Sunday, August 5, 2007

Recruiting: Four and Five Stars

In an earlier post, I tried to determine whether or not recruiting rankings mattered in terms of future team performance. One of the many conclusions of that post was that approximately 80 per cent of Alabama's roster (not the total roster, but only those on scholarship)the past few years has been comprised of players rated as three-star or lower by the recruiting services. Long story short, the argument was that number was / is entirely too low.

So, I decided to take the same time period (2002-2004) and determine how many four and five star recruits that each SEC team signed over that period.

The following is what I found, listing the number signed by each team per year, and then the total for the three year period:



So, how well did all of that correlate with overall winning percentage in a future time-frame (2004-2006)? Here is that data broken down:



Much like with recruiting rankings, as you can see, where you rank in terms of how many four and five star recruits you sign, compared to the rest of the conference, strongly correlates with overall winning percentage on down the road.

To be specific, the correlation between recruiting rank and winning percentage rank was a robust .7202. That's not quite as strong as the .8 that we saw between overall team recruiting rankings and winning percentage, but it's still very strong.

At bottom, the teams that sign the most four and five star players are generally the ones that do the best on the playing field. Certainly there were a lot of busts included in those numbers, but on the whole the teams that signed the most high-quality players did better than the rest. It didn't correlate quite as high as overall team rankings, I presume, because recruiting isn't just about signing four and five star players (although that's the majority of it), it's also about signing some players that are rated three stars and below and turning them into quality football players as well.

Nevertheless, at bottom, it just further seems to indicate that, regardless of how vehemently some people want to assert otherwise, recruiting rankings do matter.

1 comment:

Troutcreekpass said...

Approaching this from another perspective: a few years ago, someone on Tiderinsider posted an analysis over a period of years (at least 5-10 years) of the difference between 3 and 4/5 star recruits in terms of their becoming stars at the collegiate level (being recognized as All-SEC or better). The 4/5 star players only became stars around 40% of the time -- in other words, most of the time they are "disappointing" in terms of how their careers pan out. On the other hand, if I remember correctly less than 20% of 3 star types go on to reach stardom (at least in terms of all star recognition). So yes, stars don't matter in the sense that most 4 stars are disappointing and 20% of 3 stars end up doing surprisingly well. BUT -- your odds are a ton better if you sign more 4 stars over 3 stars. Not a big surprise but another reason someone is in denial if they say "stars don't matter" at least in terms of overall class and program strength.