"Here's what to expect" has never been harder. And as someone whose words are 61% "here's what to expect", it's a struggle right now.

Not just for me. Bill Connelly released his updated Returning Production and revised preseason SP+ projections today, and he talked about the same thing. How do you put together a preseason formula when the dynamics of roster construction have changed so drastically?

For me, part of the struggle is with this revised LLUOI process I chose this year. I say that I want to cover things position by position (and I already did so for the quarterbacks), but as soon as I started writing out the incoming offensive linemen I put it on hold because we had transfer linemen visiting nearly every weekend. It's May 25th and we still don't know the full landscape for the incoming offensive linemen.

Here, I'll give you an example. Just as I started writing this article, Bret Bielema tweeted one of those "good news coming soon" tweets. Any other year (even last year, probably), I'd assume that this signal is for a recruit in the upcoming class. It's May 25th, assistants have been on the road for six weeks, June camps are coming up - yeah, this is a 2024 guy.

But it's probably a 2023 guy? Because May is now about landing players in the portal and not landing high schoolers for 2024. I'm not even sure the staff has fully defined the needs for 2024 yet because the 2023 dust hasn't settled. It's crazy to say that on MAY 25TH but it's true.

Let's discuss it through the eyes of Connelly's attempt to rank incoming classes. His preseason SP+ projections have always been three things: What's your recent history, how much of that production did you lose, and how impactful have your recruiting classes been? But then the portal came along, and teams like USC last year came along (go grab the Heisman winning QB and the best college WR from the portal and instantly improve your team), so how does one rank incoming classes fairly?

So he devised the way he ranks incoming classes, creating a combo ranking of freshman classes + transfers. And those transfers come with built-in experience, so how does that play into this? USC and LSU proved last year that you can grab high-impact players from the portal and instantly improve, but didn't Georgia also prove that recruiting the best high school classes every year and developing them over 3-4 years is the best approach?

I guess I should let him explain his conundrum (from today's article):

The transfer portal has made it a lot easier for teams -- at least those pretty high up in the rankings -- to add experienced players and field more complete teams. That's not to say everyone uses the portal equally well, but now that I've had time to play with data from the transfer-heavy 2022 season, recruiting rankings for transfers have become a solid-sized piece of SP+ projections.

That means intriguing things for a team that (A) has won the past two national titles and (B) hasn't dipped into the portal all that much.

Kirby Smart used the portal to add a couple of proven wide receivers (Missouri's Dominic Lovett and Mississippi State's Rara Thomas), a young running back (Tennessee's Len'Neth Whitehead, who carried 32 times in 2021) and a young cornerback (Texas A&M's Smoke Bouie, a Georgia native and top-70 recruit in 2021). Considering 25 teams have added at least 15 transfers and 53 have added at least 10, that's not much of a haul in terms of either quantity or known production.

That has created an interesting effect. When it comes to the impact teams' recruiting hauls will specifically have on their 2023 performance, Georgia's class ranks just 22nd overall despite their freshman haul ranking second

2023 recruiting classes, ranked by potential 2023 impact:

1. LSU
2. Oklahoma
3. Oregon
4. Miami
5. Texas A&M
6. Florida
7. Ohio State
8. Auburn
9. USC
10. Michigan

That frames all of this really well, I think. He's looking for a better way to quantify teams bringing in transfers (players who will make a more-immediate impact than an 18 year-old freshman) because that better speaks to this coming season. Georgia's fantastic high school class means they'll be really good in 2025 (just like their fantastic 2020 class means they'll be really good this year), but for a team like LSU, they're building for the now, not the future (and banking on being able to do that every year after their rental players depart).

So on one hand, saying that Georgia bringing in the 22nd-best recruiting class is insane. And on the other hand, it makes sense because they bought in a ton of high schoolers who won't have much of an impact on this season. Which is what he's trying to do with these preseason numbers. But did his formula build in enough "Georgia's 2020 and 2021 classes were outstanding and those players about to break out in a big way"?

I'll give you an example from his revised preseason SP+ ratings. In February we were #44, and right now we're still at 44. But some teams who gained a lot of transfers since then moved up (like Michigan State, who jumped from 47 to 42 by adding a bunch of "returning production" through transfers. And then there's Cincinnati which lost their QB to Northwestern and dropped from 43rd to 46th.

I'm not suggesting that this is some huge deal. "Oh no, Cincinnati dropped three spots in some preseason formula!" I'm suggesting that preseason formulas have some hurdles to navigate with all these changes. That might not mean much to you, but I guarantee it means a lot to Vegas. I agree wholeheartedly with this tweet that I saw today:

This, from Bill Connelly, is exactly why the Portal is great for bettors.

If he's struggling to quantify it, then so, too, are the books.

