Link: A better (?) way to look at college football's returning experience

Re: A better way to look at college football's returning experience

Wow, I didn't realize we would be that low on that scale. As long as Coach Saban is around I'm not concerned about #'s returning. LSU may return the most, but Les still has to coach and they have the same QB as last year.

Re: A better way to look at college football's returning experience

Interesting thinking - evocative of sabermetrics. His sample size is as yet too small, and his results at the high end too inconsistent, to make for a truly compelling, predictive model, but perhaps he, or others, can refine and improve it.

Several years ago, The Wall Street Journal found that the best pre-season predictor of success was cumulative games started by returning o-linemen. I wonder if that's still valid?

Re: A better way to look at college football's returning experience

Isaiah 63:1 said:

Interesting thinking - evocative of sabermetrics. His sample size is as yet too small, and his results at the high end too inconsistent, to make for a truly compelling, predictive model, but perhaps he, or others, can refine and improve it.

Several years ago, The Wall Street Journal found that the best pre-season predictor of success was cumulative games started by returning o-linemen. I wonder if that's still valid?

Click to expand...

Somehow, "We're number 104!" does not have the same ring to it as "We're Number 1!"

Re: A better way to look at college football's returning experience

crimsonaudio said:

...Only time will tell if there is really any correlation here...

Click to expand...

The thing is, he has plenty of additional data, from past seasons. Why didn't he go back, say, 10 years and see if his model was ex post facto predictive? If it wasn't, revise the model. If it was, go back 25 and reassess. But i suppose that would've required actual, you know, effort...

Re: A better way to look at college football's returning experience

To anyone who thinks returning starters is a big deal......the 2000 Crimson Tide that were the defending SEC champions returned 16 starters from a 10-3 top ten team and were picked by many to win it all.

Of course, it was two starters they lost....Chris Samuels and Shaun Alexander (who made up a mere 73% of the offense in 1999).........that were the bigger issue.

Most years we return fewer starters than the rest of the SEC because:
a) our number of starters doesn't matter because we rotate so efficiently......some of our backups have more playing time by their sophomore year than some juniors at big schools
b) we lose more folks early to the NFL draft

I'd have to look at the data, too, but I suspect a number of our starters don't actually BECOME starters until mid-way through their junior year.

I guess it HAS hurt us, though.......we did LOSE the national championship in 2010, 2013, and 2014 so...........

Re: A better way to look at college football's returning experience

Football's numbers are always small enough to make statistical significance difficult. In something like this, you look for the extremes, and then look to see if that factor might hook up with an opportunity.

One thing that is interesting is that, for an LSU, it doesn't actually matter how many points improvement they show in 10-11 of their games. It really just comes down to whether they can beat Bama. To that end, Alabama's recruiting success and on-field track record suggests that this sort of analysis is less telling for us than it likely is for a number of other teams.

In short: cool analysis, but (like most things) not something that should be used on its own to draw conclusions.

Re: A better way to look at college football's returning experience

Isaiah 63:1 said:

The thing is, he has plenty of additional data, from past seasons. Why didn't he go back, say, 10 years and see if his model was ex post facto predictive? If it wasn't, revise the model. If it was, go back 25 and reassess. But i suppose that would've required actual, you know, effort...

Click to expand...

Because that's more work than one person can do at one time? He does so much other stats stuff he can't just decide to go that far back and have it figured out.

Will point out an issue with this comment from one of his opening graphs:

With just one year at my disposal (2014) when I wrote the original post, I wasn't able to draw any lasting conclusions, but the potential was obvious. I put together a rough formula that posited that teams like North Carolina (87 percent of production returning), Ohio (85 percent), and Temple (83 percent) were among the most experienced in the country, then watched as those teams improved from a combined 18-9 to a combined 29-12. (Others near the top didn't improve nearly that much.)

Click to expand...

Sagarin ranks those three teams' schedules 59 (UNC), 76 (Temple), and 104 (Ohio). Would like to know how much the schedule variable plays into it.

RTR91 said:

Because that's more work than one person can do at one time? He does so much other stats stuff he can't just decide to go that far back and have it figured out...

