Some of you will, of course find reason to to argue with RC's great experiment. Others do believe in the predictive potential of statistics.
Below are exerpts from the fresh-off-the-presses analysis of our QB quandary, inspired by our very own (always congenial /lol) and plentiful debates. Go here for the full story ... Clausen v Pike Compare
I, too, am wondering just how RC Fischer managed to develop this elaborate formula. But I do have a theory. I think he simply took the six undeniably elite QBs of the modern era, then looked at their final years in college. The pre-supposition must be that these six wind up in the top 70 - 80 % of QBs based on his formula. In order to do that, he (actually it's a group of people) had to put the exactly correct weight on all measurables and stats, throw out pushovers, etc. I can imagine it's a complicated algorithm. Now, he's in the business of proving it works over time. To divulge the full algorithm would be to damage a huge potential money machine.
I don't think RC is trying to fool anybody. Having corresponded with him several times now, I'm convinced that he's a mathematician who happens to love football; and what started off as a pet project is now showing great potential as a predictor of success with a high accuracy rate. I would imagine he's looking for a big payday in 3-4 years, hoping his projections pan out. Imagine how much teams, agents, journalists will pay for the information he could provide.
So, it's a work in progress, that happens to work such that it awards high scores to the six best QBs. Ryan Leak also had a high score, because, well, some things just can't be quantified. But it is information that is pertinent to our team, so I guess I'd be remiss if I don't at least put it out here. Whee doggie, we could have us another QB thread, or it could be largely ignored. It's all wonderlic'n good (or bad).
UPDATE>: In response to questions/comments below or sent to me directly, I requested further info on which games were used in the Clausen analysis. RC Fischer's response is below, however, to summarize, not all games against winning teams are created equal. The formula they use is far more complex than what it may appear to be. And basically, asking exactly how it works is like asking KFC about their eleven herbs and spices before they opened their first store.
Update on games looked at = I have 2 parts within our overall system where we rule games in and out.In one section on Clausen our system throws out Nevada and Navy for bigger difference in "class" (to say it simply).*Occasionally even a 6-7 lost a bowl game team could make into consideration if it was a more power conference team (Clausen doesn't have any of those for us) matchup that would be considered key/strong. Or a top rated defense team that was 5-7, 6-7 from a power conference...that's rare, All this is just one section of our formula looking at the QBs. We are trying to look at key games, and exclude stat compilations on just rolling up weaker teams.In another section within the same overall calc all stats against winning record teams are in with a set weighted amount factored to an overall. So (simple example) a game against USCs stats carry more weight than against Navy, but both opponents have winning records...but we don't look at them the same. If that makes it clear. These are the "adj" numbers you see in the grid of info on Clausen. Not replicable without the exact formula.Again, I won't go until full detail or publish an exact formula because I don't want to give away something (if it works fairly predictably) for free. I hope you understand on that. I would lose my edge/usefulness to my clients. If it doesn't work at all to predict anything and it's another 50/50 nice QB theory, then ultimately I will just lose my clients period.Appreciate it. Let me know any follow ups from that.RC
This past week, our work on the 2011 College QBs Draft class was debated on the Carolina Panthers fan website -- "Cat Scratch Reader". Looking at all the fan comments sent in, I was struck by the struggle of many Panthers fans on their current QB dilemma. There was/is heavy debate on -- "...is Clausen the future?" or "...is Pike really any good" or "should the Panthers find a QB in the 2011 Draft"? I thought there would be more folks giving up on both Tony Pike and Jimmy Clausen, especially on Tony Pike. However, I do see there is still some amount of fan support and/or hope shown for one or the other Panthers current QBs.
QB is virtually everything in the NFL. An elite QB can change the destiny of a franchise for years to come; the lack of an elite QB can doom a franchise into a seemingly never-ending purgatory of perpetual "rebuilding plans" and constant QB debates. The Green Packers fans, don't really care about this year's incoming QB class. Nor do the Atlanta Falcons fans care, nor the San Diego Chargers, etc. In Buffalo, Minnesota, Arizona, and Carolina (to name a few), the possible trade-worthy QBs (Kolb, Flynn, etc) and especially the incoming 2011 College QB class is almost a 24/7 non-stop debate. The value of having "your" QB in place and not having to worry about it is "nirvana".
Since there was such considerable debate on the CSR site about whether or not the Panthers should use the #1 pick on a QB or continue with Clausen and/or Pike; I thought I would add to the debate by expanding our system analysis on the two Panthers QBs. *Thanks to the fans of the "Cat Scratch Reader" and to Rick Bates for featuring and debating on our college QB hypothesis work. Hope you guys enjoy debating this one!
See, our great debates actually bring attention to the site.... but Lordy Mercy, they do take on a life of their own sometimes.
The past decade of Super Bowl matchups and their Wonderlic scores:
2010-11 = Rodgers (35), Roethlisberger (25)
2009-10 = Brees (28), P.Manning (28)
2008-09 = Roethlisberger (25), Warner (?)
2007-08 = E.Manning (39), Brady (33)
2006-07 = P.Manning (28), Grossman (29)
2005-06 = Roethlisberger (25), Hasselbeck (29)
2004-05 = Brady (33), McNabb (14)
2003-04 = Brady (33), Delhomme (?)
2002-03 = Brad Johnson (?), Gannon (27)
2001-02 = Brady (33), Warner (?)
I must admit, it's hard to argue with those results. Take away Big Ben, and there's no known winning QB under a 28.
