FO on ESPN: College Stats Don't Lie

week's ESPN column looks at the Lewin Career Forecast and what it says about the top four quarterback prospects in this year's draft. (Don't freak out if the system doesn't seem fully explained; remember, ESPN articles have size constraints.)

View Full Article

Comments

55 comments, Last at 21 Apr 2008, 6:31pm

2 Re: FO on ESPN: College Stats Don't Lie

I have high regard for Lewin's forecast, but the statement that draft position "has virtually no bearing on his NFL success" is far from the case. Draft position is obviously no guarantee, and there are occasional late round gems, but the first few QBs taken tend to dramatically outperform subsequent picks.

3 Re: FO on ESPN: College Stats Don't Lie

The difference in number of starts between Henne and the other three is really very striking. This suggests he's the 'safest' of the QBs on this list. Not sure that's true (his pocket presence is a concern), but I do really like him as a prospect for a team that has run/pass balance and likes to throw downfield.

4 Re: FO on ESPN: College Stats Don't Lie

Re: 2

More than that, its actually, to the best of my knowledge, false. In Lewin's original article he limited his research to the first two rounds, and I recall him saying that the system didn't work for QBs drafted later than that.

5 Re: FO on ESPN: College Stats Don't Lie

"Wannstedt may not be the best coach in college football, but it is hard to believe he was that wrong."

Does no one remember the Bears of the 90s?

6 Re: FO on ESPN: College Stats Don't Lie

I suspect that as far as the FO readership is concerned, this is preaching to the choir.

I would be very interested to hear Lewin's thoughts in detail on Colt Brennan, whose slide I regard as even more ludicrous than Brohm's. Saying that, I don't think it's a given that Brohm will fall that far. He could go as early as #8 to the Ravens. Chicago also seems like a fair possibility. I'd certainly be staggered if he fell out of the first round altogether.

Meanwhile, the Lions, Packers, Cowboys, Eagles, Redskins, Falcons, Buccaneers, Cardinals, Seahawks, Ravens, Jets, Colts, Jaguars, Titans, Raiders and Chargers are all still in need of GMs for the FO Reader Participation Mock Draft. If you fancy making the picks for your favourite team in the first three rounds, email me at tomrichards8464 at gmail dot com.

7 Re: FO on ESPN: College Stats Don't Lie

“Wannstedt may not be the best coach in college football, but it is hard to believe he was that wrong.”

I could not disagree with this statement's second half any more strenuously. As someone who lived in Pittsburgh during his entire tenure, it is extremely easy to believe, nay KNOW, that he was that wrong.

The lack of starts for Flacco means we don't have a way of knowing how good he may actually be. However, in this case, the lack of starts is not evidence in and of itself that he won't be good, due to the fact that he was coached at Pitt by a complete and utter disaster of a head coach.

8 Re: FO on ESPN: College Stats Don't Lie

RE: 5

Amen, brother. Wannstedt is reviled by Bears fans. His biggest weakness was talent evaluation. He simply has no clue in that regard.

That Ol' Wannie took a long look at the kid from Delaware and decided to go another way is the most ringing endorsement I can imagine, short of the ghost of Bill Walsh saying, "I'd draft him."

9 Re: FO on ESPN: College Stats Don't Lie

David-
If you had somehow been drafted in the first two rounds, how good would your projection have been?

10 Re: FO on ESPN: College Stats Don't Lie

re: 5

Or the Dolphins from early in this decade.

"What can Trent Green do that Jay Fiedler can't?"

I'm going to throw up a little now.

