In a recent episode of his podcast Revisionist History, Malcolm Gladwell applied Chris Anderson and David Sally’s argument about weak and strong links to the US education system. Their original argument was about soccer teams — essentially, they argue that improving the quality of the worst players on a soccer team is more important than getting superstar players. Another way, it’s better to raise a team’s floor players than to raise the team’s ceiling players.
Their argument doesn’t apply to all sports — in fact, basketball is almost the opposite from soccer. A single elite player, like LeBron on the Cavs for instance, can completely change the team because he is able to control so much about the game. In soccer, an elite player simply has fewer touches per game, meaning that weak link players have more opportunities to make critical mistakes.
If we were thinking about this in terms of college football, is it more important to improve the weakest positions on the team or stockpile elite talent? Is it better to go after five-star recruits or limit the number of three-stars?
There’s reason to think that football might be more similar to soccer than basketball in terms of the relative importance of weak vs. strong links. First, football has the same number of players on the field as soccer, meaning more opportunities for weak links to affect the entire team’s play. It doesn’t matter if four of five linemen are five-star pass-blockers -- it only takes one turnstile to lead to a sack.
Second, there’s been a good amount of research suggesting that NFL teams significantly over-value high NFL Draft picks. Given a limited amount of financial resources, it would be better to address a team’s personnel weaknesses by stockpiling pretty-good players with later picks than to chase the incredibly expensive and overvalued front end of the first round.
Eventually I’d love to get a ton of data and fully test the weak vs. strong link theory in college football, but for now I just wanted to look at Ohio State’s recruiting classes since Urban came to Columbus. I’m guessing that the theory applies more to weaknesses on the field than to weaknesses in a recruiting class — i.e., if you’re a recruiting coordinator, is it better to chase two four-star recruits in a position of critical need or go after a five-star, top-5 overall recruit at a position of relative strength?
But that’s not exactly what I look at here — instead, I looked at the top and bottom ends of six recruiting classes. Usually, in the 247 Composite team rankings, we only see either the overall total team score or the average rating of all players in that class. But the class average can be significantly swayed by clusters of player ratings. For example, a recruiting class average could be four stars, but that class could either be all four-star recruits or half five-stars and half three-stars. So I separated the top 25% and bottom 25% of the last six Ohio State recruiting classes by year:
|Year||Weak Links||Class Average||Strong Links|
And here's that data in chart form:
I excluded specialists from the classes in the data above because the kickers and long snappers were significant low outliers, and only a few classes actually included a scholarship specialist.
There are a couple takeaways for me. First, strong link recruiting has been consistent throughout Meyer's tenure. The Buckeyes typically get a similar number of elite recruits per recruiting class. In the past six years, the percentage of players above .9 in the 247 Composite has gone from 52%, 56%, 78%, 73%, 48%, and 56%. There are two really elite classes, but mostly hovering just above half of the class.
In contrast, the average of the lower 25% has steadily declined since the highs of that 2013 class -- that is, the lowest-ranked 25% has gotten worse since 2013. The overall class average has been varied, however. Finally, this data only goes back to the beginning of Urban's tenure -- if I had data going back further, I think we'd see a decent uptick in both the top 25% and bottom 25% from the Tressel/Fickell era.
I also looked at year-to-year changes in the average recruiting rankings, and the results show a good amount of variance:
So in the chart, the 2013 class average was an average of .032 points better than the 2012 class. Here, it's easy to see the impact of just how good the 2013 class was -- the 2014 and 2015 classes were both great, but had a marked decline relative to 2013.
It's not until the last class in 2016 that the ratings of the top 25% of the class improved over the previous year. Going back to the first chart, the average rating of the top 25% of the 2016 class -- which is equivalent to roughly the top six players -- was the fourth-best in Meyer's tenure at Ohio State.
The bottom 25% of each class since 2013 has gotten steadily worse since the 2013 class, but the rate of decline has also slowed each year. The thing is that the worst six players in each recruiting class could end up being stars. For instance, the lowest 25% in the 2012 class included Cardale Jones, Jacoby Boren, Mike Thomas, and Pat Elflein, while the 2013 lowest-rated recruits included Darron Lee, Chris Worley, and Tyquan Lewis. Malik Hooker is in the bottom 25% for 2014, and Damon Arnette, who's challenging for playing time at corner, is in the 2015 class.
It's possible that, even though the bottom 25% of each recruiting class appears to be getting steadily worse each year, the coaching staff is just especially adept at identifying and developing diamonds in the rough. We won't really know for sure until the later classes -- like 2014 and 2015 weak link recruits -- become upperclassmen.
Finally, I took a look at the middle 50% of each recruiting class -- the group that's not weak link or strong link:
The important line here is the middle orange one for the middle 50% of the class -- and specifically the middle years, in 2013 and 2014. While the upper tier of each recruiting class has stayed fairly constant year-to-year, and the bottom 25% has steadily declined in average recruiting rating, the middle 50% had a definite bump in 2013 and 2014. That means that the top 75% of each recruiting class in 2013 and 2014 was incredibly strong -- and the middle 50% especially strong relative to Meyer's first two years in Columbus.
That's particularly encouraging for 2016 and beyond, because those players are now upperclassmen. Even if you (very wrongly) assumed that the bottom 25% of each recruiting class never play a down at Ohio State, the overall talent base of the program grew significantly in 2013 and 2014, and then grew again with a stellar 2016 class. 2015 was a relative decline for the lower 75%, but it was still an impressive, seventh-ranked class in the 247 Composite team rankings. Nearly 75% of three of the last four recruiting classes were rated .92 or higher.
Overall it's difficult to see whether the Anderson/Sally theory really applies to Urban's recruiting strategy just yet. It would be better to follow up and look at weak links in the actual two deep, and see if on-field weak links lead to a shift in recruiting strategy in the following recruiting class.
But if nothing else, this data is another way of showing just how well Ohio State and Urban Meyer have recruited as of late, and that should bring positive things for 2016, and beyond.