[CTcard]

I remember that a year ago Bill Connelly (Football Outsiders and SB Nation stat guy) was developing a system that measured returning production on offense and defense as a better measure than returning starters.

Here is this year's installment for the 2016 season. It's a couple months old, but I missed it before.
http://www.sbnation.com/college-football...e-stanford

Stanford gets a couple of particular mentions

Do you have Ohio State and Stanford in your top 10?
Because if so ... well ... consider this a red flag.

Stanford comes in at #115 of 128 teams in total returning production.
Sort of hard to believe with McCaffrey returning, but he also claims returning RBs matter less than returning receivers.

And that seems to be after this correction ...

Note: This post was updated to account for some roster errors. Michigan, Notre Dame, and Stanford do not list redshirts, so some of the players listed as seniors are juniors, juniors are sophomores, etc. It's annoying, and it manages to trip me up every damn year.

[CTcard]

I remember that a year ago Bill Connelly (Football Outsiders and SB Nation stat guy) was developing a system that measured returning production on offense and defense as a better measure than returning starters.

Here is this year's installment for the 2016 season. It's a couple months old, but I missed it before.
http://www.sbnation.com/college-football...e-stanford

Stanford gets a couple of particular mentions

Do you have Ohio State and Stanford in your top 10?
Because if so ... well ... consider this a red flag.

Stanford comes in at #115 of 128 teams in total returning production.
Sort of hard to believe with McCaffrey returning, but he also claims returning RBs matter less than returning receivers.

And that seems to be after this correction ...

Note: This post was updated to account for some roster errors. Michigan, Notre Dame, and Stanford do not list redshirts, so some of the players listed as seniors are juniors, juniors are sophomores, etc. It's annoying, and it manages to trip me up every damn year.

[BostonCard]

Two comments...

Sort of hard to believe with McCaffrey returning, but he also claims returning RBs matter less than returning receivers.

1) A returning running back probably matters more at schools with a run-first offense (like Stanford) than in pass-happy schools like, say, WSU, where I sometimes wonder if they even use a running back.  If he applies the same weight to all schools, there's a problem.  A better iteration of this would be to weight according to the percentage of running plays.

1A) In McCaffrey we bring back our top running back AND our top receiver; that's probably not being considered in the analysis, since in pretty much every other team, "returning receiver" means "returning wide receiver and maybe tight end".

2) Stanford's roster is meant to reflect the fact that our student athletes are students first, and they are listed by their academic class.  It does not exist to make the life of Connelly or other lazy sports journalists easier.  It rightfully emphasizes how many years they have been in school, and de-emphasizes eligibility, which is only tangentially related to graduating.  That being said, it would be nice if Stanford had a way to denote the latter as well.

BC

[BostonCard]

Two comments...

Sort of hard to believe with McCaffrey returning, but he also claims returning RBs matter less than returning receivers.

1) A returning running back probably matters more at schools with a run-first offense (like Stanford) than in pass-happy schools like, say, WSU, where I sometimes wonder if they even use a running back.  If he applies the same weight to all schools, there's a problem.  A better iteration of this would be to weight according to the percentage of running plays.

1A) In McCaffrey we bring back our top running back AND our top receiver; that's probably not being considered in the analysis, since in pretty much every other team, "returning receiver" means "returning wide receiver and maybe tight end".

2) Stanford's roster is meant to reflect the fact that our student athletes are students first, and they are listed by their academic class.  It does not exist to make the life of Connelly or other lazy sports journalists easier.  It rightfully emphasizes how many years they have been in school, and de-emphasizes eligibility, which is only tangentially related to graduating.  That being said, it would be nice if Stanford had a way to denote the latter as well.

BC

[CTcard]

1A) In McCaffrey we bring back our top running back AND our top receiver; that's probably not being considered in the analysis, since in pretty much every other team, "returning receiver" means "returning wide receiver and maybe tight end".

That's not quite what he's doing.

He looks at returning rush yards, returning receiving yards, returning passing yards.
So, McCaffrey is accounted for - assuming no mistakes.


He then correlates these things with S&P+ numbers the ensuing season.

