NoSaint Posted February 7, 2011 Share Posted February 7, 2011 Standard scores have a mean of 100 and a standard deviation, or meaningful difference, of 15 points (with the exception of the Stanford Binet scales that have a SD of 16 points). Approximately 68% of people who are administered an IQ test score within one standard deviation of the mean (85-115). I was forced to take a lot of stats and psychometrics classes in grad school Ahhh so I got the 34% but missed the iq test end. I took more math than psych- it's been a few years though. I think generally the point I was making stands even though the numbers are slightly off. If you are below average intelligence it will be harder to win in the NFL, especially against other top players, in complex schemes. All other things equal, which they never are, I would take a smarter qb. Link to comment Share on other sites More sharing options...
Johnny Hammersticks Posted February 7, 2011 Share Posted February 7, 2011 Ahhh so I got the 34% but missed the iq test end. I took more math than psych- it's been a few years though. I think generally the point I was making stands even though the numbers are slightly off. If you are below average intelligence it will be harder to win in the NFL, especially against other top players, in complex schemes. All other things equal, which they never are, I would take a smarter qb. Based on the quarterbacks from the past 10 super bowls and their performance on the Wonderlic....you may be right. QB's, Wonderlic Scores, Super Bowls Link to comment Share on other sites More sharing options...
Orton's Arm Posted February 7, 2011 Share Posted February 7, 2011 (edited) As Kelly just didn't seem to learn from his mistakes time after time, it may have hurt his odds. Am I saying it's impossible to win with a low score? Obviously not. Am I saying it is a huge challenge with below average iq? Yes. Especially when the stage is that much bigger. Is it a coincidence that no one under 22 has before? Just like being a slow runner or having a below average arm hurts. Not an insult, just common sense. Like I said, there is some luck on the test as it's pretty short. Give 3 different versions back to back to back and you will likely get 3 different scores. Also like I said, it wouldn't be unlikely to get a 15 if you answered B to every question. That kind of score is a red flag. Clearly you still don't understand the concept of using statistics (mathematical study, not football accomplishments) accounting for variability. If you can find a set like this that accounts better - more power to you. All I am saying is that someone fitting this description is more likely to have success, and larger success- not that it's impossible to break this very loose guideline. You are as far off base as the person that said you can't win without a 22 because it hasn't happened. I'm just saying it's typically quicker to swim with the current, not against it. It's kind of like saying a more accurate kicker wouldve won the superbowl- truly Norwood could have made that kick where as someone with a higher percentage could have still missed. Statistics only start to account for variability, and this set of three does better then most out there. It's the more accurate kicker. Doesn't mean he makes them all but I'd rather put the ball on his foot, even if it could be blocked, blown by a gust of wind etc... I agree that you can use statistics to predict the odds, but seldom can use them to predict the odds with 100% certainty. That being said, a few examples where a model doesn't work do not invalidate the model--unless of course the model had been intended to predict success with 100% accuracy. For example, if 80% of the new hires with engineering degrees were able to handle the technical aspects of a position, as compared with only 20% of people who had no college degrees, a hiring manager should tend to hire people with engineering degrees. (Even though you could point to individual examples of engineers who'd failed and non-graduates who'd succeeded in the position.) It's worth noting that the Wonderlic is a flawed measure of intelligence. For example, there are general knowledge questions. These types of culturally loaded questions are something one would hope to avoid on a well-designed intelligence test. (Even if there is a correlation between knowledge of such questions and g.) There are also time dump questions--an important fact to note, because test-takers are not given enough time to complete all the questions. (Unless they are very, very fast.) I could imagine, say, Dan Marino walking into the Wonderlic test without having done much, if any, prior preparation for it. He answers several normal questions correctly, then comes to a time dump question. He burns a ton of his time there, he may or not get the right answer, and finds that when the time for the test is up he's left numerous questions unanswered. Something like that would produce an artificially lowered score that would significantly understate his true level of intelligence. (Which is why it's extremely poor test design to include time dump questions on tests where participants aren't given enough time to complete all the answers.) Despite its flaws, the Wonderlic is almost certainly strongly correlated with intelligence on a broad scale. That's why it's possible to incorporate Wonderlic scores into a model designed to work most of the time; even though you obviously couldn't use them if the model had to work 100% of the time. Edited February 7, 2011 by Edwards' Arm Link to comment Share on other sites More sharing options...