There will be significant early-season edges -- arguably as large as any season in recent memory. pic.twitter.com/EhujlweS2g

— JAF (@JonFendler) May 25, 2023

That's my whole point here. Nobody knows. And it's so rare that nobody knows. We're in the portal era now, and we don't have enough data to say how this will all play out.

If teams lean too heavily into the portal, do they kill development and continuity? If teams rely only on freshmen when they have a position of desperate need, are they hurting themselves by not adding a 22-year old senior in that spot? And will Colorado win any games at all with 50+ transfers coming in? NFL expansion teams always struggle because nobody knows what the guy next to him is doing - won't that be the case for Colorado?

If you want to know how the Big Ten is approaching this issue, look no further than the number of transfers added this offseason. We have added six - Nicario Harper, Luke Altmyer, Tanner Arkin, Denzel Daxon, Demetrius Hill, and Clayton Bush (QB John Paddock is a walkon) - and here's how that stacks up with the rest of the transfers added in the Big Ten:

1. Indiana - 23
T2. Purdue - 14
T2. Wisconsin - 14
T4. Michigan State - 13
T4. Nebraska - 13
6. Minnesota - 11
7. Maryland - 9
T8. Iowa - 8
T8. Michigan - 8
T8. Ohio State - 8
11. Penn State - 7
T12. Illinois - 6
T12. Rutgers - 6
14. Northwestern - 3

You can't really frame it better than that. Indiana went 6-18 the last two seasons. Northwestern went 4-20. Both need to turn things around quickly. Tom Allen said "to the portal!" and has added 23 new players (in addition to a high school class of 15 players). Pat Fitzgerald said "keep developing in our system!" and only added a QB and two wide receivers from the portal. Should Allen have thought more about the future than the present? Should Fitzee have thought more about the present than the future? Nobody knows! That's the whole thing!

I think it's fairly obvious from the T12 there that Bret Bielema leans in the "development" direction, not the "quick fix" direction. He's said that exact thing out loud several times ("we're a developmental program"). But that brings us back to the Georgia conundrum. We obviously don't recruit anywhere close to the level that Georgia recruits, but the concept holds. When developing a preseason formula in an attempt to predict what might happen in the fall (and this isn't just guys like Connelly - as referenced above, Vegas sportsbooks use similar formulas to set their lines), how do you balance "added a bunch of instant returning production" vs. programs that are choosing to play the long game? If you just want the caffeine hit, Folgers crystals or french press?

Let's play out a scenario. You need to replace catches from your wide receiver who just graduated. Do you replace those catches with A) a player who has been with you for several years but doesn't have any on-the-field experience or B) a player who got that on-the-field experience elsewhere but doesn't know your system yet?

That's what Connelly is trying to figure out with his formula. Returning experience is maybe the best indicator of "this season will be better than last season" because all players have growing pains when they're first out there on the field. So if a school brings in an experienced player who has already worked out those growing pains elsewhere, might that be better than a redshirt sophomore on your roster who might have a rough first season on the field? But if that redshirt sophomore knows everything you're trying to do and has been growing within your culture, learning from those who went before, isn't THAT more important?

We can see that playing out with Calvin Hart's departure. With two sophomore linebackers (Kenenna Odeluga and Dylan Rosiek) pushing for playing time (and often playing in front of Hart in scrimmage situations), Hart chose to transfer out after spring ball. He'll play his final season at Oregon State next year. But that transfer out dropped us from 71st (in the February Returning Production rankings) to 86th in the RP rankings released today (losing a starting linebacker is a decent hit, historically).

It was the same with Brian Hightower. We took a "returning production" hit when he transferred to Cal because losing a receiver with 37 catches is hard to replace. But we kind of know why he transferred...

Catches first 7 games: Brian Hightower 25, Casey Washington 10
Catches final 6 games: Casey Washington 21, Brian Hightower 12

Hightower then transferred out, giving us a returning production hit, but was that just a case of another guy passing him on the depth chart?

I just had that feeling where I realized I've made my point and now I'm digging too deep. You get what I'm saying, I think. I should wrap it up with a recap.

The transfer portal is very important. Last year, our quarterback transfer from Syracuse handed the ball off to our running back transfer from Western Michigan and it helped us get to eight wins. Northwestern is the Big Ten team that uses the portal the least, and even they went out and added a QB and two receivers this offseason because they're passing offense was just so horrific last year. You can't NOT use it.

But can you use it too much? Is Indiana smart or crazy for adding 23 transfers? And how do we predict what the season might hold if the old formula (recruit well, develop those players in your system) has been blown apart? (Or is it still intact and the coaches ignoring portal madness are the coaches who will come out on top after the dust settles?)

I guess I'm only complaining that it's hard to predict things now. The season will be what the season will be. We either approached it correctly or we didn't.

(But seriously - what's the best way to approach it?)