Class Rankings


Robert
May 04, 2020
[]
[]
7 Comments

I started writing the Prince Green LLUOI and I realized I haven't written the Samari Collier LLUOI (2021 recruit who verballed in December 2019). I started researching Samari Collier, pulled up some Deuce Spann film for comparison (I just think Spann is so good), and I realized there's no way to accurately "rank" the 2020 recruiting class. It has a kid like Spann, but it's not going to rank anywhere close to anything because there's only 13 recruits. Why only 13? Very few seniors on last year's team + this staff's focus on (obsession with?) transfers.

This is how my brain works. As I step down that ladder, I realize I have to keep going. It made me think about Matt Hinton's preseason rankings (the ones I swear by every summer) and how one of his categories is recruiting. Makes sense. His formula looks at what kind of program you've had the last few years, how much offense you have returning, how much of your defense is returning, how much overall experience you have, and how well you've recruited the last four classes. Put it all together and it spits out a preseason rankings. I swear by it for many reasons, but last year was perhaps its best year as it correctly predicted Minnesota's great season as well as Northwestern and Purdue's tumble. Experience and returning production are everything, and that's either boosted by recruiting or reduced by recruiting.

So here's the steps I walked down:

  • Wait, I can't write the Green post because I can't find the Collier post which means I never wrote it.
  • This Collier film is not good compared to Deuce Spann. I should compare and contrast the two.
  • Man, Deuce Spann is SO GOOD and nobody knows it because he's just lumped in as part of the "88th best class".
  • The other day I wrote about how 15% of our roster is now made up of transfers, and those players don't even count towards these classes.
  • Thinking about it, that means that preseason rankings that include recruiting averages (like Hinton's) won't even include 13 of our 84 players - it's a number that pretends Luke Ford and Josh Imatorbhebhe don't exist.

That brought me here. If you just went and randomly removed 13 players from Ohio State's roster (or Clemson's roster, or whoever), they're going to drop significantly in those recruiting rankings. We're obviously not recruiting anywhere close to these teams, but our classes are being "graded" on bringing in 71 players, not 84, and that's not painting an accurate picture.

I've made this point before, just in different ways. When talking about the specific recruiting class rankings, I've pointed out how the system isn't built to "rate" a class of 13 players. Composite rankings are great (average out everyone's opinions), but they struggle to accurately paint a picture of a small recruiting class. Last year, in the composite rankings on 247, Cal's class was ranked 39th and USC's class was ranked 56th. OH MY GOD EVEN CAL IS OUT-RECRUITING CLAY HELTON. Not really. USC only had 13 scholarships and Cal had 26. You can add 13 no-name mid-to-low 3-star recruits to USC's class - players who would never see the field at USC - and push it up 25 spots and past Cal in the rankings just based on quantity. Even though the 13 players added would never play at USC, the warm bodies would push them past Cal in the class rankings.

Again, this isn't the fault of the people who put together the formula. It's more or less impossible. I'll try to not get too lost in the weeds here but I think they did it right. Most college recruiting classes have between 18 and 23 players, so make that the sweet spot. Make it so a big class of so-so players is never rated above a class of 18 great players. Beyond 23 players, steep diminishing returns (it needs to slow down or else a class of 28 with a bunch of backdates is going to rate way too high). It also can't boost USC's class of 13 too much (13 is still 13) and rank it above a solid class of 20 unless the average rating has a massive difference.

A quick example and then I'll move on. I'll use Cal since we just played them in the bowl game. According to the Composite Rankings, Illinois had 13 players, average score 84.90, and that class ranks 88th. Cal had 26 players, average score 85.66, and that class ranks 39th. I looked up the Bowling Green recruiting class from 2020 and picked five two-star recruits from that class - Adrian Wilson, Fortune Woods, Matt Fortner, Myles Williamson, and JB Brown. Using the class calculator, I added those five players to both the Illinois and the Cal classes. The Illinois class moves from 88th to 69th - the Cal class stays at 39th. In their formula, the same five players add more than 10 points to the Illinois class but add 0.4 points for Cal.

This is not a "see, the Illinois class was not half bad" point. I'm just trying to show how the class rankings work. There's no perfect system (unless everyone had exactly 20 recruits or something), so they do the best they can and set up a system with diminishing returns. Those five players are the 14th, 15th, 16th, 17th, and 18th-best players in the Illinois class but the 27th, 28th, 29th, 30th, and 31st-best players in the Cal class so the numbers adjust based on that. It's not an awful system, but... it's never going to accurately compare a class of 13 to a class of 26.