It's probably true that returning rushing yards matter more for a primarily running team, but trying to account for the particulars of each team would wipe out the statistcal nature of the statistics.

[CTcard]

1A) In McCaffrey we bring back our top running back AND our top receiver; that's probably not being considered in the analysis, since in pretty much every other team, "returning receiver" means "returning wide receiver and maybe tight end".

That's not quite what he's doing.

He looks at returning rush yards, returning receiving yards, returning passing yards.
So, McCaffrey is accounted for - assuming no mistakes.


He then correlates these things with S&P+ numbers the ensuing season.

It's probably true that returning rushing yards matter more for a primarily running team, but trying to account for the particulars of each team would wipe out the statistcal nature of the statistics.

[stupac2]

We lost our QB, top TE, one of our top WRs, three starting OL, a RB TD vulture, most of our starting DL, a few key LBs, a starting safety, plus probably some guys I'm forgetting. It's not unreasonable to be down on us. Every team that's replacing a QB should be treated with skepticism, except I guess Alabama who can win with a blind hobo playing QB.

Most places have us in the 10-20 range (but still 1-3 in the Pac-12) which feels right to me.

[stupac2]

We lost our QB, top TE, one of our top WRs, three starting OL, a RB TD vulture, most of our starting DL, a few key LBs, a starting safety, plus probably some guys I'm forgetting. It's not unreasonable to be down on us. Every team that's replacing a QB should be treated with skepticism, except I guess Alabama who can win with a blind hobo playing QB.

Most places have us in the 10-20 range (but still 1-3 in the Pac-12) which feels right to me.

[CTcard]

We lost our QB, top TE, one of our top WRs, three starting OL, a RB TD vulture, most of our starting DL, a few key LBs, a starting safety, plus probably some guys I'm forgetting.

True overall, Stanford has lots of graduation losses (plus transfer).

As far as Connelly's stats go we lost 97% of our passing yards (correlation 0.264), lost 33% of our receiving yards (correlation 0.285), and 30% of our rushing yards (correlation 0.079). OL losses don't figure in at all because he found no correlation between returning OL and S&P numbers.

While those are significant losses, it's hard to believe only 8 teams of 128 lost as much or more. Makes me wonder if he actually did make the corrections with respect to Stanford listing players by academic standing rather than eligibility.

Equally surprising to me is that Connelly has Stanford with much more returning production on defense (61% vs 33%), despite big losses in his key stats of pass break ups (lose 45%, correlation 0.440), tackles (41%, correlation 0.388), sacks (lose 49%, correlation 0.194). I guess losing 97% of passing yards really dominates.

[CTcard]

We lost our QB, top TE, one of our top WRs, three starting OL, a RB TD vulture, most of our starting DL, a few key LBs, a starting safety, plus probably some guys I'm forgetting.

True overall, Stanford has lots of graduation losses (plus transfer).

As far as Connelly's stats go we lost 97% of our passing yards (correlation 0.264), lost 33% of our receiving yards (correlation 0.285), and 30% of our rushing yards (correlation 0.079). OL losses don't figure in at all because he found no correlation between returning OL and S&P numbers.

While those are significant losses, it's hard to believe only 8 teams of 128 lost as much or more. Makes me wonder if he actually did make the corrections with respect to Stanford listing players by academic standing rather than eligibility.

Equally surprising to me is that Connelly has Stanford with much more returning production on defense (61% vs 33%), despite big losses in his key stats of pass break ups (lose 45%, correlation 0.440), tackles (41%, correlation 0.388), sacks (lose 49%, correlation 0.194). I guess losing 97% of passing yards really dominates.

[winflop]

Keep Sleeping on Stanford...

[winflop]

Keep Sleeping on Stanford...

[Leftcoast]

KSOS - I like it.  We can use this after every begrudging article or comment.

[Leftcoast]

KSOS - I like it.  We can use this after every begrudging article or comment.

[Hulk01]

Silly.

As Matt nicely points out on TOS, what matters is who you have, not who you don't. 

