clock menu more-arrow no yes

Filed under:

UW Mythbusters: The 1-year Recruiting Lag

New, 63 comments

Is it true that it takes a year for recruiting to catch up to winning?

NCAA Football: Pac-12 Championship-Colorado vs Washington Kirby Lee-USA TODAY Sports

Recruiting and winning invoke the old chicken and egg conundrum. Does a team start winning only once they have quality recruits? Or do you need quality recruits before a team starts winning? The answer to both is seemingly yes. Logically, the two go together in a feedback loop.

Any analysis done in the moment on the subject makes the assumption that we’re able to accurately measure a recruit’s potential in an arbitrary 1 to 5 numbering system. It turns out though that high school seniors aren’t finished products and aren’t numbers. Who knew? This is the point in the article where I acknowledge that recruiting rankings don’t always get it right. In fact, they mostly don’t get it right. But we now have to kill another 2 months before spring ball and 7 months until the season so I’m going to barge ahead without the mandatory 3-year evaluation period necessary to get an informed opinion.

Much has been made on this site and others about the seeming lack of impact that winning a Pac-12 championship and making the College Football Playoff had on UW’s recruiting efforts. Players such as Foster Sarell, Marlon Tuipulotu, and Connor Wedington all turned down the chance to play for the defending conference champs. Many more players didn’t even let UW get in the door.

On its face, this seems perplexing and therefore people have searched for an explanation. The most popular one is that it takes at least a year for a team’s success on the field to translate to success in recruiting. This makes sense.

Recruiting doesn’t begin after the season begins but rather starts many years in advance. Many recruits had already crossed Washington off the list by the time the Dawgs started winning. With every year that UW continues to win, it follows that recruits will be more receptive at the beginning of the process and give UW an actual chance. More bites at the apple will lead to more hits and better recruiting rankings.

It’s a logical and reasonable narrative. But I don’t like narratives. Narratives tend to be crafted in place of examining the data. Let’s look at the data then, shall we?

Recruits = Wins + Time?

I looked at all teams that are currently in a Power-5 football conference going back to 2001. The reason for that cutoff point is I used scout.com for all of the recruiting ranking data and that is as far as their team rankings database goes.

The first step is to identify our selection criteria. In 2016, Washington increased their win total by 5 (from 7 to 12). That’s a pretty high bar so to try to get a larger sample size we’re going to reduce it to teams that made a 4-win jump. But we want to get more specific than that. In 2009, UW improved from 0 to 5 wins. Technically it’s the same jump in wins but there’s a big difference in perception between going from awful to mediocre than from good to great. So we’re only including teams that had 10+ wins after their leap.

We want to look at before, immediately after, and with a year gap for each of these schools so the 3 teams that made a leap in 2016 are excluded (Colorado, UW, and Penn State). That leaves 47 teams that jumped by at least 4 wins up to double digit wins in that period. I am considering Year 0 to be the year before they started winning, Year 1 to be the signing period immediately after the big year, and Year 2 to be the following season with a full recruiting period in between.

Recruiting Rankings Before and After Win Jump

Year Average Team Rank Average Star Rating
Year Average Team Rank Average Star Rating
Year 0 30.02 3.03
Year 1 25.68 3.19
Year 2 26.60 3.22

As you can see, it looks like the greatest boost in recruiting actually happened in the year immediately after the leap in wins. Teams increased their recruiting ranking by 4.34 spots in Year 1. However, that number might be misleading. Total team rankings are based not just off of star ratings but also by the number of recruits signed. It stands to reason that the reason a team sees a marked improvement in wins might be because they have a ton of seniors or draft-eligible juniors who got better. If a team has a ton of seniors then they need to replace them and therefore sign more recruits. So star rating is the better measure.

And in this case, star rating also went up by 0.16 in Year 1. If you only looked at star ratings, that is equivalent to a jump from 39th to 29th in the rankings. So better than the 5 spot improvement in total rank. It’s not as if these teams suddenly challenge Alabama but it’s usually good enough to move them up a couple of pegs in the conference pecking order. With this class, UW just improved their rank by 6 and their star rating by 0.17 so they almost exactly followed the trend of the other similar programs.

