What Goes Into Luck?

I was struck by OCC's article asking Who Were The Luckiest NFL Teams In 2011? I encourage you all to read the article if you haven't already. The short answer is that comparing predictions based on a teams points scored and points allowed versus actual wins and losses indicates that teams like the Packers, Chiefs, and Broncos were fortunate in that they amassed more wins than their on-field performance would predict. It also suggests that, conversely, teams like the Vikings, Eagles, and Dolphins were not in Lady Luck's good graces this season. Our Fair Cowboys were in the middle of the pack, as the 11th unluckiest team with a -0.6 variance. But beyond 2011, one of the other statistics OCC mentioned really caught my eye.

For most teams, these numbers tend to change from year to year. But not for all. The Cowboys have had a negative variance for the last three years in a row. No team has had more successive years of 'bad luck', with the Texans and Panthers the only other teams to also have had three consecutive years with a negative variance.

On a personal level, I have long felt that the Giants are the luckiest teams in sports and 2011 was certainly no exception. The Giants finished with a positive variance for the seventh year in a row. The next closest teams are the Cardinals with four consecutive years and the Saints with three consecutive years. Only once in the last ten years (2004) did the Giants have a negative variance. The Giants are lucky on a metaphysical level that transcends rational numbers.

These statistics raise an interesting question. Is there anything more to these trends in "luck" than just random chance? Is there a reason why the Cowboys have been so consistently unlucky according to this formula when the Little Boys Blue have been so fortunate?

Here are five possible answers to that question.

1. No There Isn't

This is the easiest answer. The safe bet is to say that luck, even in the NFL, is random and lacks a good explanation for the resulting disparities how it's dispersed throughout pro football. Things like turnovers, game-changing plays, or those few moments in the season that seem to make or break a team, could be wholly unpredictable with no good way to account for them beyond the fickle mistress of random chance.

The Giants seven-to-ten years of consistent "good luck" could simply come down to that old brain teaser – if you flip a coin ten times and each time it lands on heads, what are the chances that it will land on tails the next flip? The answer is that the odds are still 50/50. Even with random stimuli like a fair coin flip, the illusion of patterns can emerge like a succession the coin landing on heads. While those results seem like a trend, they are just outliers and the apparent consistency is belied by the complete randomness behind them.

In other words, the Giants may just be lucky when it comes to luck. Through no deliberate actions of their own, they may simply be the unwitting beneficiaries of favorable bounces and good fortune. Yet, on this theory, they do not do anything to "make" their own luck so to speak. Despite their history of favorable fortune over the last seven years, on this account the Giants' chances of good luck in 2012 would be no better than the Cowboys.

Again, this is the safe answer, but I'm tempted, however recklessly, to think there's something more to the story.

2. Leadership

There's been a great deal of talk about leadership at Valley Ranch lately in the wake of Jason Hatcher's comments that it's been lacking on the team. Could having strong leaders be the difference between outperforming your points scored/point allowed prediction and finding your team falling short? BTB's own Dave Halprin chimed in on the intersection of leadership and luck:

Take the New York Giants. The Super Bowl winning New York Giants. Today, it would be easy for the people who follow that team closely to say, this guy or that guy was a real leader and they helped us win the Super Bowl. Okay, but just imagine if Tony Romo had connected on that pass to Miles Austin in that first Giants game. Imagine Austin caught it and scored like he probably would have happened 99 out of 100 times in that exact same scenario. The Cowboys in all likelihood would have knocked the Giants out of the playoffs. No Super Bowl, no "leaders" who got them there. No nothing. Yet, the players who are acknowledged as "leaders" post Super Bowl would have been doing the very same thing all season long. Yet their efforts might go unnoticed because the Giants would have been just another also-ran.

I'm tempted to agree with Dave. It's hard to forget Tiki Barber criticizing Eli Manning's leadership ability's after Tiki abandoned ship, only for the Giants to turn around and win the Superbowl. Though Eli obviously has to contend with comparisons to his brother, he's often been criticized for not being the "take charge" sort of leader. My G-men loving friends, lucky gents though they are, have often voiced to me the same sort of frustrations that Hatcher talked about. This dirth has not seemed to have slowed New York down too much.

That's why I am skeptical as to whether leadership, or at least one particular style of leadership, can explain the difference. Eli's certainly more of a "lead by example" type of player than a fire and brimstone man like Ray Lewis. Tony Romo, Jason Witten, and DeMarcus Ware's similar leadership style should not and does not distinguish the Cowboys poor fortunes over the last few years by comparison.

3. Being Aware of the Randomness

Given the "Any Given Sunday" nature of the NFL, most team's fortunes are going to be shifting up and down in the course of a given season. In an ironic twist, perhaps part of what helps a team improve their "luck" is being aware of the effect luck can have.

If you're the Giants, you know plenty about instability and shifting fortunes. In 2007, a 10-6 record was not only enough to net them the fifth seed in a weak NFC playoff field, but it was enough to point them to a Superbowl win. A few season's later, a 10-6 record isn't enough to even make the playoffs. The Giants' best season in terms of pure record over the last five years was a 12-4 finish in 2008, but all that gave them was an early exit against Philly. Meanwhile, this year they narrowly inch their way into the playoffs at 9-7 and make another Superbowl run.

Tom Coughlin has been on the hot seat so much that he's switched to insulated trousers. Despite his recent clutch performances, Eli is known for being brilliant one week and erratic the next. With such a wild ride, how could you not build up a tolerance to the changing tide? The concept of "variance" is big in the poker world, and it can be oversimplified down to the following. The statistics of the game say that even when employing the best strategy, sometimes you're going to be ahead and sometimes you're going to be behind. The idea is to trust in your process and not react too swiftly to those hills and valleys along the way.