11 Re: FO on ESPN: College Stats Don't Lie

This article, even in short form, confirms what most of us who read here regularly already know:
1. Brohm has a chance to make some team very happy between picks 25-40--he was being touted as perhaps the #1 overall even in Oct.
2. Flacco has a small sample size, and while he may be a good QB, it would be good for him to go somewhere where he could sit on the bench for a year or 2.
3. Ryan is a risk--he seems to be the guy that could wind up being a Eli Manning type--good NFL starter, not great. Or he could become Akili Smith.
4. I never understood why Henne wasn't mentioned as much as the other 3, till recently I heard him being suggested as equals with Flacco & Brohm as early 2nd rounders.
My hope for Henne is a team like Minnesota--great RB core, good O-line, good D--where he could sit for a year (so the coaches would see the TJ is NOT a good QB--and get another good WR) and then step in and lead them into the playoffs because he didn't need to be the team saviour, just someone who could make plays when it's his turn, like Rivers has done in SD.

12 Re: FO on ESPN: College Stats Don't Lie

#5 et al,

You don't have to remember the Bears. Introducing Wannie's finest hour, the 1-15 Dolphins.

13 Re: FO on ESPN: College Stats Don't Lie

I have high regard for Lewin’s forecast, but the statement that draft position “has virtually no bearing on his NFL success” is far from the case.

I think he meant, "within the first two rounds, draft position has virtually no bearing on NFL success". Which is basically true. A QB who is taken with the first overall pick is just as likely to be successful as another QB with the same games started and completion% who is taken in the middle of the second round.

14 Re: FO on ESPN: College Stats Don't Lie

If Wannstedt carried out deeds more heinous than trading a 1st and 3rd for Rick Mirer then I don't want to know what they were. Some things are better left unknown.

(although the name John Avery keeps popping up in my mind)

15 Re: FO on ESPN: College Stats Don't Lie

Re 4. To be fair, he does mention the first 2 round limit in this article too. But even within the first 2 rounds, the 1st QB taken tends to significantly outperform the 2nd, the 2nd outperforms the third, and so on.

But that doesn't take away from his main thesis, that college starts provide a lot more data and film, and are indicators that he won't be a bust.

16 Re: FO on ESPN: College Stats Don't Lie

I would be very interested to hear Lewin’s thoughts in detail on Colt Brennan, whose slide I regard as even more ludicrous than Brohm’s.

Lewin is not a scout. The scouts have determined that Brennan doesn't have the quality to go in the first two rounds, so he falls outside Lewin's projection system.

Any arguments about his projection therefore come down to arguing with the scouting. They're wrong every few years.

18 Re: FO on ESPN: College Stats Don't Lie

Hmmmph! I talked about exactly this trend in an article last year. Who the heck are these "Outsiders" that are stealing my ideas?

19 Re: FO on ESPN: College Stats Don't Lie

There's some wording in the article that seems contradictory, but maybe that's the fault of the limit.

In the early part of the article, Lewin states that these indicators "are far better than the scouts at determining how good a player will be", and that "where a quarterback is selected in the draft has virtually no bearing on his NFL success". And yet he does acknowledge that his projection system only applies to QB's drafted in the first two rounds, because that allows scouts to weed out anyone who is actually bad.

It's probably more accurate to say that scouts do a good job, usually, of ruling out bad QB's (excepting the occasional Jake Delhomme and Tom Brady), but college stats are more useful than scouts in differentiating the remaining candidates.

And that gross draft position--i.e. what day a QB is taken on--does correlate with success, but that exact draft position (say, +/-30 spots) has little effect, and then completion percentage and games started are more important.

As for the contention that "good players start and we can trust college coaches to know who their good players are"--I strongly disagree with this statement. Consider Matt Cassel and Carson Palmer. Word is that Cassel almost beat out Leinhart for the starting job. Had Leinhart had a little muscle twinge or something when they were competing, then Cassel would have started, and given USC's talent level, probably would have been reasonably successful.

There's also many cases when a college QB doesn't start because his skill set doesn't fit the system a coach wants to run. In college, far more than the NFL, niche QB's with special skills are used, and if you're a stand-up NFL-downfield type pocket passer playing for a coach who is in love with the spread option, you're probably not going to start over a really good scrambling QB, even if you would be more successful at the NFL level. In the pros, a good coach will adapt his scheme to the strengths of the personnell he has, since there's such a scarcity of NFL-level talent. In college, coaches are far more likely to stick to whatever system they think works best or they are most familiar with. Hence starting time is probably much less a function of how good a player is and more a function of how well he fits the coach's system.