NoSaint Posted February 7, 2011 Share Posted February 7, 2011 I agree that you can use statistics to predict the odds, but seldom can use them to predict the odds with 100% certainty. That being said, a few examples where a model doesn't work do not invalidate the model--unless of course the model had been intended to predict success with 100% accuracy. For example, if 80% of the new hires with engineering degrees were able to handle the technical aspects of a position, as compared with only 20% of people who had no college degrees, a hiring manager should tend to hire people with engineering degrees. (Even though you could point to individual examples of engineers who'd failed and non-graduates who'd succeeded in the position.) It's worth noting that the Wonderlic is a flawed measure of intelligence. For example, there are general knowledge questions. These types of culturally loaded questions are something one would hope to avoid on a well-designed intelligence test. (Even if there is a correlation between knowledge of such questions and g.) There are also time dump questions--an important fact to note, because test-takers are not given enough time to complete all the questions. (Unless they are very, very fast.) I could imagine, say, Dan Marino walking into the Wonderlic test without having done much, if any, prior preparation for it. He answers several normal questions correctly, then comes to a time dump question. He burns a ton of his time there, he may or not get the right answer, and finds that when the time for the test is up he's left numerous questions unanswered. Something like that would produce an artificially lowered score that would significantly understate his true level of intelligence. (Which is why it's extremely poor test design to include time dump questions on tests where participants aren't given enough time to complete all the answers.) Despite its flaws, the Wonderlic is almost certainly strongly correlated with intelligence on a broad scale. That's why it's possible to incorporate Wonderlic scores into a model designed to work most of the time; even though you obviously couldn't use them if the model had to work 100% of the time. I generally agree with this 100% - the one thing I would add is it does measure both intelligence AND preparation. Both are important for a qb, probably more then anyone else on the field. Especially in a big game like the superbowl. I want a qb that knows when it's time to buckle down in addition to the intelligence end. As the test is no surprise, a lack of prep speaks volumes. I don't think a 28 vs a 32 is a big deal, but to me a 15 vs a 30 is a major red flag. Link to comment Share on other sites More sharing options...
ajzepp Posted February 7, 2011 Share Posted February 7, 2011 Matt Leinart should have been an NFL success. 64.8 completion perectage, 35 on the wonderlic, 30+ college starts. Where does this writer mention him in the article? I wish it was this simple to predict NFL success or failure. If they added a fourth variable, "work ethic", Leinart would fail. He is a complete waste of talent. The guy was clearly more concerned with his image, living the hollywood lifestyle, and partying than he was working to become a solid NFL QB. You can have all the ability, talent, and potential in the world but if you don't WORK, then you'll fail. Link to comment Share on other sites More sharing options...
NoSaint Posted February 8, 2011 Share Posted February 8, 2011 Based on the quarterbacks from the past 10 super bowls and their performance on the Wonderlic....you may be right. QB's, Wonderlic Scores, Super Bowls amazing how standard deviations and math talk killed this debate so quickly! Link to comment Share on other sites More sharing options...
EasternOHBillsFan Posted February 8, 2011 Share Posted February 8, 2011 Matt Leinart PASSES all three of these tests, and yet he is an NFL bust, right? (35 score, 38 starts, 64.5% completion) Sorry, but I don't buy it. Link to comment Share on other sites More sharing options...
ajzepp Posted February 8, 2011 Share Posted February 8, 2011 Matt Leinart PASSES all three of these tests, and yet he is an NFL bust, right? (35 score, 38 starts, 64.5% completion) Sorry, but I don't buy it. http://forums.twobillsdrive.com/topic/126879-26-27-60-rule-of-qbs/page__view__findpost__p__2103243 Link to comment Share on other sites More sharing options...
EasternOHBillsFan Posted February 8, 2011 Share Posted February 8, 2011 (edited) http://forums.twobillsdrive.com/topic/126879-26-27-60-rule-of-qbs/page__view__findpost__p__2103243 Another, fine: Rex Grossman (29 score, 30 starts, 61% completion) This is too easy. Edited February 8, 2011 by BmoreBills Link to comment Share on other sites More sharing options...