Looking at the Year 2 line however seems to disprove our hypothesis. The team ranking went down by almost a full spot and the average star rating saw just a minuscule increase. That isn’t a great omen for any Washington fans hoping that this analysis would reveal a likely top-10 class for next year.

Let’s try one more thing though. Those numbers don’t factor in what happened in Year 2 in these scenarios. Let’s say that next season Colorado, returning only 3 defensive starters and losing their senior QB, drops back down to only 6 wins. They’ll seemingly be less likely to keep their recruiting gains in Year 2 than if UW wins the Pac-12 North again. So we’re re-running the numbers with only teams that won at least 9 games again in Year 2.

Recruiting Rankings for Teams with 9+ wins in Year 2

Year Average Team Rank Average Star Rating
Year Average Team Rank Average Star Rating
Year 0 31.86 3.03
Year 1 27.29 3.15
Year 2 27.21 3.22

Not quite what we were hoping for. That sample of teams saw bigger increases from Year 1 to Year 2 but only because they didn’t see as much of an increase from Year 0 to Year 1 and started off in worse shape in Year 0.

There are definitely individual cases of teams seeing a 1 year lag time in their recruiting. From 2009 to 2011, Stanford with Andrew Luck went from 8 to 12 to 11 wins. Their average star rating went up by 0.23 after the big spike but went up again by 0.22 the following year. From 2013 to 2015 TCU went from 38th in the rankings to 27th to 13th after they won 12 then another 11 games. But for every example like these there’s a counter which cancels it out on the average basis. Unfortunately, I’m going to officially call the one-year recruiting lag myth: Busted.

But that’s not a satisfying note on which to end. For those of you who have seen the show Mythbusters, almost every time they fail to get the result they want they have to make good TV by blowing something up. So let’s do some bonus research. (And yes, more research is just as exciting as massive explosions).

Impact of Wins on Average Star Ratings

I took the average star ratings for every one of the schools since 2001 and checked for the correlation between it and a few different categories. Unsurprisingly, wins were the most relevant factor in determining the average star rating. I looked at wins in the current season, wins for the past 5 seasons, and wins for the past 10 seasons. The results were almost identical but the past 5 seasons was barely the most meaningful number. That number was about 1.5 times higher than the correlation for final place in the AP Poll which makes sense since Washington didn’t vault into the recruiting top-10 with the #4 finish. So while a single year isn’t enough of a lag time for recruits, it appears that 5 seasons of elite performance may make the most difference.

Just for fun, let’s look at a graph of every team during that time with their 5-year average win totals on the y-axis and their average star rating that year on the x-axis. The colors are by conference with the Pac-10/12 in Purple and the SEC in yellow. The size of the dot is based on the win total from the most recent year so bigger equals more wins. (If you’re on a mobile device, I apologize that this may be a little hard to read).

Team Average Star Ratings and 5-year Win Total Averages since 2001

You can see the clear correlation (not causation) connecting consistently winning programs and strong recruiting classes. Of the 28 Pac-10/12 conference seasons with an average recruiting ranking of at least 3.5: 14 belong to USC, 7 to UCLA, 3 to Stanford, 2 to Oregon, 1 to Cal, and 1 is this year’s UW class. That top-right most purple dot is USC after the 2006 season when they went 11-2 and had an average star rating of 4.36. (It should be noted that the only player from that class to become an eventual NFL starter was #1 overall ranked Everson Griffen).

UW’s Average Star Ratings and 5-Year Win Total Averages since 2001

Here’s another look at that same graph but with the UW seasons highlighted. Unfortunately, there are a couple of microscopic dots in there which represent the disastrous 2004 and 2008 seasons. The selected dot farthest to the northeastern corner of the graph is the recruiting class Coach Pete just signed.

The Dawgs and Chris Peterson may never reach the recruiting heights of USC, but if they keep improving at the rate that they have been it’s legitimately good enough to win a national title. Clemson’s average star ratings in the 4 signing periods before their national championship: 3.35, 3.37, 3.38, 3.68. Washington’s last 3 recruiting classes: 3.21, 3.33, 3.5. All it takes is one more step up. That may not be likely but it’s certainly possible and right now that’s good enough for me. Go Dawgs.

(And shout out to THD1 for giving me the idea for the column)