The Upside of Gravitating to Mediocrity
9 comments so far
June 17, 2009 by
The old statistician’s joke goes, “she said I was average, but she was just being mean”. But being average is only a bad thing if you’re aspiring to something greater. At long downtrodden football programs like Indiana or Duke, average would be a step up. Fortunately, there’s a tendency for things to tend toward average. Statisticians call it “regressing to the mean”. In areas of strength, it means you’re fighting gravity. Like when you try to string together over 40 winning seasons or over 35 straight bowl games. But in areas where you’re struggling, it means that improvement isn’t just possible – it’s likely.
Coming off of a 9-4 football season, the Huskers are going to have to work to maintain success in some areas. Nebraska was in the top twenty-five nationally in time of possession, first downs, third down efficiency (both offensively and defensively), red zone defense, total offense, passing offense, scoring offense, pass efficiency, punt returns, rushing defense and in getting sacks. Repeating or improving those numbers won’t be automatic. In fact, there will be a pull in the opposite direction to overcome.
But some areas where the Huskers want to get better almost have to get better. NU finished 107th in turnover margin in 2008, 106th in net punting, 103rd in kick coverage, 102nd in penalties per game and 98th in penalty yardage. You’d expect an uptick there almost by default. The Huskers were also 82nd in pass efficiency defense, 80th in scoring defense, 82nd in tackles for losses that they allowed, 88th in fourth down conversion percentage and 89th in pass defense. Those also seem like areas where improvement would be likely.
It’s not only the number of areas where the Huskers are likely to improve that could be seen as encouraging, but the categories. Turnover margin is a huge predictor of wins and losses on the field. They create cheap scoring opportunities and stop drives. One turnover clearly seemed the difference against Texas Tech. The special teams coverage will help the defense (and offense) immensely as the Huskers can stop giving up free yards with each change of possession. The penalties would seem the easiest thing to correct since they are self-inflicted. It was a penalty that might have cost the Huskers the game against Virginia Tech and there were some really costly ones against Texas Tech for that matter. Scoring defense is obviously a critical number because it’s what goes on the scoreboard that ultimately determines who wins and who loses. The tackles for loss and fourth down conversions ultimately cost the offense chances to be more effective.
All in all, gravitating to mediocrity in areas of weakness in 2008 would be a good thing for Nebraska in 2009. If the Huskers climb to the middle of the pack in those areas, the improvement will be obvious. And why stop there? If the Huskers can take an area of weakness and actually make it a team strength (e.g. turnovers), then ultimately they can go from being just competitive in the Big 12 to becoming true contenders.
Did you enjoy this article?
Get Husker news by email
Follow us on Twitter
Related Stories:
Post a comment
darren Jun 17 09
Excellent points about the nature of stat turnarounds, Steve. Reasons for optimism, I think.
I also know that is the nature of football (and often college sports in general) to be strong in a former area of weakness, just because of personnel growth/changes.
If a unit played a lot of young or inexperienced players one year, and got taken advantage of, they return the next year (or 2) as more tested and experienced players and actually lead the team.
caveman99 Jun 17 09
I like the sentiment for NU, and I agree that NU has to improve in those areas and most likely will. This doesn't explain ISU, KSU, Indiana, Washington State, etc. The statiscally premise is based on the main factor being even across the board. In this case talent is the base factor for all teams. If all teams had similar talent, then yes teams would flux in the rankings year to year and gravitate towards the median. This happens in the NFL more because the level of talent is more even, dynasties are harder to make these days. In college fooball, this doesn't apply as much. To do this, we would need to relook at what we consider the median for NU to be. It sure isn't 60th out of 120 teams. NU is on a different talent plane than 80 teams. Is the median for NU 20th then?
Steve Jun 17 09
caveman,
Your point makes sense about talent being the great "unequalizer". But when it comes to a number of these statistics the ranges can be similar. In part, because conferences tend to be more equal than the full range of college football. A team like Hawaii may not be the best in the country, but it can still lead the nation in passing for example.