Dorkington Posted February 8, 2011 Share Posted February 8, 2011 I guess that depends on your definition of success. I would rather have rings that records. Marino went to one SB early on, then was our B word for the rest of his career. If all you want is big numbers, Bledsoe was a "franchise" QB for us. I'd take Marino over Fitzpatrick. Also, Marino plus the 90's Bills would have done just as well as Kelly in regards to team success. Put Marino on 90's Cowboys, probably would have won a few rings there. A QB can't carry any entire team, imho. Link to comment Share on other sites More sharing options...
NoSaint Posted February 8, 2011 Share Posted February 8, 2011 Another, fine: Rex Grossman (29 score, 30 starts, 61% completion) This is too easy. i guess im not sure what argument you are making.... that qbs that dont fit those criteria are better in general? I'd take Marino over Fitzpatrick. Also, Marino plus the 90's Bills would have done just as well as Kelly in regards to team success. Put Marino on 90's Cowboys, probably would have won a few rings there. A QB can't carry any entire team, imho. very true -- pick possibly the greatest ever not to fit the criteria, and compare to one that played football in the ivy league, and is very much in the lower tier of the list in success.... and you came out ahead on your end. would you disagree that someone with below average IQ would likely have a harder time playing qb (making reads, leading men etc..), especially in the biggest games? i agree that there are many other factors, but i guess im not sure what the point you are trying to argue is? we have all accepted it is not a law of the universe but instead something that helps predict and generally manage the huge number of variables that you have to take into account when evaluating a qb. essentially by taking the three numbers you are saying someone that is mentally sharp, well prepared, consistently accurate, with a great deal of experience and limited injuries tends to be a better prospect then someone who is missing some of those qualities. i suppose i dont get the arguments against this? Link to comment Share on other sites More sharing options...
dave mcbride Posted February 8, 2011 Share Posted February 8, 2011 The thing is, people talk about statistical tendencies, but then I look at the Bills and see the following. Brohm meets the criteria and he sucks. Todd Collins met the criteria (27 starts; 64.3 percent completion rate lifetime; good Wonderlic score) and he sucked (although I do give him credit for lasting 16 seasons). Rob Johnson sucked yet he started over 30 games and had a lifetime completion rate of 64.6 percent and a wonderlic of 26. Kelly wasn't a bright guy (evidenced by his actions and his wonderlic score), yet he was good. Link to comment Share on other sites More sharing options...
hooper Posted February 8, 2011 Share Posted February 8, 2011 Wow a whole SH$% ton of numbers being thrown around here, the number that I like is the number of years that Aaron Rodgers sat on the bench....Groom a QB for a while and then let him bring you the sweetness....Go muthu fu$%#ng BILLS!!!!!!!!!!!! Link to comment Share on other sites More sharing options...
Hapless Bills Fan Posted February 8, 2011 Share Posted February 8, 2011 I believe your data is wrong Blaine Gabbert, I believe has started 26 games, When? He has only been the starter for 2 years. College season is 12 games. Plus a bowl. Link to comment Share on other sites More sharing options...
Albany,n.y. Posted February 8, 2011 Share Posted February 8, 2011 (edited) Obviously there is something wrong with the CBS stat site so let's try this again with the site ddaryl provided me... Blaine Gabbert - 29 starts...60.9% PASS/PASS Cam Newton - 20 starts...65.4% FAIL/PASS Jake Locker - 40 starts...53.9% PASS/FAIL Ryan Mallett - 37 starts...57.8% PASS/FAIL Christian Ponder - 35 starts...61.8% PASS/PASS Colin Kaepernick - 51 games....58.2% PASS/FAIL Andy Dalton - 50 starts...61.7% PASS/PASS Rick Stanzi - 38 starts...59.8% PASS/FAIL Greg McElroy - 35 games...66.3% PASS/PASS Again, thanks to ddaryl for catching the errors. This revised list shows many more options that meet 2 of 3 parts of the 26-27-60 rule: Gabbert, Ponder, Dalton and McElroy. Anyone know when the Wonderlic tests are given? PTR Looking at your original data & ddaryl's it's pretty obvious what the problem was: your source didn't include the 2010 season. Edited February 8, 2011 by Albany,n.y. Link to comment Share on other sites More sharing options...