Stanford lost 97% of its passing yards.  Well, Oregon lost Marcus Mariota and, if Vernon Adams hadn't gotten hurt, the W-L impact would not have exceeded two games. 

We are replacing a first team All Con TE with one who will be, very soon, even better.

We are replacing a very productive QB with a QB who, for the first time in five years, gives us a short, intermediate, and wide passing game, and a brilliant short, intermediate, and wide route runner (Trent Irwin) with near-perfect hands and a 6'7" catch radius.

We are backing up the Voldemort Trophy runner up running back with a guy who compiled nearly identical numbers in his first year, and who may be the second best running back in the conference--and yes, that includes Royce Freeman.  Love is much more dangerous--and big plays really really matter.

Chryst/Burns throwing and running, Mac/Love running, Irwin/Rector/Schultz/Mac receiving, Marx/Williams pounding when we have to pound, the long-awaited return of a total passing game, red zone weapons (Taboada, Schultz, and how about Rector's back of the end zone grab in the Spring game).  If the OL is merely serviceable, this team will hit 35 points at half time several times.  As a former DB, an offense like this--although I never saw one remotely like it--would scare the hell out of me.  I'd want two more guys out there with me.

If the OL--watch out. 

[Hulk01]

Silly.

As Matt nicely points out on TOS, what matters is who you have, not who you don't. 

Stanford lost 97% of its passing yards.  Well, Oregon lost Marcus Mariota and, if Vernon Adams hadn't gotten hurt, the W-L impact would not have exceeded two games. 

We are replacing a first team All Con TE with one who will be, very soon, even better.

We are replacing a very productive QB with a QB who, for the first time in five years, gives us a short, intermediate, and wide passing game, and a brilliant short, intermediate, and wide route runner (Trent Irwin) with near-perfect hands and a 6'7" catch radius.

We are backing up the Voldemort Trophy runner up running back with a guy who compiled nearly identical numbers in his first year, and who may be the second best running back in the conference--and yes, that includes Royce Freeman.  Love is much more dangerous--and big plays really really matter.

Chryst/Burns throwing and running, Mac/Love running, Irwin/Rector/Schultz/Mac receiving, Marx/Williams pounding when we have to pound, the long-awaited return of a total passing game, red zone weapons (Taboada, Schultz, and how about Rector's back of the end zone grab in the Spring game).  If the OL is merely serviceable, this team will hit 35 points at half time several times.  As a former DB, an offense like this--although I never saw one remotely like it--would scare the hell out of me.  I'd want two more guys out there with me.

If the OL--watch out. 

[stupac2]

All I know is that until I see Burns/Chryst playing well in a hostile environment, I am fine with people being skeptical about us. How many teams do you think should get the benefit of the doubt when changing QBs? Alabama, OSU, maybe FSU? It's not many. This team has the chance to be very, very good next year. But it's still just a chance, a lot needs to go right for it to happen.

All I know is that I don't think any pre-season has been this optimistic since 2011.

[stupac2]

All I know is that until I see Burns/Chryst playing well in a hostile environment, I am fine with people being skeptical about us. How many teams do you think should get the benefit of the doubt when changing QBs? Alabama, OSU, maybe FSU? It's not many. This team has the chance to be very, very good next year. But it's still just a chance, a lot needs to go right for it to happen.

All I know is that I don't think any pre-season has been this optimistic since 2011.

[CTcard]

As Matt nicely points out on TOS, what matters is who you have, not who you don't.

This is actually a theme Matt and I developed years ago over on the Bootleg.

It's certainly true in assessing individual teams, but it's not very practical for incorporating into statistical models. Some of those models try to use recruiting rankings as a way to account for this effect, but I don't believe that either the service evaluations themselves, or the method of ranking, is sophisticated enough to cover the "who you have" question.

This is why even a good statistical model might be pretty good overall, but not have that much relevance for any one particular team.

Stanford lost 97% of its passing yards.  Well, Oregon lost Marcus Mariota and, if Vernon Adams hadn't gotten hurt, the W-L impact would not have exceeded two games. 