20 Re: FO on ESPN: College Stats Don't Lie

Brohm did not play in a gimmicky offense at Louisville. It was pro-style. You can say he had a strong supporting cast on offense though.

BTW, Lewin is dead wrong on Flacco. Palko was an accurate WCO style passer, was a far better fit for their offense than Flacco. Since he was a folk here in the Pittsburgh area, Flacco would have never gotten a fair shot at the job.

Also, I wouldn't compare Ryan to Eli at all. He's the next Matt Hasselbeck. Ryan is actually a very accurate passer, he just thinks he's Brett Favre and forces the ball far too much.

6: Brennan was terrible at the Senior Bowl, and played the worse schedule imaginable. NFL scouts have done their homework there.

21 Re: FO on ESPN: College Stats Don't Lie

15: Can you really say with confidence that earlier picks outperform later picks for reasons other than playing time? Sure, a first-overall pick gets a lot more playing time because of the huge investment, but that doesn't automatically make him better than a second-rounder the coach felt comfortable discarding.

22 Re: FO on ESPN: College Stats Don't Lie

Michigan has done a very good job at creating NFL quarterbacks the past 20 years, and Henne should be no exception.

Grbac, Collins, Griese, & Brady were all NFL starters with varying levels of success. Throw in Henson and Navarre as NFL players (although neither were that good).

Henne will probably be the last one for a while though while Rich Rodriguez is there.

23 Re: FO on ESPN: College Stats Don't Lie

Brennan was terrible at the senior bowl, but I'm quite willing to believe that was largely due to Mike Martz. I love Martz as an NFL coach, but I wouldn't expect a college quarterback to perform well when suddenly confronted with his methods for a short period only. None of the QBs on that squad did well.

I know Brennan played a weak schedule, but it's not like he was on a powerhouse team surrounded by NFL level talent himself, and he absolutely tore it up for two straight seasons. I'm not saying he's a first round talent, but I absolutely think he's a first day talent, and I think his struggles in the Sugar Bowl are unduly influencing the way he's assessed. Brady or Manning would not have had a good game playing in the Hawaii offense against the Georgia defense. It's not helpful or reasonable to compare that performance to one by a QB at a big school playing against a quality opponent.

24 Re: FO on ESPN: College Stats Don't Lie

Re:15 But even within the first 2 rounds, the 1st QB taken tends to significantly outperform the 2nd, the 2nd outperforms the third, and so on
I don't think so.
Here's the QB draft order for years with multiple QBs in rounds 1 and 2 since 1996. I'd say in 2 of 9 years the 1st QB was clearly the best.
1997: Druckenmiller, Plummer
1998: Manning, Leaf, Batch
1999: Couch, McNabb, Smith, Culpepper, McNown, King
2001: Vick, Brees, Carter, Tuiasosopo
2002: Carr, Harrington, Ramsey
2003: Palmer, Leftwich, Boller, Grossman
2004: Manning, Rivers, Roethlisberger, Losman
2005: Smith, Rodgers, Campbell
2006: Young, Leinart, Cutler, Clemens, Jackson

25 Re: FO on ESPN: College Stats Don't Lie

15, 21: I think there's a cause-and-effect issue here. Did you ever think that maybe players are drafted earlier in the first two rounds because they're better and will likely have a more productive career?

26 Re: FO on ESPN: College Stats Don't Lie

Re 21-
Of QBs drafted from 1980-2000, the first QB taken tends to outperform the second (in yards per attempt, adjusted for interceptions).

QB Pick / Adjusted YPA
1 / 5.57
2 / 5.24
3 / 4.87

After the 3rd QB taken, it gets pretty random. By the way, those are pretty big differences (~0.5 YPA).

Also, the first QB taken has better career efficiency stats than the second QB 62% of the time.

