Why John Beck is Likely to Win the Job

[Oldfan]

ASF ~ You can take pot-shots at my methods all you want. I expect it. People are correct that I use selective stats. They view this as a defect, and I view this as the point of the exercise. I am testing various ideas relating to short sample sizes and selective stats as predictors. Some of these have far more grounding than people suspect.

The only thing you can test for by offering your creative stats in this forum is to see how many people take them seriously.

Small sample sizes can be useful when there is a logical basis for an hypothesis. A small sample that supports the logic, would encourage one to spend the time for a larger study. For example, I believed that most NFL teams were drafting and keeping their best players. So, I did a quick study of the 2006 All Pro team and found that 53 players of the 56 member squad still represented the team that drafted them.

One year is a small sample size, but the one-sided result would be enough to indicate that my time wouldn’t be wasted if I wanted to collect a 20-year sample to get a more reliable average to support the theory.

[ouvan59]

This is one reason that I'm pretty confident that Beck's performance looks better than it is. In his only two opportunities against a legit NFL starting defense playing hard (Ravens), he looked very poor. .

Are you effing serious? His two opportunities resulted in one touchdown and one badly thrown ball that skipped at Moss' feet. So essentially his looking very poor is one pass. That is a really, really, really, really small sample size.

I should also add that Grossman bounced two passes against the Ravens. I'm guessing those were great plays by him though.

[Atlanta Skins Fan]

The only thing you can test for by offering your creative stats in this forum is to see how many people take them seriously.

Actually, I post them for entertainment value, especially in OPs.

Most of my threads are for entertainment. I enjoy turning large orderly groups into brawls. This group is especially fun.

Occasionally I get some very useful feedback amid all the slapstick and pies in faces. It's like Animal House with a few nuggets of good stuff.

[tiger187126]

Actually, I post them for entertainment value, especially in OPs.

Most of my threads are for entertainment. I enjoy turning large orderly groups into brawls. This group is especially fun.

Occasionally I get some very useful feedback amid all the slapstick and pies in faces. It's like Animal House with a few nuggets of good stuff.

so you basically just defined trolling.

[Oldfan]

Actually, I post them for entertainment value, especially in OPs.

Most of my threads are for entertainment. I enjoy turning large orderly groups into brawls. This group is especially fun.

Occasionally I get some very useful feedback amid all the slapstick and pies in faces. It's like Animal House with a few nuggets of good stuff.

Gotcha.

I make predictions. Most of them are made seriously, like this one, but every now and then I throw in a trumped up argument for a very unlikely event. One of my favorites was the prediction that Pat Ramsey would return as our QB. I think that one generated about 650 replies during a slow period in the offseason.

[Atlanta Skins Fan]

so you basically just defined trolling.

No, I'm creating very entertaining debate. There's a big difference.

Most of my threads contain a seed of really surprising info that I believe in. But it rides in dressed as a clown, riding on a donkey. It invites attack, but is surprisingly resilient. The seed that is strong will be recognized, interpreted and reasserted by the second wave of attack, attacking the first wave. It's almost always this way.

Some pretty great threads have resulted.

[bird_1972]

Actually, I post them for entertainment value, especially in OPs.

Most of my threads are for entertainment. I enjoy turning large orderly groups into brawls. This group is especially fun.

Occasionally I get some very useful feedback amid all the slapstick and pies in faces. It's like Animal House with a few nuggets of good stuff.

Disappointed to hear you say that, ASF. I actually regard (or regarded) you as one of the more serious posters here on ES that genuinely wanted to foster discussion about football in an intelligent way. So you're basically saying that you just engineer convoluted statistical arguements for entertainment value?

Very disappointing indeed.

[tiger187126]

No, I'm creating very entertaining debate. There's a big difference.

Most of my threads contain a seed of really surprising info that I believe in. But it rides in dressed as a clown, riding on a donkey. It invites attack, but is suprisingly resilient. The seed that is strong will be recognized, interpreted and reasserted by the second wave of attack, attacking the first wave. It's almost always this way.