Well, presumably this is one of the reasons that returning QB is not as high on the correlation chart as one would think. However, in general, not having a good QB is obviously a brobdingnagian hit to any team. Last year Oregon lucked into a very good QB (at least when healthy). You saw how bad Oregon was when they didn't have a very good QB, i.e. when Adams was out. Thus the danger.

We are replacing a first team All Con TE with one who will be, very soon, even better.

What we are really doing is replacing the TE duo of 1.Hooper, 2.Schultz with the duo of 1.Schultz, 2.Taboada. That seems just fine to me, but we don't actually know how they will do yet.

We are replacing a very productive QB with a QB who, for the first time in five years, gives us a short, intermediate, and wide passing game, and a brilliant short, intermediate, and wide route runner (Trent Irwin) with near-perfect hands and a 6'7" catch radius.

We're replacing Hogan with Irwin? [wink]

Obviously our QB will be different this year. Quite likely stronger at some aspects. BUT, it's a pretty tall order to assume QB overall will be an upgrade this year.

We are backing up the    Trophy runner up running back with a guy who compiled nearly identical numbers in his first year, and who may be the second best running back in the conference--and yes, that includes Royce Freeman.  Love is much more dangerous--and big plays really really matter.

Love is a drastically different kind of player than Freeman. Calling him better is certainly assuming a lot. Among other things, in comparison, Love has hardly played.

Love is much quicker, and probably more likely to take it to the house on any given play.

But, Love also goes down fairly easily when he doesn't have space - see the number of zero to negative plays in the spring game. This is where he is well behind McCaffrey, and Freeman for that matter.

If the OL--watch out.

Yeah, every year we can go down our roster during the spring and muse about how wonderful it is. The most important thing is "If the OL", though I would combine that with "If the QB".

The two most important position groups on offense are the QB and the OL. We have turnover at both. We have reason to be optimistic for both, but nothing is for sure until it happens.

-------
By the way, the other team Connelly warns about is Ohio State, which returns substantially less production than we do. I don't know anything specific about Ohio State's roster, but I am sure it is full of inherently talented athletes. Will be interesting to see how they turn out this year.

[CTcard]

As Matt nicely points out on TOS, what matters is who you have, not who you don't.

This is actually a theme Matt and I developed years ago over on the Bootleg.

It's certainly true in assessing individual teams, but it's not very practical for incorporating into statistical models. Some of those models try to use recruiting rankings as a way to account for this effect, but I don't believe that either the service evaluations themselves, or the method of ranking, is sophisticated enough to cover the "who you have" question.

This is why even a good statistical model might be pretty good overall, but not have that much relevance for any one particular team.

Stanford lost 97% of its passing yards.  Well, Oregon lost Marcus Mariota and, if Vernon Adams hadn't gotten hurt, the W-L impact would not have exceeded two games. 

Well, presumably this is one of the reasons that returning QB is not as high on the correlation chart as one would think. However, in general, not having a good QB is obviously a brobdingnagian hit to any team. Last year Oregon lucked into a very good QB (at least when healthy). You saw how bad Oregon was when they didn't have a very good QB, i.e. when Adams was out. Thus the danger.

We are replacing a first team All Con TE with one who will be, very soon, even better.

What we are really doing is replacing the TE duo of 1.Hooper, 2.Schultz with the duo of 1.Schultz, 2.Taboada. That seems just fine to me, but we don't actually know how they will do yet.

We are replacing a very productive QB with a QB who, for the first time in five years, gives us a short, intermediate, and wide passing game, and a brilliant short, intermediate, and wide route runner (Trent Irwin) with near-perfect hands and a 6'7" catch radius.

We're replacing Hogan with Irwin? [wink]

Obviously our QB will be different this year. Quite likely stronger at some aspects. BUT, it's a pretty tall order to assume QB overall will be an upgrade this year.

We are backing up the    Trophy runner up running back with a guy who compiled nearly identical numbers in his first year, and who may be the second best running back in the conference--and yes, that includes Royce Freeman.  Love is much more dangerous--and big plays really really matter.

Love is a drastically different kind of player than Freeman. Calling him better is certainly assuming a lot. Among other things, in comparison, Love has hardly played.