The link below has more if you're interested.

27 Re: FO on ESPN: College Stats Don't Lie

I seem to remember Casey Weldon starting over Brad Johnson, Billy Joe Hobart starting over Mark Brunell, and Brian Griese starting over Tom Brady (As well as Drew Henson taking half the snaps afterwards). This is off the top of my head.

I'm not sure I agree with the rationalization that college coaches are going to play the better quarterback. Systems definitely come into play here, as well as the coaches preference for a kid that he personally recruited or against a kid whom he doesn't like personally.

28 Re: FO on ESPN: College Stats Don't Lie

Let me clarify a few points:
1. For quarterbacks drafted in the first two rounds 1997-2007 pick number (not what number quarterback they were, but their actual draft position) is not a statistically significant predictor of NFL success. This specific point is not debatable, although the more general one Brian makes might certainly be true.

2. My initial sentence with respect to Wannstadt was, "If Flacco turns out to be as good as the scouts say he is, and Wannstadt didn't notice then Wannstadt is a worse coach than I think he is, and I think he is a terrible coach." This got changed in the editing process but the final version vastly understates my distaste for Wannstadt and his coaching 'abilities'.

29 Re: FO on ESPN: College Stats Don't Lie

I know it was basically said by MKJ/19 but there's something almost annoying about describing games started as "far better than the scouts" when perhaps the biggest reason games started works is because of scouting.

30 Re: FO on ESPN: College Stats Don't Lie

Also this article doesn't most accurately represent my thoughts on all the relevant issues here. This was short piece for a general audience. For more extensive discussion of this research I would recommend my three previous FO articles on the topic, and my two very in depth PFP (06 and 07) articles.

31 Re: FO on ESPN: College Stats Don't Lie

not what number quarterback they were, but their actual draft position

I'd actually be surprised if a player's actual draft position was a reliable statistical indicator of success for ANY player taken in ANY round at ANY position, especially near the top of the draft, because it's affected by so many variables that have nothing to do with how good a player is. Specifically, how good the talent at other positions in the draft is, and what the needs of teams drafting in a set order are.

Consider a player that is indeed worth exactly the #12 spot in the draft. Say he plays CB, and he's the second best CB in the draft. That player could go much earlier if two of the top five teams are desparately in need of a CB and are set at most other positions. Or if there are many good players at other positions but only two good CB's in the whole draft. Or that player could go much later if there are many good CB's in the draft and the talent drop off form Mr. #12 to the third and fourth best CB's in the draft isn't huge. Or if none of the top 12 teams need a CB.

Basically, even if scouts were perfect and always knew how high a player should be ranked, the ranking you could infer from a player's draft position has got to have an error associated with it of at least +/- 10 spots.

Which is why rookie pay scales closely tied to draft position are stupid. But that's a discussion for another day.

32 Re: FO on ESPN: College Stats Don't Lie

Re: 19 (Cassel v Leinart)

I'm not entirely sold that Leinart actually is a better QB than Cassel. We haven't seen what Cassel can really do in the NFL, but we've seen enough to know that Leinart sure can't do a whole lot.

33 Re: FO on ESPN: College Stats Don't Lie

Pro Football Reference looked at whether draft position predicted QB success, and what they found (linked below) is that QBs who are picked first overall do substantially better than anyone else, while there isn't much of a difference between picks 2-40. Then around pick 40 there is another dropoff.

The interpretation of this is that there are 3 tiers of college QB prospects: guys who are clearly great prospects (like Manning) who get snapped up with the first pick, guys who are good who will be gone by the early to mid 2nd, and guys with potential who get taken mid 2nd or later who are less likely to succeed. NFL teams are pretty good at figuring out which tier a QB belongs in, but within each of the lower two tiers they aren't very good at figuring out who will succeed.