Some pretty great threads have resulted.

It's surprisingly resiliant because you:

A) make up the rules

change them if they don't work out for you

And yet you claim to want to stir some **** up with what you say.

No, I think you are just a clown riding a donkey disguising a troll.

[Atlanta Skins Fan]

Disappointed to hear you say that, ASF. I actually regard (or regarded) you as one of the more serious posters here on ES that genuinely wanted to foster discussion about football in an intelligent way. So you're basically saying that you just engineer convoluted statistical arguements for entertainment value?

Very disappointing indeed.

No, you've got it backward. I develop convoluted statistical studies quite seriously, and these take a lot of time. The way in which I post them is designed to create a big debate, which is both entertaining and useful by shaking out nuggets of perspective that I had not considered.

I learned this approach accidentally. Years ago I had some serious posts which tanked because they weren't provocative enough. It's pointless to post a serious thread and get 10 responses. Then I made a few posts that produced incredible mob attacks, and these were more successful.

What I learned is that you won't get the nuggets of great feedback (pro or con) without a vigorous debate. One way to create vigorous debate is to provoke opponents into attacks. It's best done in a light and entertaining style.

The style is for entertainment value, to stimulate debate. But the fundamental insight, the seed of the post, is real. I'm testing that idea and seeing how it holds up under harsh debate.

[Oldfan]

...The style is for entertainment value, to stimulate debate. But the fundamental insight, the seed of the post, is real. I'm testing that idea and seeing how it holds up under harsh debate.

There's nothing wrong with that. I haven't used that approach on this board, but I have in other debate forums.

If your theory can survive the harsh critics populating internet boards, you gain confidence in it.

[Pelado]

I learned this approach accidentally. Years ago I had some serious posts which tanked because they weren't provocative enough. It's pointless to post a serious thread and get 10 responses. Then I made a few posts that produced incredible mob attacks, and these were more successful.

I've definitely seen this. People are less likely to respond without provocation of some kind.

[Atlanta Skins Fan]

If your theory can survive the harsh critics populating internet boards, you gain confidence in it.

Right, and I actually do learn from critics at times. For example, one flaw in my original college QB model was that it didn't account for gimmick offenses (run 'n' shoot, etc.). This allowed Colt Brennan to slip through and pass the original model.

I hardly ever watch college football, but I was able to learn from those who do. They pointed out that various systems are gimmick offenses designed for almost continuous passing, and that this inflates a QB's stats. That QB will typically not transition well to the NFL.

Another great tip I get from others is to look at so-and-so player as an example who either defies or supports the theory. The members of the forum collectively have their eyes on many more players than I do. So, threads will very often surface very interesting test cases, which will sometimes blow up the theory or give it instant credibility with a strong data point.

A lot of people here bring expert insights in this way. Then there are the "Grossman sux dude" crowd. While those posts are tedious, they tend to incite a counter-attack by someone with something interesting to offer, supporting the original case. Then that post will incite a thoughtful response, attacking Grossman more intelligently, or attacking a fundamental idea (e.g. passes per drive) more intelligently.

If you watch great threads, they work in waves like this. The McElroy thread was one of the very best.

The main trouble with the Grossman debate (vs McElroy) is that people are too sure that they know "who Grossman is," usually in the negative. This dumbs down the debate. Far too many QBs have surprising late careers to be so sure about any QB, especially one who has shown a lot of promise at times. As for myself, I have a theory about who Grossman is, but it's only that. I think there's a reasonable chance that Grossman will flame out and show "Bad Rex" all over again. But, I think there's a better chance that he'll figure out (or has already figured out) how to play "Good Rex" most of the time, with great help from the Shanahans and their system.

