SITE NEWS: We are moving all of our site and company news into a single blog for Sports-Reference.com. We'll tag all PFR content, so you can quickly and easily find the content you want.
Also, our existing PFR blog rss feed will be redirected to the new site's feed.
Pro-Football-Reference.com ยป Sports Reference
For more from Chase and Jason, check out their work at Football Perspective and The Big Lead.
ShareThisThe History of the Black QB: Part III
In Part I of this series I chronicled the emergence of the black quarterback, from Fritz Pollard to JaMarcus Russell. In Part II, I looked at some of the greatest seasons by a black QB in NFL history, with Daunte Culpepper's 2004 season topping the list.
There have been twelve black quarterbacks to start over 60 games in the NFL in their career, and Charlie Batch (49) and Byron Leftwich (44) might join them one day. David Garrard, Vince Young, Jason Campbell and JaMarcus Russell are young QBs that should stick around for awhile, too. Here's a list of the twelve:
passyd rushyd gs Warren Moon 49325 1736 203 Steve McNair 31304 3590 147 Randall Cunningham 29979 4928 135 Donovan McNabb 25404 2962 98 Daunte Culpepper 22422 2540 84 Jeff Blake 21711 2027 100 Aaron Brooks 20261 1534 90 Doug Williams 16998 884 81 Rodney Peete 16338 1139 87 Tony Banks 15315 881 78 Kordell Stewart 14746 2851 87 Michael Vick 11505 3859 67
Moon started his career in the Canadian Football League (and didn't enter the NFL until age 28), but the other 11 QBs all started in the NFL in their early or mid-20s. Listed alphabetically, here's how many rushing yards each QB had in his first season with at least eight games started:
age team year rshyd gs att
Tony Banks 23 ram 1996 212 13 368
Jeff Blake 24 cin 1994 204 9 306
Aaron Brooks 25 nor 2001 358 16 558
Daunte Culpepper 23 min 2000 470 16 474
Randall Cunningham 24 phi 1987 505 12 406
Donovan McNabb 24 phi 2000 629 16 569
Steve McNair 24 oti 1997 674 16 415
Rodney Peete 23 det 1989 148 8 195
Kordell Stewart 25 pit 1997 476 16 440
Michael Vick 22 atl 2002 777 15 421
Doug Williams 23 tam 1978 23 10 194
Avg 23.6 407 13.4 395And then in the QBs' 4th year starting:
age team year rshyd gs att
Tony Banks 26 rav 1999 93 10 320
Jeff Blake 27 cin 1997 234 11 317
Aaron Brooks 28 nor 2004 173 16 542
Daunte Culpepper 26 min 2003 422 14 454
Randall Cunningham 27 phi 1990 942 16 465
Donovan McNabb 27 phi 2003 355 16 478
Steve McNair 27 oti 2000 403 15 396
Rodney Peete 26 det 1992 83 10 213
Kordell Stewart 28 pit 2000 436 11 289
Michael Vick 26 atl 2006 1039 16 388
Doug Williams 26 tam 1981 209 16 471
Avg 26.7 399 13.7 394Those numbers are pretty close. Vick's huge year boosts things for sure, but it appears as though rushing doesn't decline too much just yet.
But after six years of starting (a year of starting is considered a season with at least eight games started), the black QBs stop running. This comes at around age 30, which has long been touted as the year when players start to lose their athleticism. (Curiously, the top 18 rushing seasons by all QBs age 30 or older from 1970-2007 were by white QBs; only two of the top 17 rushing seasons by all QBs of any age were by white QBs.) Here's a look at the rushing numbers for all QBs in their sixth year of starting or later:
yrstart age team year rshyd gs att 7 Donovan McNabb 30 phi 2006 212 10 316 7 Steve McNair 30 oti 2003 138 14 400 7 Randall Cunningham 35 min 1998 132 14 425 7 Jeff Blake 32 rav 2002 106 10 295 7 Rodney Peete 36 car 2002 14 14 381 8 Jeff Blake 33 crd 2003 177 13 367 8 Steve McNair 31 oti 2004 128 8 215 9 Steve McNair 32 oti 2005 139 14 476 10 Steve McNair 33 rav 2006 119 16 468 32.4 129 12.6 371While we should be a bit concerned about a Simpson's paradox problem -- especially since Vick isn't on this list -- note how the games started and attempts dropped slightly, while rushing yardage crashed. The QBs averaged about 30 rushing yards per game in their first and fourth years starting, but just 10 yards per game by the end of their careers. Rushing yards and pass attempts were practically even in their early years, but by the last seasons their careers the QBs had nearly three times as many pass attempts as rushing yards. McNabb, McNair and Cunningham transformed themselves from running to passing quarterbacks. Kordell Stewart never stopped running, and was entirely ineffective once he hit thirty-one years of age. Who knows what would have/will happen to Vick, although I think a pretty honest debate could take place as to whether he was a better passing QB than Stewart. Kordell actually improved a bit as a passer, after starting off pretty poorly. Vick started off as a great QB and a pretty good passing QB, but actually regressed as a passer each season; his adjusted net yards per attempt dropped each of his four years as a starter.