Mango Posted February 8, 2011 Share Posted February 8, 2011 I'd take Marino over Fitzpatrick. Also, Marino plus the 90's Bills would have done just as well as Kelly in regards to team success. Put Marino on 90's Cowboys, probably would have won a few rings there. A QB can't carry any entire team, imho. COMPLETELY DISAGREE. Marino was all about Marino, never would have given the ball to Emmet or Thurman. Superbowl 26(I think), Jimbo did the same thing, determined to win the game all by himself. The guy tossed the ball something like 50 times and 4 ints. Would have been worse with that whore Marino back there Link to comment Share on other sites More sharing options...
Hapless Bills Fan Posted February 8, 2011 Share Posted February 8, 2011 (edited) Obviously there is something wrong with the CBS stat site so let's try this again with the site ddaryl provided me... Blaine Gabbert - 29 starts...60.9% PASS/PASS Cam Newton - 20 starts...65.4% FAIL/PASS Jake Locker - 40 starts...53.9% PASS/FAIL Ryan Mallett - 37 starts...57.8% PASS/FAIL Christian Ponder - 35 starts...61.8% PASS/PASS Colin Kaepernick - 51 games....58.2% PASS/FAIL Andy Dalton - 50 starts...61.7% PASS/PASS Rick Stanzi - 38 starts...59.8% PASS/FAIL Greg McElroy - 35 games...66.3% PASS/PASS Again, thanks to ddaryl for catching the errors. This revised list shows many more options that meet 2 of 3 parts of the 26-27-60 rule: Gabbert, Ponder, Dalton and McElroy. Anyone know when the Wonderlic tests are given? PTR, maybe it's a nit, but I don't see how the above can be correct. Isn't the college football season 12 games long? So Gabbert, a 2 year starter, has started in 24 regular games plus a bowl game. In 2008, he PLAYED in a few games and racked up 5 completions in 13 attempts -- but my boy Chase Daniel started all 12 and Gabbert only got the nod when the game was ridiculously out of reach. So 29 games maybe...but the rule you cite is STARTS, not games, Gabbert has 25. I think you're oscillating, the first site was inaccurate low, this site is inaccurate high Edited February 8, 2011 by Hopeful Link to comment Share on other sites More sharing options...
vincec Posted February 8, 2011 Share Posted February 8, 2011 (edited) I'd take Marino over Fitzpatrick. Ya think? COMPLETELY DISAGREE. Marino was all about Marino, never would have given the ball to Emmet or Thurman. Superbowl 26(I think), Jimbo did the same thing, determined to win the game all by himself. The guy tossed the ball something like 50 times and 4 ints. Would have been worse with that whore Marino back there People on TSW are the best. Wait, are you being serious? You would rather have Fitzpatrick than Marino? Edited February 8, 2011 by vincec Link to comment Share on other sites More sharing options...
Orton's Arm Posted February 8, 2011 Share Posted February 8, 2011 I'd take Marino over Fitzpatrick. Also, Marino plus the 90's Bills would have done just as well as Kelly in regards to team success. Put Marino on 90's Cowboys, probably would have won a few rings there. A QB can't carry any entire team, imho. I agree that Marino was a bit better than Kelly, and considerably better than Fitz. That said, the people who created the model in question stated that it's only supposed to apply to first round QBs, which clearly excludes Fitz. One of the (many) reasons why Fitz was taken in the seventh round was because the defenses he faced while playing for Harvard weren't exactly the best college defenses in the nation! Even against those lesser defenses he was still only able to complete 59.9% of his passes; which puts him right at the borderline of the model's 60% completion percentage threshold. Link to comment Share on other sites More sharing options...
nick in* england Posted February 9, 2011 Share Posted February 9, 2011 Someone tell the how the 26-27-60 rule works for the following and I'll judge it's success: JP Losman Trent Edwards Todd Collins Drew Bledsow Alex Van Pelt Doug Flutie Rob Johnson Billie Jo Hobert Kelly Holcomb Travis Brown Brian Brohm Levi Brown Link to comment Share on other sites More sharing options...
Recommended Posts