[Pelado]

Beck clearly has upside. He's also a guy who did very well in my original college QB model (but not the sophomore model). I've been curious about interpreting the differences between those projections. My working theory is that QBs who pass the sophomore model are "natural elite QBs," while those who pass the original model, while failing the sophomore model, are "learning QBs." Some of these "learning QBs" have made it in the NFL, while many have busted. They may need more time and more of a perfect system fit to succeed.

How do Steve Young, Rich Gannon, Brad Johnson, and Trent Dilfer grade out in your original college model and your sophomore model?

[Oldfan]

Right, and I actually do learn from critics at times. For example, one flaw in my original college QB model was that it didn't account for gimmick offenses (run 'n' shoot, etc.). This allowed Colt Brennan to slip through and pass the original model.

I hardly ever watch college football, but I was able to learn from those who do. They pointed out that various systems are gimmick offenses designed for almost continuous passing, and that this inflates a QB's stats. That QB will typically not transition well to the NFL.

Another great tip I get from others is to look at so-and-so player as an example who either defies or supports the theory. The members of the forum collectively have their eyes on many more players than I do. So, threads will very often surface very interesting test cases, which will sometimes blow up the theory or give it instant credibility with a strong data point.

A lot of people here bring expert insights in this way. Then there are the "Grossman sux dude" crowd. While those posts are tedious, they tend to incite a counter-attack by someone with something interesting to offer, supporting the original case. Then that post will incite a thoughtful response, attacking Grossman more intelligently, or attacking a fundamental idea (e.g. passes per drive) more intelligently.

If you watch great threads, they work in waves like this. The McElroy thread was one of the very best.

The main trouble with the Grossman debate (vs McElroy) is that people are too sure that they know "who Grossman is," usually in the negative. This dumbs down the debate. Far too many QBs have surprising late careers to be so sure about any QB, especially one who has shown a lot of promise at times. As for myself, I have a theory about who Grossman is, but it's only that. I think there's a reasonable chance that Grossman will flame out and show "Bad Rex" all over again. But, I think there's a better chance that he'll figure out (or has already figured out) how to play "Good Rex" most of the time, with great help from the Shanahans and their system.

In a nutshell, my approach to grading QBs goes like this:

1) Ignore the bandwagon opinions;

2) Use statistics only to verify what can be seen on the field;

3) Realize that all QB performances are heavily influenced by team and scheme;

4) Grade the QB as a scout would with heavy emphasis on the tangibles;

5) Turn a deaf ear to anyone who claims expertise on a QB’s intangibles;

6) Realize that, although the intangibles can’t be reliably graded, they are important;

7) Realize that you might never know if you are right when you support a QB because it is rare when one is lucky enough to fall into an ideal support system.

[santanathegreat]

This has been pointed out to ASF on more than one occasion. ASF is a stat cherry picker, and can throw out some statistics that aren't even statistics at all. Almost every stat he posts is cherry picked in some way. Don't get too upset about it, its the way of ASF.

What I don't get are the "well at least he supports his posts with facts" responses trying to defend him.

Like many of the stats he posts are complete and utter BS. A half-decent statistician would throw pretty much of all of his "analysis" in the crapper. No offense to the guy...I see him more as an entertainer than a knowledgeable stats guy, but the other people who try and use his numbers to defend Grossman and ASF afterwards are much more annoying.

[dballer]

Right, and I actually do learn from critics at times. For example, one flaw in my original college QB model was that it didn't account for gimmick offenses (run 'n' shoot, etc.). This allowed Colt Brennan to slip through and pass the original model.

I hardly ever watch college football, but I was able to learn from those who do. They pointed out that various systems are gimmick offenses designed for almost continuous passing, and that this inflates a QB's stats. That QB will typically not transition well to the NFL.

Another great tip I get from others is to look at so-and-so player as an example who either defies or supports the theory. The members of the forum collectively have their eyes on many more players than I do. So, threads will very often surface very interesting test cases, which will sometimes blow up the theory or give it instant credibility with a strong data point.