Which, I should note, is why I'm happy they added the "roster talent" page. All that really matters is your 85 guys against their 85 guys. Those rankings aren't out for 2020, but in 2019, the Illini were 37th on the "Team Talent Composite" rankings. Boosted, of course, by the fact that Wole Betiku was a 5-star back in high school and Trevon Sidney/Richie Petitbon/Brandon Peters et al were 4-stars. And that gets complicated because Betiku obviously wasn't a 5-star anymore. More on that in a bit, but if you're wondering how 247 could rank Lovie's classes all right around #50 and then they do a roster talent page and Illinois is suddenly 37th, well, it's the only ranking on their site that acknowledges the transfers.

There are other parts of this that make this difficult (and I've always said this about juco recruits). Junior college kids are only giving you two years vs. the other players giving you four years (in theory), so shouldn't they be rated lower? The same goes for graduate transfers. If you only get one year out of Wole Betiku, that's not the same as drafting some high 3-star defensive end at the same talent level (something like a .8700 composite ranking) and developing him over four seasons. Then again, two years of Brandon Peters is going to provide a lot more than four years of some low 3-star quarterback, so there's a lot of value in higher-ranked graduate transfers.

It's ranking them accurately that's the problem. Betiku was a five-star in high school. That's recruiting for "future first-round NFL pick" since they only name 25-30 five-stars every year. He obviously wasn't a first-round pick (signed an undrafted free agent contract with the Giants), which means his 5-star rating was inaccurate. 247 attempted to correct this by issuing new ratings to transfers - Betiku was re-rated with a .8700 rating as opposed to his .9884 rating coming out of high school - but they only did that for a certain number of transfers. For Illinois, of the last 13 transfers, only 3 have been re-rated. Which is why that "talent-wise, Illinois is 37th" is probably a bit inaccurate.

Although, it should be noted, Ohio State and Alabama and Clemson have several 5-stars on their rosters who really should be re-rated as 3-stars because they've already been passed up on the depth chart, and on the talent composite rankings, they're still getting credit for them as 5-stars. So maybe transfers shouldn't be re-rated at all and a "here's where these 85 players ranked coming into school and here's how that compares to other schools" is the best way to do it.

All of that to get to this: I'm going to attempt to re-rank the 2019 and 2020 classes by adding the transfers to each class. And I'll attempt to re-rank each player with an accurate rating of what I think Illinois is getting. Blake Jeresaty, for example, was a 0-star who went to Wofford. But he was a first-team All American there and reportedly had offers from "all five Power Five conferences" when he put his name in the portal, so that's not a 0-star.

I'm also making my cutoff July 1 (each class generally arrives on campus in June, so July 1 seemed like the correct date), so Milo Eifler (transferred in July of 2018) is added to the 2019 class and Derrick Smith and Chase Brown (transferred in August of 2019) are added to the 2020 class. That's probably better anyway because Eifler couldn't play until 2019 and Smith & Brown can't play until 2020. I considered doing the same with Luke Ford (another sit-out transfer), but he enrolled last spring so it feels right to put him in the 2019 class.

Here's the ratings for the three players who 247 re-ranked as transfers:

Luke Ford - from .9753 four-star to .9300 four-star
Richie Petitbon - from .9696 four-star to .8500 three-star
Wole Betiku - from .9884 five-star to .8700 three-star

(I'm not sure why Luke Ford's rating dropped after transferring after one redshirt freshman year, but whatever.)

Here's my ratings for the other ten transfers:

Josh Imatorbhebhe - .9000 four-star
Brandon Peters - .8900 four-star
Brian Hightower - .8800 three-star
Milo Eifler - .8800 three-star
Chase Brown - .8700 three-star
Blake Jeresaty - .8600 three-star
Trevon Sidney - .8500 three-star
Derrick Smith - .8472 high school rating seems accurate
Christian Bell - .8400 three-star
Brevyn Jones - .8317 high school rating seems accurate
Desmond Dan - .7783 high school rating seems accurate

I'm trying to be conservative here. I think Jeresaty will provide much more than a .8600 three-star performance, but that's what I'm hoping to see, and hope always gets in the way. If I let my hope out, unleashed, Jeresaty would get a .9400.

Now we'll add these players to their respective classes. Smith, Brown, Dan, Jones, Bell, Jeresaty, and Hightower go with the 2020 class and the others all go with the 2019 class. Below are the lists of classes with the player rankings (either their composite ranking, the 247 "re-rank", or my re-ranks listed above). I had to add a rating for Devon Witherspoon since he didn't get a rating at all (unknown recruit added in August). I went with .8400. That's probably way low (he was our best freshman), but I'm trying to remain a bit conservative. If I could sing my hopeful song, Chase Brown would be a .9300 four-star. Because he's totally a four-star.