This, in the end, is a thinly veiled "what might happen to Vince Young" post. The other young black QBs -- Campbell, Garrard and Russell -- are more passers than runners. Russell's not a runner, Garrard can run but looks to pass first, and Campbell has a strong arm and isn't very fast. But the young Vince Young resembles the young McNair/McNabb/Cunningham pretty well. On the other hand, he also fits the young Vick/Stewart model.
It's rarely wise to view the world in black and white, especially when it comes to race. But there aren't many white QBs in the modern era that have resembled what Young's done to start his career. Steve Young fits, but the rest of his career was pretty peculiar: after posting somewhat similar stats to Young through two seasons, he was on the bench for the next four, and then played with the greatest WR ever for the next eight.
Last week we looked the best value added seasons by a black QB. Value added is a very good way to encompass a QB's attempts, yards, rushing yards, TDs, INTs and league average into one stat. Let's use that again, but ignore the rushing compoment for a soon to be obvious reason.
So the formula will be the difference between QB adjusted net yards per attempt (with attempts defined as pass attempts plus sacks) and the league average adjusted net yards per attempt, multiplied by total attempts.
age team year lgavg nay/a value rshyd gs Ran. Cunningham 24 phi 1987 4.71 4.56 - 70 505 12 Ran. Cunningham 25 phi 1988 4.69 4.68 - 7 624 16 Ran. Cunningham 26 phi 1989 4.88 4.49 -225 621 16Donovan McNabb 24 phi 2000 4.85 4.44 -247 629 16 Donovan McNabb 25 phi 2001 4.82 5.02 103 482 16 Donovan McNabb 26 phi 2002 4.97 5.20 88 460 10Steve McNair 24 oti 1997 4.79 4.55 -108 674 16 Steve McNair 25 oti 1998 4.91 5.24 172 559 16 Steve McNair 26 oti 1999 4.81 5.37 197 337 11Kordell Stewart 25 pit 1997 4.79 5.03 108 476 16 Kordell Stewart 26 pit 1998 4.91 3.36 -764 406 16 Kordell Stewart 27 pit 1999 4.81 3.18 -484 258 12Michael Vick 22 atl 2002 4.97 5.57 272 777 15 Michael Vick 24 atl 2004 5.23 4.49 -271 902 15 Michael Vick 25 atl 2005 4.98 4.23 -314 597 15Vince Young 23 oti 2006 5.02 4.20 -312 552 13 Vince Young 24 oti 2007 5.11 4.21 -365 395 15 Vince Young 25 oti 2008 ???? ???? ??? ??? ??McNabb and McNair had similar first two seasons: both were below average passers in year one, above average passers in year two, and improved their passing in year three (but both missed time due to injury). Both also had strong rushing numbers each season. Each went on to have very good careers, and won a conference championship game.
Stewart and Vick regressed as passers each season, although Stewart would revive his career in 2001. Both were well below league average passers in their second seasons. Cunningham was pretty consistent all three seasons, although the NFL average improved by his third year. He actually finished with a below average net adjusted yards per attempt ratio in all but one of his seasons as an Eagle. Here's something interesting about Cunningham: in six of his seven seasons as starting QB of the Eagles, he led the NFL in sack yards lost. I think most would consider Cunningham a superior QB to McNair. While they have almost identical QB Ratings, yards per attempt and adjusted yards per attempt, Cunningham gets destroyed when you factor sacks into the equation. He's got four of the worst ten seasons in sack yards lost, including the worst ever. McNair is 59th all time in sack yards lost, and 23rd all time in attempts -- he was difficult to sack. As a result, McNair averaged 0.7 more net yards per pass than Cunningham, and that's why Cunningham was almost always under the league average.
Young has had some awful WRs. I understand that completely; they might be the worst in the league. But most of the other QBs on this list have had similar complaints at some point during their career. He also started right away, unlike McNair or McNabb, and age plays a huge part in QB success. But he wasn't playing his college games at Alcorn State, either, so he was probably a bit more NFL-ready than McNair was. His well-below league average numbers last season have to be worrisome.
Two bad comparisons is far from the death knell for Vince Young. I've been a big fan of him since the National Championship Game, and he was my favorite of the Big Three Qbs coming out of college. I think he has the ability to improve his passing game, and reinvent himself as a passer. Giving him some good WRs could make a huge impact on his passing numbers.
For me, it's easy to get excited about an athletic QB who posts a winning record at age 23 and age 24, especially on a team without much offensive talent. Studies like this help me to reign in my enthusiasm quite a bit.
One other, pro-Young note. Vick and Stewart are 6'0 and 6'1, respectively. McNair and McNabb are both 6'2, Cunningham is 6'4, and Young is 6'5. Lots of people think height matters (myself included), especially if you've got a funky delivery.