My personal theory on the Cowboys team meltdown in 2010 is that they had a rash of bad luck early–a much of the season. There were certainly a myriad of factors at play, but in that season and many others, the fans have seen Dallas seemingly become disheartened after a string of ups-and-downs. There's been a great deal of turnover in both organizations, but has their been enough institutional consistency in Dallas and New York for the G-Men to have built up a better tolerance for the NFL's usual slate of variance than the Cowboys? Do they persevere and not let their bad runs get them down or their winning stretches breed complacency?

It's an attractive theory, but I fear that it may prove too much. Any player or coach who has spent significant time in the pros knows how quickly a team's successes can turn to failures and vice versa. With the persistent turnover in division winners and playoff teams, any squad must have at least a passing familiarity with the fickle nature of the game. Certainly the Giants' path has been somewhat unique in how their largely similar finishes have led to vastly different results, but with variance a fact of life for every team across the league, it's difficult to say that the experience from their prior paths is enough to make the difference.

4. Discipline

Discipline is difficult to quantify. It's hard to categorize and measure things like mental mistakes, instances where players are out of formation, or when one individual missed an assignment. One admittedly incomplete method of measuring discipline, however, is looking at how many penalties a team has to overcome in a given season. I looked at rankings over the past five years for which NFL teams had the fewest overall penalties. The tale of the tape is fairly stark, and not at all pretty for Cowboys fans.

That's right, even in the Cowboys' best season in recent years, they were still ranked 25th in the league for penalties, and they averaged a rank of 28th over the past five seasons. In other words, over the past five years, in a given season the Cowboys would average more penalties than all but four teams in the NFL. By contrast, the Giants average rank is 15th, putting them right around the league average. In addition, both of their Super Bowl winning years were the ones where they had the fewest penalties. Could this disparity in penalties help explain the disparity in luck?

It may be part of the story, but there's one big problem with it. Shouldn't the luck formula implicitly take this sort of statistic into account? If an offense continues to commit penalties, wouldn't that limit the number of points they score, and likewise woudln't a defense committing penalties increase their points allowed? I believe the Cowboys' penalty problems are worthy of a more focused look, and that they have a great deal to do with the team's ceiling over the past half-decade, but it's not clear that it affected their "luck."

On the other hand, as I mentioned, discipline is difficult to measure. It might be that part of what makes up the discipline of a team is not captured, either explicitly or implicitly, in points scored or points allowed, and that can change a team's luck beyond the bounce of the ball.

5. Mental Fortitude

I harped on it at the beginning of the season. I harped on it during the season. I harped on it at the end of the season. One of the biggest problems for the Dallas Cowboys since the Wade Phillips era has been that the team cracks under pressure. They do not play well from behind and they have even more trouble cutting off a furious comeback. It's continued in the Jason Garrett administration, given how many fourth quarter meltdowns the fans watched this year.

The Cowboys played an inordinate number of close games this year. Eight of their match ups were decided by four points or less, with a ninth decided by an overtime touchdown. What makes up the rest of the record? Three blowout wins, two blowout losses, one comfortable victory, and one comfortable loss. As OCC mentioned, that's more close games than all but two other teams. To boot, the Cowboys had 11 games decided by a touchdown or less in 2010.

The 2011 Cowboys went 4-4 in those games decided by four points or less. That record looks even worse if you add the overtime loss to the Cardinals making the team 4-5 in close games this season. By comparison, the Giants played only five games this year that were decided by four points or less, but they one every single one of them. That's right, the Little Boys Blue went 5-0 in close games this year. If you include the post-season, New York was 7-0 in games decided by four points or less.

Now maybe that's just randomness within randomness. Maybe those seven wins in close games and those seven years of a positive "luck" differential are just a dice roll that happened to come up in the Giants' favor. But maybe there's something more to it.

If you play a number of close games and you're inconsistent under pressure, then you can expect a pretty average "luck" differential. In other words, if you do not always finish, whether in terms of mounting comebacks or staving them off, then you can expect to split your close games. The Cowboys pretty much did that and ended up with a middle-of-the-pack luck ranking this year.

In general, a number of close games ought to keep your "luck" ranking pretty square in the average since they should keep your ratio of points scored to points allowed to be pretty tight. Improvement in one's "luck" comes when a team excels in these tight situations. If you win a number of close games, then your ratio of points scored to points allowed is not going to change drastically, but your number of wins certainly will.

This helps explain the disparity between how many games the "luck" formula predicts the giants ought to win based on their points scored/points allowed differential and the number of games they actually end up winning. While the points scored and allowed in those close games even out for a team like the Giants, by coming up big in those games, they're able to make their number of wins outstrip their point differential.

Over the past few years, the Cowboys have seemed to fold when the going gets rough. Sure, 2011 was a banner year with the heartbreaking losses to the Jets, Lions, Patriots, and of course, the Giants, but it's been an issue for years now. Now perhaps success in those close games is just as much of a coin flip, but I believe there's something more there. There's something to how a team responds when the pressure is on that bears itself out in these numbers. Mental toughness is a difficult thing to teach–how not to fold when things are tight–but if the Cowboys want to improve their luck in 2012, building fortitude when the pressure is on would be a big step in the right direction.

Even after all of this, I'm genuinely not sure if there's really any way to "change your luck" here. What do you guys think? Is there a good reason for why the Giants have been so much "luckier" than the Cowboys in recent years or are they, as OCC puts it, just randomly on "extreme ends of the statistical probability bell curve"? I'd love to hash it out in the comments.

Enjoy other randomness at The Andrew Blog.