This could be roughly consistent with Lewin's system, if we see Lewin as providing a way to distinguish between the QBs within tier 2 (and also providing the occasional warning against a QB who is mistakenly seen as tier 1 - though generally Lewin & NFL teams will agree on who deserves to be in tier 1). There's also some disagreement about the boundary between tiers 2 & 3 (pick 40 vs. pick 64), but some of that may have to do with the different time frames used (the PFR analysis goes back to 1971).

34 Re: FO on ESPN: College Stats Don't Lie

I said during the season that the time to discuss this is when this article gets published, but I'm strongly inclined to believe that the accuracy of the Lewin Career Forecast depends precisely on people not noticing and paying attention to precisely those variables the Lewin Career Forecast uses to perform its analysis. Basically, if teams start only drafting in the first 2 rounds college QBs who completed a lot of passes (because completing a pass is really, really important) and played a lot of games (helps ensure the scouts are correct in their observations), then these two factors become less important.

If that looks like a familiar story, that's because shares some elements with the post-Moneyball story of OPS in baseball. The A's were able to get high OPS players for value because they were mispriced in the market relative to their value. Same with NFL QBs-to use a prominent recent example, Billick massively overvalued Kyle Boller based on his arm strength.

Take a player in the current draft-Chad Henne. For most of his career, he's shown occasional tendencies to badly miss receivers and, more damning, revert to poor footwork and poor throws under pressure. My impression was that he'd fallen down in the rankings, but has made a post-bowl rise back to where he's considered a solid 2nd round pick, precisely because he was a starter at Michigan for so long and did reasonably well throwing the ball. So, he ends up as a second round pick who's projected to be an above-average QB by the Lewin Forecast. And as he and guys like him end up in the second round and not turning into great QBs (Kevin Kolb may end up a good example of this), the Lewin Forecast, already inaccurate for the 3rd-7th rounds, stops being accurate for the 2nd round, too. There will probably always be exceptions, and as the factors for the Lewin Forecast rise in importance, the exceptions may increase percentage-wise, precisely because the decision to ignore the factors becomes harder to justify.

David, please don't take this as an attack on you personally-I don't mean it that way at all. I just think the Lewin Career Forecast is a wasting asset.

35 Re: FO on ESPN: College Stats Don't Lie

Re 32: Really? Leinart had the best rookie season of the quarterbacks taken in his class. Then his coach got fired, he needed to learn his second offense in two years and he got hurt early in the season. Isn't it a tad early to go writing him off?

I for one would be surprised if Leinart didn't have a better career than any quarterback drafted this year or last year, or if he turned out to be better than Vince Young or Jay Cutler (though Cutler is looking good). But that's me.

36 Re: FO on ESPN: College Stats Don't Lie

12-Introducing Wannie’s finest hour, the 1-15 Dolphins. The guy was years and two head coaches removed from that team. How can anyone blame the 1-15 Dolphins on him?

37 Re: FO on ESPN: College Stats Don't Lie

Am I along in thinking Chad Henne has little chance for NFL success? Well, I guess not, as someone up above said the same thing, but I really don't think he has anything to offer beyond arm strength. And it's not like his arm strength is off the charts for an NFL QB. He has no mobility, no pocket sense (he makes Drew Bledsoe look like Steve Young), doesn't seem to be very accurate, and who knows where or why he's throwing the ball half the time.

38 Re: FO on ESPN: College Stats Don't Lie

I don't think Brennan's stock really fell, because nobody ever saw him as a first day pick. Brohm was just supplanted as the number one overall QB prospect; he's still a well-regarded player who will probably go in the first round.

I don't think Brennan is an NFL starter - he's a really raw, developmental prospect with bigger gaps in his game than those of Matt Ryan or Brian Brohm.

39 Re: FO on ESPN: College Stats Don't Lie

#34 I largely agree with your statement. I'm not quite sure what stage we are at in the process right now. I can also say that completion percentage is predictive of NFL success regardless of round so it seems to be important, not just undervalued.