This entry was posted on Friday, February 15th, 2008 at 10:51 am and is filed under History, Statgeekery. You can follow any responses to this entry through the RSS 2.0 feed. Both comments and pings are currently closed.
David Garrard is 30, which is normally not considered "young". However, he doesn't have the wear and tear of the average QB that has been starting for 6 years, so perhaps he's "football young?"
Garrard's status as a 5-year backup is unusual, so similarity scores and comparisons to peers are probably not totally accurate. Assuming his 2007 performance isn't a fluke, are there any legitamately good QBs that sat on the bench for so long?
Here, you are comparing him to the league average, which includes all the lead-footed passers. However, the NAYPA will overstate Cunningham's sack rate, and thus understate his passing value. In most cases, a rush attempt is a sack avoided, every bit as much as a pass incompletion or completion. Sure, there are going to be a few attempts that are by design, but for the most part, these attempts are occurring on original passing plays. But unlike the quarterback who throws it away, Cunningham does not get credit for getting away from a sack. His sack rate would still be high, but would be reduced by about 1.5% if you calculated NAYPA as [sacks/(sacks + pass attempts + rush attempts)].
Quarterback A throws 24 passes, runs the ball 8 times, and has 2 sacks. Quarterback B throws 32 passes, no rushes, and has 2 sacks. QB A has a sack rate of 8.3%, while QB B has a sack rate of 6.3%, but I would say they are roughly equal in their ability to avoid sacks.
I know you were touching on this with your previous post. I'm not sure if I can suggest an alternative that fairly captures it for a formula like AYNPA, because you would have to quantify the value of the rushing yards vs the passing yards to also include rush attempts in the denominator.
My gut reaction is I don't have as much a problem with simply adding in both rush yards and rush attempts to the NAYPA formula, without any adjustments such as rushes over 3.0 YPC, etc. Sure, there's some noise, but I wouldn't be too concerned over the kneel downs too much.
Bowman,
Garrard just turned 30 yesterday, so he was 29 last season. Warren Moon, Trent Green, Kurt Warner, Jeff Garcia, Brad Johnson, Jeff Hostetler, and Bill Kenney all come to mind as guys who got their first real opportunities to start between ages 27-29, and had pretty good careers.
Of course, when the age increases, you also get a lot more guys like Steve Bono, Jay Fiedler, Hugh Millen or Tommy Maddox, than you would if he was age 24.
I would guess that the interception portion of his 2007 season was a fluke that will be difficult to repeat. This does not mean that he was not responsible for the low int's, just that it is the least consistent "QB skill", because (1) it is a relatively rare event per attempt, (2) things such tipped passes at the line falling into a defender's hands or not, a receiver juggling a pass which goes in to a defender's hands or not, or a pass going into a defender's hands, and whether he holds it or not, are out of the qb's control, and (3) game context, will the Jags be trailing in the 4th more in 2008? are all factors.
However, his YPA and TD% are reasons to think he is not an overall "fluke".
There's no doubt Cunningham had an awful line when he was in Philly, but I once heard someone say that the reason he took so many sacks was that, because of his scrambling, the offensive linemen didn't know where to block for him.
JKL, I agree that your quarterback A/quarterback B comparison is one reason why Randall Cunningham and other scramblers might lose points for their seemingly high sack totals. The other question, though, is what do those quarterbacks do with those pass attempts?
If we assume they are the same in their 24 shared attempts and two sacks, we then compare where they're different: A's 8 rushes to B's 8 passes. 8 passes are very likely to gain more yardage than 8 rushes. That's why QB B gets rated higher in a pure "yards per attempt"-based model. If QB A was less of a runner and more of a smart/strong-armed/quick-release/cool-headed (take your pick) QB who could find the open man and pick up 15 yards rather than scramble for 6, he'd be rated higher.
I started a poll over on the Footballguys message board. Here's something I posted there for those who think Cunningham's sack #s are just the product of a bad offensive line.
That's far from clear and convincing evidence that the Philadelphia OL under Cunningham was at least average. But I think it might shift the burden of proof a bit.
Just so I'm clear, I'm not arguing that Cunningham was good at avoiding sacks early in his career, or that it was a product of his line. I think a QB is at least as responsible for sack rate as the line. I'm only saying that the sack rate as calculated will overstate Cunningham's sack rate, because it doesn't consider the sacks he avoided by running. His rate is still on the well above average even if we added all rushes into the denominator.
Jason W, I would agree with you, to an extent. I would want to know how many first downs QB A picked up on those 8 throws relative to how many QB B picked up on the runs. If A gained more yards, but B picked up a higher % of first downs, it may be closer to a wash.
I am pessimistic about Vince Young. He ran a 1 read offense in college. Look at your 1 option, if covered, scramble. Norm Chow had him in a 1 read offense in his first year starting to ease his transition to the NFL. This season they asked him to move beyond 1 read, and it produced awful results. Yes, he does have bad receivers, but he can't read defenses well, forces balls into coverage, and has passes batted down often for a guy who is 6 foot 5.