Click to expand...

Didn't Bill James derive and test the original Pythagorean Expectation, based on years of data, by himself, with pen and paper while guarding canned food at night?

Re: A better way to look at college football's returning experience

selmaborntidefan said:

To anyone who thinks returning starters is a big deal

Click to expand...

It is a big deal for most teams. There's just two big factors...
A: Returning starters can't make a bad coach into a good one.
B: Nick Saban defies the returning starter thing.

Two things though, LSU will be scary good. This will be the best LSU team in a long time. He's always been loaded with talent, but LSU has had attrition problems comparable to Alabama. Somehow he held it together though and the result is a team that will be great unless something happens to make the wheels fall off. On the other hand, Ohio State is decimated and Urban hasn't shown the ability to contend with a completely rebuilt roster. I'd expect LSU to beat Ohio State handily for instance.

Re: A better way to look at college football's returning experience

KrAzY3 said:

Two things though, LSU will be scary good. This will be the best LSU team in a long time. He's always been loaded with talent, but LSU has had attrition problems comparable to Alabama. Somehow he held it together though and the result is a team that will be great unless something happens to make the wheels fall off.

Click to expand...

LSU has to prove that they have a real offensive game plan and a QB capable of executing it before I worry about them beating Alabama with your run defense. They will likely be the most talented team in the country in week one, but they don't play Alabama in week one. By the time that they play you guys, you will have a team of very experienced players.

As for OSU - serious rebuilding year, but I still like our chances of winning the B1G which I think will take a step backwards over-all this season.

Re: A better way to look at college football's returning experience

B1GTide said:

LSU has to prove that they have a real offensive game plan and a QB capable of executing it before I worry about them beating Alabama with your run defense. They will likely be the most talented team in the country in week one, but they don't play Alabama in week one. By the time that they play you guys, you will have a team of very experienced players.

As for OSU - serious rebuilding year, but I still like our chances of winning the B1G which I think will take a step backwards over-all this season.

Click to expand...

Let's hope that Alabama has the offense figured out by the time they play LSU, they'll need it. LSU and Alabama both are interesting in that they've had a lot of success in recent years without much in the way of NFL talent at the QB position. Les finds a way to muck things up, but don't forget he had one of the best BCS teams ever with a big question mark at QB. This is probably going to be the best LSU team since that 2011 team. I'd wager they can go a long way with just the running back and defense, which doesn't sound that much different from what Alabama did in 2009 for instance.

I'd agree about a chance for Ohio State to win the conference. The talent is there, the coach is there, but in order for people to have expectations beyond that I think they have to overlook the lack of experience all over the field.

Re: A better way to look at college football's returning experience

KrAzY3 said:

It is a big deal for most teams.

Click to expand...

I agree (likely with you totally) UP TO A CERTAIN POINT it does matter. But that's where other factors offset the advantage, including two you cited.

KrAzY3 said:

Two things though, LSU will be scary good. This will be the best LSU team in a long time. He's always been loaded with talent, but LSU has had attrition problems comparable to Alabama. Somehow he held it together though and the result is a team that will be great unless something happens to make the wheels fall off. On the other hand, Ohio State is decimated and Urban hasn't shown the ability to contend with a completely rebuilt roster. I'd expect LSU to beat Ohio State handily for instance.

Click to expand...

LSU returns 18 starters this year......that's after returning 17 last year.

They DO look VERY good on paper, but they're going to need a QB.

Re: A better way to look at college football's returning experience

This reminds me of the old joke with Vanderbilt football:

The good news for Vandy is that they return 20 starters from a 2-10 football team.

The bad news for Vandy is that they return 20 starters from a 2-10 football team.

Re: A better way to look at college football's returning experience

I love meaningful, important stats but I am not sure of this one yet.

I think the bottom line is something most Bama fans have known and felt since 2011. LSU will have a very good team - and on any given Saturday they could put the pieces together and have a good chance at beating any team in the nation, including Alabama.

It will be easier to make a prediction the week before the game. QB or no QB, when we play LSU, we had better show up.