OK, the lists.

2019
Marquez Beason .9564
Isaiah Williams .9479
Luke Ford .9300
Shammond Cooper .9027
Josh Imatorbhebhe .9000
Brandon Peters .8900
Milo Eifler .8800
Moses Okpala .8784
Keith Randolph .8701
Wole Betiku .8700
Seth Coleman .8599
Kyron Cumby .8581
Richie Petitbon .8500
Trevon Sidney - .8500
Dalevon Campbell .8499
Joseph Thompson .8451
Tarique Barnes .8451
Devon Witherspoon .8400
Evan Kirts .8343
Nick Fedanzo .8366
Casey Washington .8323
Griffin Moore .8248
Josh Plohr .7998

2020
James Frenchie .8864
Reggie Love .8844
Brian Hightower .8800
Chase Brown .8700
Tre'Von Riggins .8613
Blake Jeresaty .8600
Cooper Davis .8567
Deuce Spann .8566
Johnny Newton .8519
Anthony Shipton .8506
Derrick Smith .8472
Phifer Griffin .8424
Lavar Gardner .8420
Kevin Tyler .8411
Christian Bell .8400
Quinton McCoy .8336
Brevyn Jones .8317
Blaise Sparks .8247
Tahveon Nicholson .8056
Desmond Dan .7783

OK, one thing left to do. Use the class calculator, find players in those classes with those ratings (the system won't allow you to add 2016 Josh Imatorbhebhe to the 2019 Illini class), add them, and then see where the class ranks once the transfers are added. For example, I can't add Luke Ford to the 2019 class, but I went to find a 2019 player with that same re-ranked player rating (Jacob Lacey who went to Notre Dame) and added him to the Illinois class. The results:

2019 original point total and ranking: 184.37, 53rd nationally
2019 point total and ranking once I added the seven transfers to the class (plus Witherspoon): 222.29, 27th nationally

2020 original point total and ranking: 152.90, 88th nationally
2020 point total and ranking once I added the seven transfers to the class: 182.53, 51st nationally

It worth noting that I'm a tiny bit premature because there's still another scholarship that will likely go to another transfer for 2020. If it's a Derrick Smith-like transfer (.8500-ish), the class moves to 49th. If it's a Luke Ford-like transfer (.9300-ish), the class moves to 45th.

That feels... pretty accurate to me? Adding guys like Ford and Peters to the class of Beason, Williams, and Cooper really should move it up near the top-25. Especially when you then have a class of 23 players, meaning their formula is maxed-out. And then the 2020 class, both in high school recruits and transfers, has been a step back. If I just look at the averages of the scores I used there, the average transfer player in 2019: .8814. That translates to a player ranked 500 or so (and using my basketball calculation - divide by 6.54 - that's like the 76th-best basketball recruit). The average transfer player this spring: .8438, which translates to a player ranked 1300 or so (a basketball recruit ranked 198th).

Yes, I know that I'm dropping Hightower from a .9366 to a .8800. And I'm dropping Bell from a .8800 to a .8400. But nether was able to crack the lineup at Miami or Wisconsin, so I don't think they can be rated like they were in high school. Also, we only get one year of service from Bell and two years from Hightower.

The other complication here before I wrap this up: other schools aren't getting credit for their transfers. No other school is bringing in 14 transfers in two years, but some are bringing in 3 and then 2. Add those transfers to those classes, re-rank them accurately, and some of this "the class climbs to 27th nationally" probably drops to "the class climbs to 31st nationally" (or whatever). But this wasn't an attempt to re-rank the entire 2019 and 2020 recruiting classes. I just wanted to see where our classes would land once the transfers were added.

And where do we land? I'd say 27th feels right for Lovie's 2019 class. Beason, Williams, Cooper, Ford, Peters, Bhebhe, Betiku, Sidney, Randolph, Coleman, Okpala - if you combine the transfers with the high schoolers, it's Lovie's best class.

And then 2020 was a big step back, even when you get the class up to 20 players so that the "it's only 13 players it can't be accurately ranked" is eliminated. Once that 21st player is added (some transfer in the next few months) it's going to be maybe the 49th-best class. Which means, using this "add the transfers" method, 2019 to 2020 almost doubled from 27 to 49.

Which is sustainable as long as the 2021 class is another big step forward.