A lot of people here bring expert insights in this way. Then there are the "Grossman sux dude" crowd. While those posts are tedious, they tend to incite a counter-attack by someone with something interesting to offer, supporting the original case. Then that post will incite a thoughtful response, attacking Grossman more intelligently, or attacking a fundamental idea (e.g. passes per drive) more intelligently.

If you watch great threads, they work in waves like this. The McElroy thread was one of the very best.

The main trouble with the Grossman debate (vs McElroy) is that people are too sure that they know "who Grossman is," usually in the negative. This dumbs down the debate. Far too many QBs have surprising late careers to be so sure about any QB, especially one who has shown a lot of promise at times. As for myself, I have a theory about who Grossman is, but it's only that. I think there's a reasonable chance that Grossman will flame out and show "Bad Rex" all over again. But, I think there's a better chance that he'll figure out (or has already figured out) how to play "Good Rex" most of the time, with great help from the Shanahans and their system.

And yet at times you seem to be the leader of the "Beck sux dude" crowd. Like having an OP labeling him "Captain Checkdown", or was it "Admiral Checkdown" after about a half of a preseason game, as if you know "who Beck is", always in the negative. I think the point people are trying to make about your stats is that you already have a theory, which is "Rex is better than Beck, and it's not close," and then you find stats (using that term loosely) to support your theory.

The problem I see is that you are quantifying things that are not meant to by quantified. Stats like "passes per drive" and "passes per game" are entirely irrelevant. Even if you have them over a decent period of time they are still irrelevant, at least for this argument. I am not even going to comment on "outstanding qb plays and horrible qb plays". Median air distance per "catchable pass"? First of all, of course, you are determining the catchable pass, so I am guessing, because you made it "catchable" pass, you aren't including Becks passes to Stallworth and Davis. But in reality just attempting the deep ball often has the same effect as completing it. It makes the defense respect it. Also, as the coaches are judging, they want to see the QB run the offense the way they design it, and in this offense there are supposed to be a lot of short passes. Not that I don't appreciate your effort, but sometimes in sports you need to just watch, see how everyone else responds to the QB, watch the throws he makes, where he makes them, and how comfortable he looks making them. The only stats that are truly relevant when it comes to a QB is completion percentage, Td/Int ratio, and Ws and Ls. You just can not use your "advanced" stats to determine the better QB in preseason games where they are sharing snaps for a half or 3 quarters, there just isn't enough data for anything to work like that.

[Atlanta Skins Fan]

How do Steve Young, Rich Gannon, Brad Johnson, and Trent Dilfer grade out in your original college model and your sophomore model?

Dilfer did poorly in my sophomore model. I don't think I went back that far with the original model.

Young and Gannon pre-date the sophomore model (1990-2011). Also I typically focused on QBs picked in the first 32 picks. (Gannon was 4th round, Johnson was 9th round)

I haven't established a model for picking QBs after the first round. The original model and sophomore model extension was designed to screen QBs who were drafted in the first round or who were considered candidates to be drafted in the first round. This approach allows scouting to create a short list, then use the model to grade the short list.

Not factoring scouting can create some big misses, especially with the original model.

The sophomore model is very strong, and may hold up better than the original model even for QBs drafted below the first round. Its strength is identifying naturally dominant QBs, as opposed to QBs who learn how to be dominant. (That's my interpretation.) Grossman is the former, Beck the latter, per this theory.

[Pelado]

Dilfer did poorly in my sophomore model. I don't think I went back that far with the original model.

Young and Gannon pre-date the sophomore model (1990-2011). Also I typically focused on QBs picked in the first 32 picks. (Gannon was 4th round, Johnson was 9th round)

I haven't established a model for picking QBs after the first round. The original model and sophomore model extension was designed to screen QBs who were drafted in the first round or who were considered candidates to be drafted in the first round. This approach allows scouting to create a short list, then use the model to grade the short list.

Not factoring scouting can create some big misses, especially with the original model.