40 Re: FO on ESPN: College Stats Don't Lie

Personally, I think pretty much every QB would be better off sitting their first year in the NFL. Has their ever been a more NFL-ready QB than NFL Royalty, Peyton Manning (who would have probably been the #1 pick the year before if he had come out as a JR with a Bachelor's degree)? He took a beating that first year and was lucky to not get injured. Anyone less experienced (lifelong coaching by some of the best) might have learned some bad habits... take a look at someone like David Carr or Ryan Leaf.

I think coaches can be counted on to play the best player for their system. This may not be the best player in another system. Michael Vick, Vince Young, or Alex Smith will never be the best pocket QB.

I think this year has some decent potential NFL QB's. I am not sold that any of them will be among the elite on their first contract. That basically means that any of them who are drafted early in the first round may be a financial drag on their team and may help prevent their team (and maybe their personal development) from competing at the top level. Yes, a QB touches the ball and has more impact than any other player. However, early first round picks are insanely overpriced for how little they have proven at the NFL level. On that note, I hope Long takes a below market contract and realizes that this will help his team (and maybe the NFL draft as a whole).

41 Re: FO on ESPN: College Stats Don't Lie

Brennan absolutely was seen as a potential first day pick in many quarters, including by David Lewin, per his PFP 2007 article, and by the guys at www.scout.com as recently as December. I agree that he is not ready to be an immediate NFL starter, but no quarterback chosen in the second round has been a successful starter in his first two seasons for more than a decade. Brennan is smart, very accurate and reasonably mobile. He is ill-suited to some NFL offenses (including Martz's) because of his lack of arm strength, but in a West Coast offense focussed on the short pass and YAC, I think he could become a very effective player.

42 Re: FO on ESPN: College Stats Don't Lie

Peyton Manning (who would have probably been the #1 pick the year before if he had come out as a JR with a Bachelor’s degree)? He took a beating that first year and was lucky to not get injured

I don't know where people keep getting this idea, but it's not even remotely true. Peyton Manning did not take a beating in his rookie year. He was sacked 22 times. You know how many times he was sacked last year? 21. Ooooh, big difference! The coaches really should've kept him on the bench! He turned into the best QB in the NFL, and he's never missed a game.

Are rookie QBs going to be great players? No, but that doesn't mean they're going to get beat up. Sometimes, they're just going to throw a lot of picks. That hurts the team that year, but there's no permanent harm done, since a team starting a rookie QB isn't going anywhere special anyway, in most cases. And it's not like throwing lots of picks one year prevents a QB from having a low interception% in later years.

So stop acting like starting Peyton Manning as a rookie was such a terrible thing. The only thing that went wrong was that Manning threw 28 picks while starting for a 3-13 team that wasn't going to be any good regardless of who was starting. Not exactly the end of the world.

43 Re: FO on ESPN: College Stats Don't Lie

well, I didn't mean IMMEDIATE nfl starter - i just meant I don't see him as a starter ever. qb is a developmental position and it's unreasonable to expect immediate high-level production from even the most elite prospects. it's just that there are a bunch of guys in the draft who simply have better stuff than brennan - better technique, better size, better accuracy, better strength. I think studying his college production is a little bit of a dead-end because of how unusual his situation was - the system, yes, and also the fact that he excelled entirely against some really, really awful competition. That Hawaii team wouldn't have won 8 games in the pac 10.

44 Re: FO on ESPN: College Stats Don't Lie

"So stop acting like starting Peyton Manning as a rookie was such a terrible thing."

I agree with you about Peyton, but at the same time I don't think you can extrapolate any broader points about rookie QBs from his career path. The speed of his progression through the development process was phenominal and pretty much necessitated that he start that first year. But if you throw a rookie QB to the wolves too early in their development process, it's akin to taking a new Ferrari and careening down some country backroads in it.

45 Re: FO on ESPN: College Stats Don't Lie

#36
Go look at the Dolphins roster and tell me how many players are left that Wanny drafted. I am almost sure there are only 2 left (Carey, Bell). Where Miami is now has as much to do with Wany as it does Saban or Cameron. Probably more.

