How Lucky Was Miami in 2013 And What Does That Mean For The 2014 Season?

facebooktwitterreddit

Nov 2, 2013; Tallahassee, FL, USA; Miami Hurricanes running back Duke Johnson (8) runs the ball past Florida State Seminoles defensive back Lamarcus Joyner (20) during the first half at Doak Campbell Stadium. Mandatory Credit: Melina Vastola-USA TODAY Sports

Many like to pretend that there’s little to no luck in football and it’s an understandable position. Football is one of the few things in life that bring us black-and-white outcomes—either you win or you lose. One team exerts its superiority over the opposing team and wins the game as a result. But things are obviously far more complicated than that and it would be shortsighted to simply look at outcomes to decide which team was better. Similarly, over the course of a season, a team’s record isn’t necessarily representative of how good of a team it was that year, especially in just a short 12-game sample that a college football regular season gives us. Whether we like to admit it or not, luck plays a role in determining outcomes of games and seasons. In this post, I’ll identify and explain several luck-based factors in football and examine how they affected Miami in 2013—and what that means for this coming season. I’m going to look at conference play only because games against Florida Atlantic and Savannah State certainly skew last year’s data set. Quick thanks to Bill Barnwell over at Grantland for inspiring this post.

More from Miami Hurricanes Football

Pythagorean Win Expectation

This concept holds true for every sport: a team’s scoring differential is a far better measure of its quality than the team’s win-loss record. Without delving into why this is (there’s ample information on this out there and a quick google search will do the trick if you are interested), luck comes into play here when a team’s win-loss record either outperforms or underperforms its pythagorean win expectation. For example, a team that scored the same amount of points as it allowed would theoretically have an even 4-4 record in ACC play. If the team went 5-3 that would be relatively lucky, just as it would be relatively unlucky to finish 3-5.

In 2013, Miami finished 5-3 in the ACC while scoring 250 points and allowing 261, for a point differential of -11. As you could guess, the Hurricanes outperformed their pythagorean expectation, but not overwhelmingly so. Their expected record would’ve been 4-4. So Miami luckily won an additional game in the ACC in 2013. For this year, that’s not a good sign. For one, the ACC slate is more difficult. Instead of a cozy Atlantic matchup against Wake Forest, Miami has to travel to Louisville at night to face a very good Cardinals team, albeit not as good as last year’s outfit. Add in the likelihood of small regression and we should move to the next category before things get a little dark.

Record In Close Games

While those that spout narratives love to talk about how the teams that just know how to win the close ones are more successful in tight games, data tells us this just isn’t true. The outcome of close football games is largely random and numbers show that replicating great records in one-score games doesn’t happen from year to year. (One exception to this rule is that teams with elite quarterbacks tend to win more close games than teams that don’t. Stephen Morris was hardly elite last year and the Hurricanes certainly don’t have an elite option in 2014.) This means teams with a really good record in one-score games will likely regress the next season as it’s highly unlikely the ball will bounce their way quite so many times two years in a row.

Miami finished 3-0 in one-score games in 2013, including one of the luckiest wins I’ve ever witnessed in a cardiac arrest-inducing 21-16 win against Florida. (Thanks, Jeff Driskel!!) The other two came on the road against North Carolina and at home against Wake Forest, respectively. Unfortunately, that means the Hurricanes are due for some regression here as well—it’s highly unlikely Miami will finish undefeated in one-score games again in 2014. Maybe we’ll find something to salvage in the next few categories.

Defensive Touchdowns

Defensive touchdowns are very random—while good teams are generally better at avoiding turnovers on offense and forcing them on defense, the number of occurrences that lead to the seas parting for a defensive player with the ball in his hands are something that even the 1972 Miami Dolphins couldn’t continually replicate. If a team picks up a staggering number of defensive touchdowns, its likely to regress in points scored the next season when it isn’t quite so lucky. On the same token, a team that allows an abnormal number of defensive scores is likely in for a positive regression the following year.

In 2013, Miami recorded four defensive scores: Ladarius Gunter picked off a pass and took it to the house against Georgia Tech, Gunter caught a blocked field goal and ran is back at North Carolina (yes, this is technically a specials teams score but I’m counting it), Tracy Howard recorded a pick-six against Virginia and David Gilbert memorably rumbled his way to a scoop and score, also against the Cavaliers. Miami allowed zero defensive scores. Gulp. No matter what luck-based metric I use, it’s becoming apparent that Miami was quite lucky in 2013. Let’s look at one more.

Fumble Recovery Rates

Overall, history shows us that it is largely impossible for a team to be skillful at recovering fumbles. If a team recovers more than 50% of balls on the turf over the course of a season—no matter if the team or its opponent coughed it up—they were fortunate that year. Since there’s really no rhyme or reason where a football bounces, it stands to reason that inferior teams still have roughly a 50% chance of recovering them against superior opposition, so I’ll include at 13 of Miami’s games for this category in an attempt to draw data from the largest sample size possible.

Last year, Miami games featured 31 total fumbles, 14 from the Hurricanes and 17 from opponents. Miami recovered 14 of them, or about 45%. Finally! We found something that could benefit Miami in 2014, although quite modestly. In reality, this is a pretty standard outcome—Miami really only was unlucky by one or two fumbles over the course of thirteen games.

Conclusion

There are a couple other luck-based factors that come to mind, such as injuries (Miami was pretty unlucky injury-wise in 2013, losing Duke Johnson, Phillip Dorsett and Malcolm Lewis for much or all of the season and having a couple key players—namely Stephen Morris—have to struggle to play) or strength of schedule (Miami’s was much easier in 2013 than it will be this year). But overall the data is pretty clear—Miami was a very lucky team in 2013. Now, before everyone starts panicking and calling me negative, I want to stress a few caveats here:

  1. Despite the fact that many subscribe to the gambler’s fallacy, just because Miami was lucky in 2013 doesn’t mean they’re due to be unlucky in 2014. In fact, they have just as much of a chance of being lucky again as they do of being unlucky. The more likely outcome is that they regress back towards the mean, which unfortunately would see them fall short of the 2013 team in some key ways.
  2. This sample size is REALLY SMALL! Eight games is hardly anything. You can find an eight-game sample of baseball this year where the Cubs were the best team in the league over those eight games. While this is an extreme example, the luck-based outcomes from last year’s Miami team are hardly 100% definitive. However, don’t take that to mean you can completely disregard those numbers either because they do comprise the best sample we have. Just take them with a grain of salt.
  3. Despite the uncertainty at quarterback, this team is improved from last season. While the 2013 squad needed a fair amount of luck to get to 9-4, this team definitely won’t need as much and their main goal for this year should definitely be winning the Coastal even though their ACC schedule is more difficult.
  4. I really want to hammer home the fact that I’m not at all a pessimist and not everything I post in here will be so gloomy. I’m merely drawing logical conclusions from the data.

Bottom line: While it’s entirely possible for this team to defy expectations, last year the Hurricanes were quite lucky and that unfortunately means it’s fair to expect at least a modest drop-off in quality of play this year.