{Robert has left the chat and has resumed clinging to hope}

Comments

IBFan on May 04 @ 06:43 PM CDT

Thank you. Thank you for all the leg work. Now if you can spread this logic across Illini nation the haters can move on to Lovie still can’t coach with decent classes and isn’t recruiting enough high schoolers who grew up in this cesspool of a state that can’t wait to get out. Seriously, great job. There are holes as you know as far as players not getting credit for summer camps or even senior year, the Illinois commitment reduction factor, etc. Classes still can get better. The staff loved Casey Washington and wanted him regardless of other people’s numbers, stars. As you have said before the staff is picking out the players that fit their style. Another speedy, long, athlete for the secondary again.

uilaw71 on May 05 @ 08:22 AM CDT

Adding to IBFAN’s point, when called out on Loyalty the Lovie haters are quick with denials and justifications. The measure of a program is how those who played for a coach speak of him. Of the 11 HCs in my lifetime, only 1 who wasn’t fired stands out from that perspective, and chances are few if any on here or Loyalty will agree on the name. So what’s my point? This program has come light years from where it was in the aftermath of the Slush Fund. Character-wise it is head and shoulders above where any other HC in that time has taken it. We may not have the talent today that others had for a fleeting moment, but Louie’s kids embody the Fight in Illini.

uofi08 on May 05 @ 10:15 AM CDT

Very good, freakishly informative article. I think you’re adjusted ratings definitely make sense. You can argue that Peters and Bhebhe actually played to their lofty prep rankings. I’m guessing transfers have never really been ranked just out of convenience and lack of information. Really hard to gage how good (or bad?) a player is after not seeing much playing time while on a college roster. I think nearly all of last years transfers out performed even the most optimistic predictions, outside of Sidney but he was injured.

In general I think recruiting rankings are more for a “feel” for how much talent a team is bringing in and not an exact science. The problem with trying to give transfers an accurate reranking is that the rest of the roster doesn’t get that same scrutiny. Players like Jake Hansen have outperformed their prep rankings. Someone like an Owen Carney has done the opposite. If you’re not going to do it to the rest of the roster, what difference does it make for a couple transfers? The instances where it really makes a difference is last years team, adding that magnitude of quantity and quality of transfers.

Robert on May 05 @ 02:27 PM CDT

That's true, and I considered leaving everyone's ranking alone because of it. Every team has busts and surprises, and when the dust settles, it probably all evens out back to your base talent level. Maybe a little higher for coaching staffs who are really great at HS talent evaluation (like Jerry Kill at Minnesota as I discussed last week).

But I'd also say that many of our transfers are here because they didn't live up to their ranking at their other school. I think it's fair to say that a roster of 85 players at any school will balance (as many "way better than their ranking coming in" as "way worse than their ranking" players, meaning it all evens out), but a collection of transfers will all weigh heavily in the "worse than their ranking" category. Just look at the players who transferred away from Illinois the last four years. Nearly every one of them was in the "worse" category, with most of them transferring down to G5 schools and FCS schools.

So if "transfer" generally means "didn't cut it at this level", then I think transfers should get a new rating in an exercise like this.

uofi08 on May 05 @ 02:45 PM CDT

100% agree. I guess that opens up the next question of how do you actually then “rate” a transfer, especially one with very little playing experience in college. At least with prep prospects there’s a couple years of tape, a history of recruiting interest from teams/coaches. Transfers are really wildcards where you hope history and physical attributes give you an indication. Just look at the QB comparisons between Illinois and northwestern last year. Going into the year most analysts assumed northwestern had the better transfer. NU had the 5 star transfer with a year to learn the system. Illinois has the 4 star with barely a summer to learn the system. Peters looked the part while Hunter Johnson looked like a walk-on.

Hoppy on May 05 @ 03:33 PM CDT

What would the 2020 class have looked like using your new ratings AND if we had been able to get Jadon Thompson through admissions?

pfruin@maynardcooper.com on May 05 @ 04:17 PM CDT

Good Read. The Team Talent Composite I think is the best measure I have reviewed that properly measures overall team talent. It is a one year snap shot. A perhaps better approach would be to weight the players on that based on class year. A .85 LB senior is more valuable this year than a .85 LB freshman. So from that standpoint, maybe transfers should not be lowered. Sure many have not lived up to hype, but you are also getting them after they have matured by a couple of years. So they are better than they were as freshman. It goes back to the what we say about O lineman. Redshirt them, and let me grow in obscurity for two years. They can fill in a bit as redshirt sophomores, and then watch them thrive in their 4th and 5th years. Richie Petitbon may not have been a .95 player, but he was a better O lineman than the average .85 lineman because he was a fifth year senior.

Speak Your Mind

Please login or register to post comments on the IlliniBoard.