Love is much quicker, and probably more likely to take it to the house on any given play.

But, Love also goes down fairly easily when he doesn't have space - see the number of zero to negative plays in the spring game. This is where he is well behind McCaffrey, and Freeman for that matter.

If the OL--watch out.

Yeah, every year we can go down our roster during the spring and muse about how wonderful it is. The most important thing is "If the OL", though I would combine that with "If the QB".

The two most important position groups on offense are the QB and the OL. We have turnover at both. We have reason to be optimistic for both, but nothing is for sure until it happens.

-------
By the way, the other team Connelly warns about is Ohio State, which returns substantially less production than we do. I don't know anything specific about Ohio State's roster, but I am sure it is full of inherently talented athletes. Will be interesting to see how they turn out this year.

[Hank 91]

Silly.

As Matt nicely points out on TOS, what matters is who you have, not who you don't. 

Stanford lost 97% of its passing yards.  Well, Oregon lost Marcus Mariota and, if Vernon Adams hadn't gotten hurt, the W-L impact would not have exceeded two games. 

We are replacing a first team All Con TE with one who will be, very soon, even better.

We are replacing a very productive QB with a QB who, for the first time in five years, gives us a short, intermediate, and wide passing game, and a brilliant short, intermediate, and wide route runner (Trent Irwin) with near-perfect hands and a 6'7" catch radius.

We are backing up the    Trophy runner up running back with a guy who compiled nearly identical numbers in his first year, and who may be the second best running back in the conference--and yes, that includes Royce Freeman.  Love is much more dangerous--and big plays really really matter.

Chryst/Burns throwing and running, Mac/Love running, Irwin/Rector/Schultz/Mac receiving, Marx/Williams pounding when we have to pound, the long-awaited return of a total passing game, red zone weapons (Taboada, Schultz, and how about Rector's back of the end zone grab in the Spring game).  If the OL is merely serviceable, this team will hit 35 points at half time several times.  As a former DB, an offense like this--although I never saw one remotely like it--would scare the hell out of me.  I'd want two more guys out there with me.

If the OL--watch out.

I have nothing to say about this. I'm only quoting it here so that it appears more often. Wouldn't want anyone to miss it.

[Hank 91]

Silly.

As Matt nicely points out on TOS, what matters is who you have, not who you don't. 

Stanford lost 97% of its passing yards.  Well, Oregon lost Marcus Mariota and, if Vernon Adams hadn't gotten hurt, the W-L impact would not have exceeded two games. 

We are replacing a first team All Con TE with one who will be, very soon, even better.

We are replacing a very productive QB with a QB who, for the first time in five years, gives us a short, intermediate, and wide passing game, and a brilliant short, intermediate, and wide route runner (Trent Irwin) with near-perfect hands and a 6'7" catch radius.

We are backing up the    Trophy runner up running back with a guy who compiled nearly identical numbers in his first year, and who may be the second best running back in the conference--and yes, that includes Royce Freeman.  Love is much more dangerous--and big plays really really matter.

Chryst/Burns throwing and running, Mac/Love running, Irwin/Rector/Schultz/Mac receiving, Marx/Williams pounding when we have to pound, the long-awaited return of a total passing game, red zone weapons (Taboada, Schultz, and how about Rector's back of the end zone grab in the Spring game).  If the OL is merely serviceable, this team will hit 35 points at half time several times.  As a former DB, an offense like this--although I never saw one remotely like it--would scare the hell out of me.  I'd want two more guys out there with me.

If the OL--watch out.

I have nothing to say about this. I'm only quoting it here so that it appears more often. Wouldn't want anyone to miss it.

[oman]

I'm naturally gloom and doom, and so I am going to be worrying about OL until three or four games in (when we will know a whole heck of a lot)

AND I wish we could switch the Rice game with the K-State game, doggone it.

[oman]

I'm naturally gloom and doom, and so I am going to be worrying about OL until three or four games in (when we will know a whole heck of a lot)

AND I wish we could switch the Rice game with the K-State game, doggone it.