And a lowly Iowa State can still be 12th in kick returns, and have a player in Austen Arnaud who finishes 23rd nationally in total offense. Indiana has a guy finish 6th in the country in tackles for loss, and 14th in sacks. The Hoosiers had a guy finish 25th in all-purpose running and 24th in punting. Kansas State was similarly in the top 25 in a bunch of categories (scoring offense, passing offense, sacks allowed, kickoff returns) and had a top 25 all-purpose runner and top 20 receiver. Washington State was pretty abysmal across the board, but I'd be shocked if we don't see at least some statistical improvement over their god-awful performance of a year ago.
Brandon Jun 17 09
Not to keep going to the well, but one of them most interesting tidbits of info in the new Phil Steele was his adjusted pass efficiency defense rankings. When adjusted for the passing offenses NU faced, Steele had the Huskers ranked 39th in PED. Iowa State finished last.
caveman99 Jun 17 09
Level of talent in each conference allows Hawaii to finish that high. Put Hawaii in the SEC and they may finish high, but not that high. This also doesn't really explain the teams that perenially do better than the median. Every team can have really good stats in one area and not so much in another, but why do the really good teams stay in the upper portion of the majority of the stats? Call me crazy, but I don't think that UF is going to gravitate to the median of 60 in the vast majority of these categories any time soon. I don't think NU should settle for that either. The whole concept of the mean is that over the long run, that is where the trend should lie. If I flip a coin I may get heads 10 times in a row, but over 100 flips it will come out very close to 50 heads vs. 50 tails. NU was good for 35 years or so, does that mean that we have to have multiple years of bottom 20 defenses because NU used to alwasy be in top 20? I would love to see a chart outlining how the perenially good teams did in the major categories vs. the perenially bad teams.
Steve Jun 17 09
caveman,
Agreed about Florida (in most categories), but the point here is that in order to be at the top consistently, you have to defy gravity. Think of it as wearing weights around your ankles when you run. In 2006, Florida won the national championship and finished 6th in scoring defense. A year later, with the same head coach, defensive coordinators, and the same caliber of athlete, they dropped to 46th.
Did the players get dumber or slower in one year? No. Rather 6th is an extremely high bar, and they moved back toward the middle of the pack. Consider that the ankle weights effect.
Regression to the mean isn't absolute or universal. But it does happen and it's why for example when you're measuring the effect of a study aid (like Hooked on Phonics), you can't just look at the kids with the poorest test scores prior to using the study aid. Some improvement in test scores will happen due to regression, but some kids who have struggled will continue to struggle and some that have always excelled will continue to excel. Sometimes Michael Westbrook comes up with a Hail Mary to win a game, and sometimes Michigan will bat it down and preserve a victory.
So regression to the mean isn't a cure-all or as you point out even your best hope to get to your goals. But the point here is you can call it almost dumb luck that could make the Huskers better in some areas because there's sort of a pull in that direction.
caveman99 Jun 18 09
Steve,
caveman99 Jun 18 09
Steve,
I get it on the UF question, and I partly agree with your stance that teams will gradually go back to the mean. My question is can you look at the median of the national rankings as the mean for all teams? Does it make more sense to split up the rankings into zones that historically programs fall in? To take it one step further, do you break it down by area for each team? Hawaii traditionally has a good offense, but horrible defense. One year they may have a pretty good D, but eventually they will gravitate back down to their mean. I would love to find a site that has historical statistical rankings together to analyze. I think it would give a better picture of what the mean for the Huskers should be.
wetwilly Jun 18 09
Nebraska's level each year is subject to mathematical analysis but in the end the old platitude will factor in strongly and their history of having done it before probably gives a mild advantage but they still have to get in shape both physically and mentally. The mental parameter is hard to measure mathematically but I'm betting on the Pelini brothers and crew to beat the 'mean' by a large degree. The platitude of course is "Your altitude is determined by your attitude".