The sophomore model is very strong, and may hold up better than the original model even for QBs drafted below the first round. Its strength is identifying naturally dominant QBs, as opposed to QBs who learn how to be dominant. (That's my interpretation.) Grossman is the former, Beck the latter, per this theory.

Could you post a list of which QB's have passed each model?

[Atlanta Skins Fan]

Could you post a list of which QB's have passed each model?

Here's the sophomore model thread. Scroll down to post #38 for a full list from v1.0 of the sophomore model.

I've made some refinements, but that should give you the picture.

I don't think I've posted a full list for the original model, which went through a few versions, starting around 2006.

Also, I have limited time to get into this, so I'm trying to oblige your request by pointing to prior posts. I will need to decline requests for new lookups not already done, for time reasons right now.

On topic, if you want to know why Beck failed the sophomore model and passed the original model, compare his college stats by year:

http://www.sports-reference.com/cfb/players/john-beck-1.html

He's what I call a "learner" or an "improver." Starts off not elite, then has stats later that are elite. These guys are frequently busts, but not always. See Tom Brady.

[MumboSauce]

hmmmm....uhmm...darn. i agree with OldFan. lol that was tough.

[GWinSkins83]

Here's the sophomore model thread. Scroll down to post #38 for a full list from v1.0 of the sophomore model.

I've made some refinements, but that should give you the picture.

I don't think I've posted a full list for the original model, which went through a few versions, starting around 2006.

Also, I have limited time to get into this, so I'm trying to oblige your request by pointing to prior posts. I will need to decline requests for new lookups not already done, for time reasons right now.

On topic, if you want to know why Beck failed the sophomore model and passed the original model, compare his college stats by year:

http://www.sports-reference.com/cfb/players/john-beck-1.html

He's what I call a "learner" or an "improver." Starts off not elite, then has stats later that are elite. These guys are frequently busts, but not always. See Tom Brady.

You can say that about Luck as an improver.

[Oldfan]

hmmmm....uhmm...darn. i agree with OldFan. lol that was tough.

Booooooooo!

[BADKCLARK]

I don't know how this will add to the discussion, but there have been comments surrounding "Arm Strength" and other things. Having watched John Beck play all thru college and into the Pros, the reason his arms don't "seem" strong is because he throws, and by extension, moves so effortlessly, that it looks weaker. But, in fact, his throws are so strong from the wrist, that he doesn't need a lot of movement on them. Scouts coming out of college hinted at this a lot.

I take you back to the Draft in 06, and a link provided by Phinsider.

http://www.thephinsider.com/2007/5/2/0358/03650

In there is another link to John's 70 yard throw in 2005 at BYU. It was a hail-m pass to end a half that was incomplete, but was a great demonstration of his arm strength. I also remember the reports of his throwing a football thru the uprights at the Indoor practice facility at BYU. This was from his knees from HIS 45 (behind the 50). Consider that's 65 yards from the upright, just to start, and getting it thru the uprights is something considerable.

Also, in the Combine, he had the 2nd fastest velocity on this throws of any of the QB's tested. He's no noodle-arm.

I also watched his first season in Miami. What a waste. The 2nd worst team in NFL history, and John was thrown to the wolves. Cameron is solely responsible for making Beck's career path harder than it was. Folding under pressure, he took Beck from the Scout team (they were trying to groom him along slowly like you should with most rookies) 1 WEEK before he was to start. Behind a line that was decimated and had no running game support at all (starters were out).

I will say that, during this time, I recall the Dolphins Steelers game. It was in one of the worst conditions I've EVER seen a football game. Complete storms, and the Steelers grounds crew had just dropped new turf during the week. There were 1 foot divots going up all over the place, and the field was terrible. Even though the Phins lost the game (Beck started), he actually outplayed Big Ben R. For the conditions they were in, it was something I remembered standing out to me. Cameron made a ton of coaching errors that year, and like John "I know what happened" there. Not as much as he did, but I have followed him his entire career.