[fullmetal]

I bet K-State could be thinking the same thing.

[fullmetal]

I bet K-State could be thinking the same thing.

[winflop]

KSOS - I like it.  We can use this after every begrudging article or comment.

You'd think they would learn after six years of doubting and being dead wrong five of those six years.

Then again, look at the history of the Axe and it's clear that the continuing doubters went to Low APR U.  They learned nothing after a few decades...

"We'd like to take some pictures"

[winflop]

KSOS - I like it.  We can use this after every begrudging article or comment.

You'd think they would learn after six years of doubting and being dead wrong five of those six years.

Then again, look at the history of the Axe and it's clear that the continuing doubters went to Low APR U.  They learned nothing after a few decades...

"We'd like to take some pictures"

[Jabbahop]

How important is the QB's ability to get the team into the right play in terms of production. Seems like Shaw called out Luck and Hogan as being excellent in that regard. My question/worry is that the offense steps back not because our new QB isn't effective throwing/running but in reading the defense and getting us into the optimal play.

Thoughts?

[Jabbahop]

How important is the QB's ability to get the team into the right play in terms of production. Seems like Shaw called out Luck and Hogan as being excellent in that regard. My question/worry is that the offense steps back not because our new QB isn't effective throwing/running but in reading the defense and getting us into the optimal play.

Thoughts?

[stupac2]

How important is the QB's ability to get the team into the right play in terms of production. Seems like Shaw called out Luck and Hogan as being excellent in that regard. My question/worry is that the offense steps back not because our new QB isn't effective throwing/running but in reading the defense and getting us into the optimal play.

Thoughts?

Important but I think it's overblown. If you execute a play well, it should gain yards. Some plays are going to work better against some defenses than others. But the best offense is going to make it so the defense can't win unless the offense screws up. That's what has guys like OF and Hulk so excited about this team, if you put Irwin, Schultz, Love, McCaffrey, and a FB/TE on the field you can give so many different looks that the defense will always be wrong. Here's an article talking about the Patriots with Gronk and Bennett, but the same idea could work with us. If the defense comes with nickel, then flex out one of the RBs and run it against a 6-man box. If they play base, throw to whoever gets matched up with an LB, because it's a mismatch whoever it is.

If the coaching staff are smart, that's how they'll build the offense. Make it so that, no matter what, the defense can't match your personnel. It might not work flawlessly against, say, Alabama, but it should shred the Pac-12 provided the QB and OL are even just competent. If they're good? Yikes.

[stupac2]

How important is the QB's ability to get the team into the right play in terms of production. Seems like Shaw called out Luck and Hogan as being excellent in that regard. My question/worry is that the offense steps back not because our new QB isn't effective throwing/running but in reading the defense and getting us into the optimal play.

Thoughts?

Important but I think it's overblown. If you execute a play well, it should gain yards. Some plays are going to work better against some defenses than others. But the best offense is going to make it so the defense can't win unless the offense screws up. That's what has guys like OF and Hulk so excited about this team, if you put Irwin, Schultz, Love, McCaffrey, and a FB/TE on the field you can give so many different looks that the defense will always be wrong. Here's an article talking about the Patriots with Gronk and Bennett, but the same idea could work with us. If the defense comes with nickel, then flex out one of the RBs and run it against a 6-man box. If they play base, throw to whoever gets matched up with an LB, because it's a mismatch whoever it is.

If the coaching staff are smart, that's how they'll build the offense. Make it so that, no matter what, the defense can't match your personnel. It might not work flawlessly against, say, Alabama, but it should shred the Pac-12 provided the QB and OL are even just competent. If they're good? Yikes.

[terry]

Welcome Jabbahop!


I think it's pretty important for a QB to get into the right play. A defense may come up to the line showing a look where they have too many bodies to block at the point where the play is supposed to go. You need a QB who recognizes that stuff and can make the adjustment.

[terry]

Welcome Jabbahop!


I think it's pretty important for a QB to get into the right play. A defense may come up to the line showing a look where they have too many bodies to block at the point where the play is supposed to go. You need a QB who recognizes that stuff and can make the adjustment.