Divisional Winners, Playoff Winners in the last 2 Seasons
Maybe just looking at Super Bowl participants evokes a "well, the elite teams have an elite surrounding cast" response. OK, let's quickly look at the past 2 seasons on "general" NFL success...and QB's with a 25+ Wonderlic.
2010 = 92% of the 13 teams with a winning record in the NFL were led by a QB with a 25+ Wonderlic score QB (Vick exception, Cassel unknown, so 11 of 12 "known" teams hit it)
2010 = 100% of the NFL teams that won at least one playoff game were led by a QB with a 25+ Wonderlic score QB. No QB with a Wonderlic score under 25 won a playoff game.
2009 = 86% of the 15 NFL teams with a winning record were led by a QB with a 25+ Wonderlic score (McNabb, Favre exceptions...Warner unknown), so 12 of 14 teams hit it)
2009 = 83% of NFL teams who won a playoff game were led by a QB with a 25+ Wonderlic score (Favre breaks rule).
2009+2010 = 88% of NFL teams with a winning record in the past 2 seasons had a 25+ Wonderlic QB
- QBs drafted in the last decade:
2009+2010 = 92% of teams that won at least one playoff game were led by a 25+ Wonderlic QB
2009+2010 = 96% of the NFL teams have had a winning record in the past 2 seasons were led by a QB with a 25+ Wonderlic score on a test taken in the past 10 years (2010 Vick the only exception)
- All this 80%+, 90%+ trending for QBs with a 25+ Wonderlic score cannot make anyone feel great about Jimmy Clausen (or any QB) with a Wonderlic score of 23.
2009+2010 = 0% of the NFL teams have won at least one playoff game in the past 2 seasons IF they had a QB drafted/took the Wonderlic test in the last 10 years and that QB scored a 24 or less on the test.
Oops, Jimmy, say it ain't so!
One last small red-flag on Clausen
Clausen is also slightly dinged in our system with having small hand measurement. Hand measurements are controversial and unknown on many QB the further you go back in time. Again, it is not an absolute -- but small handed (9.0' or lower) QBs have not been as successful historically speaking. Tony Pike has a larger than average hand measurement. Big hands don't equal success, but small ones can be trouble. I know you can find a QB who bucks the trend, I know -- but there are many, many more QB failures or just mediocre for every 1 exception. Again, do you want to take the 20% odds, I'll take the 80%. The hand-size thing is just a slight system ding because of the probabilities/history of the past.
As well, standing just under 6'3 is not a great attribute either. Clausen is 6'2.6" and Tony Pike is a more "model" QB at 6'5.8". Below 6'2 is the real data cutoff line (Drew Brees a major exception to the rule) where the trend swings hard against future QB success (which makes Greg McElroy at 6'1.8" a potential problem). All things being equal, you would rather have a 6'5 QB with large hands to the hold/protect the ball, than a 6'2 and half QB with "tiny" hands. Not an absolute, but a definite trend in your favor with a bigger QB -- all else being equal.
The NFL QB Jimmy Clausen most compares to statistically coming out in the 2010 draft...
Let me get "pro-Clausen" fans excited for a minute, then pull the football out a la Lucy/Charlie Brown....
When I take the QBs we have in our system and start filtering them down to likely matches with Jimmy Clausen, as I go through all the various Passing Performance Metrics (there are 10+ categories with various weights) -- still left standing among 8 potential matches are names like Peyton Manning and Carson Palmer, along with a collection of mediocre/low/no performers. As I mentioned earlier, our Passing Metrics show pretty well for Clausen and Pike. Then, I filter to the QBs who are closer to Clausen's physical size and the list drops to 3 other likely matches (noted below). The closest matches are a cluster of good Completion Percentage QBs with very high ratings (almost off the charts) for Pass Attempts per INT, who are under 6'3.
Sadly, (here comes Lucy pulling the football away...) for Clausen fans the most likely match is John David Booty. Similar smaller size for both, "safe" passers with their accuracy...and just to pile on -- both under 24 on the Wonderlic. Graham Harrell scored under 24 as well.
I'm starting to lose confidence in Sir Speedo.
QB Yr College H W adj Comp Pct Adj Yds per Comp adj Pass per TD adj Pass Per INT Yds per game 35 Att TDs per game 35 Att INTs per game 35 Att Clausen, Jimmy 2009 Notre Dame 74.6 222 63.5% 11.0 19.7 126.3 244.7 1.8 0.3 Harrell, Graham 2008 Texas Tech 73.1 223 64.5% 12.5 15.9 83.4 282.4 2.2 0.4 Booty, John David 2007 USC 74.4 218 62.5% 10.8 14.1 101.7 236.6 2.5 0.3 Beck, John 2006 BYU 74.1 215 63.8% 12.9 16.1 72.6 287.3 2.2 0.5
Jimmy Clausen Overall Score = 0.492
AAAAARRRRRRGGGGHHH!!! Not looking like Mr. Franchise (by the numbers, anyway)
The NFL QB Tony Pike most compares to statistically...Carson Palmer and David Carr.
The coin flip on which way it would fully judge a match for me on Pike in our system is the Wonderlic (sick of that word yet?). Palmer scored a "safe" 26. Future NFL bust David Carr scored a magical bust-level of 24. Again, we don't have a reliable score on Tony Pike -- so we have a safe 25 Wonderlic score rated in there, but that's a huge assumption....an organizational game-changing assumption.