John P. Lopez is an expert practitioner in Aggie logic, which is to say, an expert in reasoning that isn’t logical at all. Perhaps the most famous graduate of TAMU’s defunct Journalism department (snort- and Tatu was the greatest talent in indoor Dallas soccer history), he has been toiling in the more lucrative alternative media (radio and internet) the past few years. This allows him to pander to Ag fans in the subscription commentary market, while still selling content to widecast media (Sports Illustrated’s webpage). His SI niche is NFL commentary, Houston-centric by nature, not design. His first SI article, last fall, explained why the Texans were very smart to draft Mario Williams instead of VY. Sure, if you’re not focused on winning or marketing, which may actually be the case with the Texans.
This displays a consistent theme of Lopez’ writing – "VY will fail". When with the Chronicle, he explained why the Texans were committing to Carr and not drafting Young, and kept touting Carr during the season http://blogs.chron.com/lopezblog/archives/2006/10/would_you_trade.html
Until he stopped… http://blogs.chron.com/lopezblog/archives/2006/12/does_mcnair_mak.html
But, even after the beatings, he never really believed in Vince, as his first SI article showed. His first Texags.com private board post provided some Vince Wonderlic bashing…
http://www.texags.com/mwr/articles.asp?issue_id=825
which is, of course, sweet nectar to the Ag fanbase.
This obsession over Vince, and Lopez’ fascination with the Wonderlic as a validation of his VY hatred, plays out again in Lopez’ latest work, where he explains the magic formula for predicting QB success in the NFL-
http://sportsillustrated.cnn.com/2010/writers/john_lopez/07/08/qb.rule/index.html
He claims that three metrics, available before the draft to all teams, can reduce the variability in the drafting of quarterbacks. These metrics are Wonderlic score, games started, and college completion percentage (Note- he makes it clear through example that the level of college play- D-1A or D-1AA- makes no difference in the use of this formula). Wow. If true, this is the NFL scouting discovery of the decade, similar to when Gil Brandt discovered that huge guys that run really fast work out better than huge guys that are slow, even if they come from bad college teams. But, is he right? Does it work?
In his article, he gives examples. He has a list of several QBs since 1998 that exceed his defined metrics – minimum Wonderlic of 26, minimum of 27 college starts, and minimum passing percentage of 60%. He also provides a list of QBs that fall short in one or more metric. The first list definitely has better QBs than the second list. He then has a third list of a few guys who are doing well despite falling short in one or more metric (can you guess which of the latter two lists Vince is on? Come on, this is John Lopez! Of course, he’s on the list with Akili Smith!).
This Lopez theory is a complete failure. Here’s why.
1. He uses selected examples. He does not survey all QB prospects since 1998, nor all high drafted QBs. Why were first round picks like Patrick Ramsey, Kyle Boller, Rex Grossman, Carson Palmer (!), J.P. Losman, Alex Smith, Aaron Rodgers, Jason Campbell, Brady Quinn, Matt Leinart and Drew Stanton not included? If they supported his argument, he would have used them. He doesn’t. There seems to be no rhyme nor reason to which players he surveys. He uses some high draft picks, and some low draft picks (Ryan Fitzpatrick?), and even an undrafted player (Romo) to support his argument. If you were really willing to test your hypothesis, you would apply the test to a complete list of players that meet a common standard (perhaps 1st round picks, or drafted QBs, or even Combine QBs), and see how it worked. That would tell you how it worked as a forecast tool.
2. The demarcation lines for his metrics are arbitrary. Look at the list he provides. Is the difference between a 26 Wonderlic and a 25 Wonderlic that significant? How about the 27 starts? If Lopez weren’t just ignoring Matt Leinart, I would suspect he set the bar at 27 simply to keep Leinart off the good list. The two best pro QBs today are Manning and Brady. Brady had a 33 Wonderlic, completed over 60% of his passes, and started…25 games at Michigan. What good is a QB rating tool that rejects Tom Brady?
How about 60%? Doesn’t he know there is a huge difference between 60% completions in a 1990s college offense, and 60% completions today? Why not just say "top 25% in completion percentage"?
If you asked Huck how these three metrics relate to NFL quarterbacking, he would give you a multiple regression, and tell you the forecast variability and strength of each factor. Lopez just picks arbitrary examples to support his argument and runs with it.
3. He ignores the obvious. Go back and look at the lists of stars and busts. How many of the stars didn’t have to start until they had a few years in the NFL as backups? How many of the busts were thrown into the fire? Think that might have something to do with it?
If Lopez had discovered a tool that helped GMs differentiate between college QBs they were comparing (pool of all QBs, or just the pool of touted QBs), he might have something. If he had a tool that meant something to QBs preparing for the league, he might have something. He has neither. He has nothing.
The news isn’t all bad for Lopez. The day will surely come when the world realizes Vince Young can’t be an effective NFL quarterback. That day will be in 2023.