches will call super conservative plays deep in the opponent's territory, knowing they don't need to try for the touchdown.
Can you recognize what that rule change does, from a game perspective? It practically guarantees a winner. Sure, there have been those few games that still end in a tie, even with sudden death rules, but the rest of the time, there is always a winner as long as someone scores. The sudden death rules decrease the chances of a tie.
Here's what the new rules functionally do: by allowing the other team to have an offensive possession if the first team scores a field goal, the rules actually increase the chances of a tie. Here's the perfect example: a couple of years ago, the Vikings and Packers went to overtime. The Packers got it first and scored a field goal. The Vikings got a possession, by rule, and proceeded to score a field goal of their own. With a normal fifteen minute period, that doesn't leave a lot of room for additional possessions. There was about six minutes left in the period when the Packers got the ball back for their second possession. Plenty of time to kick a game winning field goal, but when teams have to drive 60 to 70 yards down field for a high percentage chance, it might take a little more.
In this case, the new overtime rules directly led to a game ending in a tie, whereas with the old rules, there would have been a winner (and it would have been the Packers, who conveniently are always complaining about OT rules since they keep losing in the post season - seriously, the Packers have lost their last four playoff games that went to overtime). What's more is this same situation happened with the Bengals and Panthers! That game would have had a winner if not for the new overtime rules. These situations highlight how the new rules can sometimes defy their purpose.
From a game design perspective, you generally don't want a set of rules that can off-set the function of them. If the overtime rules are meant to determine a winner, a system of rules that increases the chances of ties defeats the purpose. (We almost had this same situation happen again this year, by the way, when the Colts scored on their opening possession, and the Panthers scored a tying field goal too. Fortunately, with about five minutes left, the Panthers defense got a takeaway to set up excellent field position for the game winning field goal.)
So ok, the stats don't really back the idea that the old rules really were so terrible, but they just felt wrong. The results don't mean anything about "fairness," it's argued. The whole premise that it's not fair is built on the conceit that the primary element of football is the offense. Even after the rule change in 2012, we still hear people argue that it isn't fair that the other team "didn't get the ball."
Fun fact about football: it's not a turn-based game. Sure, it's not a free-flowing game like soccer or basketball or hockey. And yes, we do like to make analogies to chess. Thing about chess though? It is turn-based. For a while, it was considered an unfair advantage for players who get the first move. After some time, players got really good at developing strategies that there are many who think going second is advantageous. Either way, the point is that chess, like checkers or othello, are purely turn-based games. When one player goes, the other player cannot do anything but watch.
Football does not actually work like that. It's maybe easy to think of offense as "getting a turn" with the ball, but it's not purely turn-based. When the other team has the ball - or has their "turn" - you have your defense on the field. You don't sit idly by while the other team moves the ball into field goal range. When it's not "your turn," you still have the opportunity to act in a way that can earn the win. Because both teams field players at the same time, there is nothing inherently "unfair" about it. If the concern were that both teams get a possession in overtime, then there should be equal focus given to ensuring both teams have an equal number of possessions during regulation too. How many times does a team win after getting the ball first in the game, then getting the ball and trying to kill the clock? This entire strategy is built around ensuring your team has one more possession than the other team. Is that any more "fair"?
The main reason why I don't like all of these rule changes is sort of a fundamental design thing for me. Given that some teams build around offense - like all of those Colts teams with Peyton Manning - and others build primarily around defense - like the Seahawks - it seems weird that the argument is that we need to ensure offenses get a chance. At the end of the day, we're trying to mold the rules around the teams, rather than having the teams build around the rules.
It's a highly questionable approach to design, and one that I don't think the league should get into. Even the idea of college rules - which I think is very entertaining - takes away special teams and some key coaching aspects. Field position is removed from the equation entirely, as are all of those decisions that make or break teams and coaches.
So yeah, I still don't think the rules ever had to change in the first place. The new rules are worse, even if slightly more entertaining. And I do worry a bit about what any more knee-jerk rule changes will do.
No comments:
Post a Comment