46 Re: FO on ESPN: College Stats Don't Lie

Martz hurt Brennan at the senior Bowl.
But he hurt Woodson still more.
He was viewed as a first-round prospect and now, he isn't even mentionned in the secound round...

Can someone explain me what happened ?
Thx.

47 Re: FO on ESPN: College Stats Don't Lie

Beyond the aforementioned questions about excelling in a system that gave us Timmy Chang leading the NCAA in several categories and the lack of competition, surely Brennan's unfortunate escapades at Boulder lower his draft stock as well. Fourth round seems right.

50 Re: FO on ESPN: College Stats Don't Lie

but it’s not like he was on a powerhouse team surrounded by NFL level talent himself

Brennan wasn't surrounded by NFL-level offensive linemen. He was certainly surrounded by NFL-level wide receivers.

51 Re: FO on ESPN: College Stats Don't Lie

But if you throw a rookie QB to the wolves too early in their development process, it’s akin to taking a new Ferrari and careening down some country backroads in it.

Sure, but I think NFL teams are pretty good at knowing when a QB is prepared enough to play without suffering permanent harm. Manning isn't the only QB that started as a rookie that ended up being elite. McNabb and Roethlisberger both started during their rookie years. Eli Manning and Jay Cutler have both done well, and they started as rookies.

And if you go back into past years, there are plenty more. Here's a list of HOF QBs that started 5+ games as rookies: Troy Aikman, Dan Marino, John Elway, Dan Fouts, Terry Bradshaw, Bob Griese, Joe Namath, Fran Tarkenton, Sonny Jurgensen, Y.A. Tittle, Sid Luckman, Sammy Baugh. That's 12. With Peyton Manning, 13. If starting a QB in his rookie year is so bad for him, how did all these guys have so much success?

Most of the examples people bring up of QBs that were supposedly hurt by starting too early were QBs that had poor projections anyway. So pointing to, say, David Carr or Ryan Leaf, ignores that they both had poor projections coming out of college, and were unlikely to be very good regardless of how long they sat on the bench. And lots of QBs that had bad projections that did sit for a while ended up being bad as well, so it seems unlikely that starting early hurts QBs.

I guess the way to test this, if it hasn't already been done, is to compare the performance of QBs who had X starts in college, Y completion%, and started as rookies, to the performance of QBs who had X starts, Y completion%, and sat on the bench for a year or two. If the second group isn't outperforming the first group, then that'd be pretty good evidence that starting early doesn't usually hurt QBs.

Quick question–how is QB success defined in Lewin’s criteria?

DPAR/Game

52 Re: FO on ESPN: College Stats Don't Lie

"Re 32: Really? Leinart had the best rookie season of the quarterbacks taken in his class. "

Yeah, he did. He also has two monster receivers.

I think this past year has shown what one really good receiver can do for a qb.

53 Re: FO on ESPN: College Stats Don't Lie

Absolutely. Part of the reason why I expect Leinart to be a big success is because he's in a good situation- or at least he will be once the offensive line is straightened out. Situation matters. But I'm baffled that people would write him off as a bust based on what was a pretty successful rookie season and an injury-shortened sophomore one. (And I should point out that Leinart is actually on offensive coordinator number three, not two, as Dennis Green made a switch right after the Bears loss...and, not coincidentally, that's when Leinart began struggling.)

54 Re: FO on ESPN: College Stats Don't Lie

50: To piggy back on your comment, UGA has some of the best talent in the country year in and year out, and their front 4 dominated Hawai'i's line. It wasn't until the (no pun intended) dogs were called off that Hawai'i started to move the ball effectively.

Brennan performed about as well as any other QB could have in that situation.

55 Re: FO on ESPN: College Stats Don't Lie

Yeah, I make no prediction regarding Brennan in the pros, but that game against Georgia is useless as a predictive tool, given that Hawaii's offensive line just got flat-out whipped on nearly every single play.