I freely admit I'm a homer. I really want to see him get his chance, and feel he's done a great job here. It's like seeing Beck of 06. He completely dismantled Oregon 38-8 in the Las Vegas Bowl that year. Given the right team to work with, Beck can be an extremely dominating QB. I've no doubt that if he gets his opportunity here, he'll be ready for it. He's a workhorse, a gym rat, and all of the statements coming out of camp regarding his work ethic and staying late smack of everything I know about Beck. And it's not just "to learn." Any QB resting on his laurels will get relegated to the back door, or be a career backup. That isn't what Beck is preparing for. He's been preparing to be a starter his whole career.

I hope we get to see it. If he craps out, Rex will always be there. But, Beck has done his diligence, and it's up to the coaches.

[Atlanta Skins Fan]

You can say that about Luck as an improver.

Luck as a sophomore easily passed the sophomore model. Post #1 shows passing grades for Bradford, Grossman, Brees, Luck, both Mannings, Freeman, Roethlisberger, Rodgers and Rivers, in that order. All those QBs are either considered elite, on the way to being elite, or have appeared in a Super Bowl. Grossman is the biggest outlier by NFL career. The sophomore model suggests that he is deeply underrated in the NFL. I've taken that to be a unique downgrade due to bad receivers / system / OC in Chicago. He's a very interesting test case.

[GWinSkins83]

Luck as a sophomore easily passed the sophomore model. Post #1 shows passing grades for Bradford, Grossman, Brees, Luck, both Mannings, Freeman, Roethlisberger, Rodgers and Rivers, in that order. All those QBs are either considered elite, on the way to being elite, or have appeared in a Super Bowl. Grossman is the biggest outlier by NFL career. The sophomore model suggests that he is deeply underrated in the NFL. I've taken that to be a unique downgrade due to bad receivers / system / OC in Chicago. He's a very interesting test case.

So you couln't do that for Beck. I mean come on now Beck's coaching staff was BAD, OLINE=BAD, WR=BAD, RB=BAD. And Luck RS Freshman year was not elite it wasn't even Landry Jones good and Jones had more pressure and tougher position than Luck. You say Bears had a bad WRs corp but berrian and muhammed was above adverage WRs than. He had a good running game behind him and for the most part had great field position cause of the Defense or Devin Hester. I think you are overrating Grossman to much. Just last year you was probably mad at Mike for benching McNabb for Grossman. You probably was like this GROSSMAN? FORREAL GROSSMAN? FIRE MIKE NOW! But he have one great half against the Cowboys who secondary sucked and still does even when they beat us on a pick that he threw lol. Now you coming in talking about Rex is ELITE if he stays consitent. Well he is going to stay consitent to what he is he is going to have 4 games where its just his day or to have a good day to inferior competition. But when it is a tough challenge he's going to fold and chock and make bone head mistakes that costs the team a win. Thats what he is and he is consitent with being that.

---------- Post added August-30th-2011 at 03:24 PM ----------

So you couln't do that for Beck. I mean come on now Beck's coaching staff was BAD, OLINE=BAD, WR=BAD, RB=BAD. And Luck RS Freshman year was not elite it wasn't even Landry Jones good and Jones had more pressure and tougher position than Luck. You say Bears had a bad WRs corp but berrian and muhammed was above adverage WRs than. He had a good running game behind him and for the most part had great field position cause of the Defense or Devin Hester. I think you are overrating Grossman to much. Just last year you was probably mad at Mike for benching McNabb for Grossman. You probably was like this GROSSMAN? FORREAL GROSSMAN? FIRE MIKE NOW! But he have one great half against the Cowboys who secondary sucked and still does even when they beat us on a pick that he threw lol. Now you coming in talking about Rex is ELITE if he stays consitent. Well he is going to stay consitent to what he is he is going to have 4 games where its just his day or to have a good day to inferior competition. But when it is a tough challenge he's going to fold and choke and make bone head mistakes that costs the team a win. Thats what he is and he is consitent with being that.