Team Recruiting Rankings

[olsontex]

I started writing the following novel in the '2022 QB Josh Hoover' thread in response to btownqb's comment:

"A commitment from Hoover moves us to like 33rd overall and we'd be the highest ranked team with 6 commits." 

I decided to be a respectful poster and not clutter the John Hoover thread with a comment that was only marginally related at best.  It's hard sometimes to have the foresight that a post is going to veer too far off course to belong in a specific thread, then again sometimes it should be blatantly obvious.  I don't know if it bothers other folks all that much but it's a pet peeve for me.

OK, moving on...  btownqb's comment reflects exactly how I think about the quality of a recruiting class, and I doubt I'm alone.  I've never understood the logic behind team recruiting rankings being based on an accumulation of points, even when individual players are weighted in a non-linear manner.  This approach works much better for basketball than football but even then it's not ideal. 

When I was a little kid I wasn't always the best brother.  I would trade "money's" with my sisters and feign consternation over giving up the "prettier" looking coins, larger sized coins, or MORE coins (most applicable to the current team ranking system).  A typical trade would be something like 1 quarter, 5 nickles, 10 5-pfennig coins, and 6 pennies in exchange for 1 franc, 6 dimes, and 7 20-pfennig coins.  If this multi-currency example seems bizarre, it's a product of growing up as an ex-pat living in both Belgium and Germany.

For my example I didn't choose these coins or denominations randomly.  My sister landed a recruiting class of 22 coins that includes 1 high 4-star, 15 3-stars inside the top 247 1250 players, and 6 3-stars or 2-stars outside the top 1250. 

- I landed a recruiting class of 14 coins that includes 1 5-star, 6 4-stars, and 7 3-stars inside the top 247 1000 players.  My class includes 1 major difference maker, 6 high impact players, and 7 useful players most of whom will eventually start and see regular action. 

- My sister's class includes a player who will gain national attention, 8 players who will becomes average starters, 7 players who will see action but not enough to make much significant impact, and 6 players who will function as practice dummies.

To real this back to football, roughly 55-60% of players on scholarship will actually play regularly and determine the success or failure of the program.  Let's be generous and set it at 55 players, and not forget that a few transfers will come in and take someone's playing time too.  With the Redshirt system our team is comprised of 5 recruiting classes.  A few players from each class will transfer out, become too injured to continue, or simply wash out.

The question is how many QUALITY players does a team need on average in each class from a playing time perspective, and how many players in a class typically make an impact on a program.  In my mind that's where the cutoff should be set for ranking recruiting classes.

In high school I was a basketball and baseball player but was intent on playing a sport in all 3 seasons to help stay in shape.  I made the football team but my 6'3 175 pound wirey frame was not designed to play TE and DE, nor was I fast enough to play CB.  I flipped to Cross Country the next season thinking it would be less taxing (boy was I wrong).  I bring this up because the scoring system for Cross Country is close to what I think makes sense for ranking College Football recruiting classes.  In Cross Country the team consisted of 15 runners of which 12 could participate in a given race.  The top 7 finishes of the 12 runners would count for scoring.

This number is obviously too flow for football recruiting rankings but conceptually I think it's the right approach.  Especially if combined with an intelligent weighting system based on player rankings.  Here's what I think makes sense:

Base the team score on the 15 highest ranked commitments and employ a sophisticated non-linear points system within ranking tiers.  If a class only has 14 players then they receive 0 points for the 15th slot.  If the 16th ranked player in a team's class is a 4-star ranked 350 they don't count.  I'm not suggesting depth doesn't matter, but at a certain point the likelihood of a player receiving enough playing time to matter passes a threshold of insignificance.  We have perfect examples of how this plays out if we look at some of the transfers IU has secured the past few years.   They may be positioned to be big difference makers at IU but there was simply too much competition on their higher profile team for them to make an impact on par with their recruiting ranking. 

There is ample data available to assess the typical career track at each ranking range.  This can be used to assign points based on the probability of a player ranked XXX making a specific level of impact on the field.  "Impact" could be defined as some combination of games started, minutes played, and percentage of plays on the field.  I mentioned "range" a moment ago, which I think is important.  It exists to a small extent today but I think it needs a major revision.  When I look at 2 WRs, one ranked 200 overall and the other 235, the idea that there is enough information to determine any real separation between the two seems unlikely.   In my mind the team should be awarded the same number of points for both of them.  The current system awards 6.6 points for #200 and 4.4 points for #235.  I chose an example of two highly ranked players but obviously splitting hairs once you get outside of the top 250 becomes even more silly.

I don't know exactly where the points of separation lay within the rankings, but I know I could easily calculate them based on historical data.  There may be a positional component that needs to be considered as well.  I imagine the washout rate for a Guard ranked inside the top 600 is fairly low but I think the exact opposite for skill positions that tend to be disproportionately higher ranked.  I also don't know the degree of sophistication or effort that is currently being used to determine the current points system but on the surface it appears shallow and lacking in nuance.

Anyway, this was an extremely drawn out way of saying the average player ranking of IU's 2022 recruits that btownqb pointed out is WAY more important than the overall points ranking.  However, even the current more meaningful perspective of average ranking is flawed in my opinion.  I'm a bit tired today and I know I didn't lay this discussion out particularly well, especially with the diversions on my youth.  I also know I could improve the current points system that is currently in place on 247 by creating a dynamic non-linear model that adjusts year-over-year to reflect changes in potential (ranking) versus results (on field impact).

[Leathernecks]

Haven't had a chance to read through your post, but there is a thread in the recruiting forum dedicated to general discussions and rankings. This would be a good post for that thread, and we can have a discussion there.

Looking forward to reading through your thoughts when I get some time!

[olsontex]

Thanks Leathernecks.  Is there a way for me to reassign my post to that thread or do I simply need to copy and paste my post and then delete this topic?

[Leathernecks]

23 minutes ago, olsontex said:

Thanks Leathernecks.  Is there a way for me to reassign my post to that thread or do I simply need to copy and paste my post and then delete this topic?

I would guess copy and paste would be the best option.

[btownqb]

31 minutes ago, olsontex said:

Thanks Leathernecks.  Is there a way for me to reassign my post to that thread or do I simply need to copy and paste my post and then delete this topic?

Hey this is good stuff though 

Take it over there and we'll talk more