QB dept. : " Sheer luck buoyed Mark Sanchez's stats " ~ ~ ~

[JohnnyHector]

And what if that doesn't happen? Then what?

Outlier.

[Matt39]

Outlier.

Mark Sanchez- the biggest outlier in the history of football.

I'll take it honestly.

[dbatesman]

And what if that doesn't happen? Then what?

Then nothing. Improvement would entail throwing fewer passes to defenders entirely, be they catchable or uncatchable, and I see nothing in Graham's piece nor the FO writeup suggesting that such an improvement would be impossible or even improbable.

[Matt39]

Then nothing. Improvement would entail throwing fewer passes to defenders entirely, be they catchable or uncatchable, and I see nothing in Graham's piece nor the FO writeup suggesting that such an improvement would be impossible or even improbable.

I know that. And me being a fan more so than I statistician...I can hope for the best. You guys seem to think its impossible, no?

Didn't you say the Jets season was over the day they drafted Sanchez?

[dbatesman]

I know that. And me being a fan more so than I statistician...I can hope for the best. You guys seem to think its impossible, no?

Didn't you say the Jets season was over the day they drafted Sanchez?

Now you've completely lost me.

[Matt39]

Now you've completely lost me.

Arent you on the record saying the Jets cant win with Sanchez?

[dbatesman]

Arent you on the record saying the Jets cant win with Sanchez?

What does that have to to with anything? Also, no.

[Matt39]

What does that have to to with anything? Also, no.

I just figured you don't see Sanchez improving at all accuracy wise(judging by most of your posts), while two posts ago...you're saying it isn't impossible or improbable.

Or should I not lump your thoughts in directly with FO?

[dbatesman]

I just figured you don't see Sanchez improving at all accuracy wise(judging by most of your posts), while two posts ago...you're saying it isn't impossible or improbable.

Or should I not lump your thoughts in directly with FO?

You should absolutely lump my thoughts in with FO, because neither of us attributes any predictive value whatsoever to the dropped-INT stat.

[Matt39]

You should absolutely lump my thoughts in with FO, because neither of us attributes any predictive value whatsoever to the dropped-INT stat.

I don't really see any reason reporting it then.

[dbatesman]

I don't really see any reason reporting it then.

If you're honestly arguing that a stat's inability to augur the future renders it irrelevant, this conversation was over long before it started.

[CTM]

I know that. And me being a fan more so than I statistician...I can hope for the best. You guys seem to think its impossible, no?

I like sanchez, have since the rose bowl in fact.. doesn't mean I have to pretend he's not a terrible qb more often then not..

he's very young and inexperianced, so of course he can get better..

[#27TheDominator]

If you're honestly arguing that a stat's inability to augur the future renders it irrelevant, this conversation was over long before it started.

It does kind of render the whole making a mint gambling by using these "stats" thing kind of hard to believe.

[Matt39]

If you're honestly arguing that a stat's inability to augur the future renders it irrelevant, this conversation was over long before it started.

:lol:

[CTM]

It does kind of render the whole making a mint gambling by using these "stats" thing kind of hard to believe.

What dbatesman is saying is that ithis particular stat doesn't predict sanvhez's future as a qb outside of the idea that he is likely to regress to the mean next year and have more of his potential int's caught.. He could very likely throw 15 potential int's next year rather the 28 or whatever this year.. But if he does throw 28 next year, bet on more the 13 being actually caught

[JohnnyHector]

If you're honestly arguing that a stat's inability to augur the future renders it irrelevant, this conversation was over long before it started.

:lol:

Personally, I would have gone with "soothsay" instead.

[Matt39]

Personally, I would have gone with "soothsay" instead.

I must be too conservative for this board.

[#27TheDominator]

What dbatesman is saying is that ithis particular stat doesn't predict sanvhez's future as a qb outside of the idea that he is likely to regress to the mean next year and have more of his potential int's caught.. He could very likely throw 15 potential int's next year rather the 28 or whatever this year.. But if he does throw 28 next year, bet on more the 13 being actually caught

So nobody is getting rich off of "this particular stat"?

BTW, I gave up posting from my whiteberry. I have enough trouble with the regular keyboard.

[CTM]

So nobody is getting rich off of "this particular stat"?

BTW, I gave up posting from my whiteberry. I have enough trouble with the regular keyboard.

Maybe.. it's something you can factor into to make a better decision on any number of wagers. Team overall wins, SB odds, division winner odds. How about say O/U on # of sanchez interceptions this season, a normal bettor might consider how much he thinks Sanchez will improve this season and maybe factor in potentional changes in offensive philosophy, a sharp bettor would want more information. And of course, I'd hope anyone here can see that this stat would be particularly valuable in this kind of wager. (Fantasy football too)

and lol.. i still iphone it.. even thought i can't type on the damn thing

[RutgersJetFan]

It does kind of render the whole making a mint gambling by using these "stats" thing kind of hard to believe.

You know there's a fine line between making use of the adjusted sack rate numbers for gambling matchups and "zomg they're saying Sanchez won't be good this is stupid."

[T0mShane]

This whole discussion reminds me of the conversation in White Noise where the father is talking to his kid about the rain and the kid questions how anyone knows it's really raining and the father gets frustrated and says "Is it raining or isn't it?" And the kid is all smug, like, that was the whole point of the discussion they just had. DeLillo, baby. Hate him.

[dbatesman]

Personally, I would have gone with "soothsay" instead.

Too many bad memories tied up in that word.

[RutgersJetFan]

Too many bad memories tied up in that word.

It's not your fault.

[#27TheDominator]

Maybe.. it's something you can factor into to make a better decision on any number of wagers. Team overall wins, SB odds, division winner odds. How about say O/U on # of sanchez interceptions this season, a normal bettor might consider how much he thinks Sanchez will improve this season and maybe factor in potentional changes in offensive philosophy, a sharp bettor would want more information. And of course, I'd hope anyone here can see that this stat would be particularly valuable in this kind of wager. (Fantasy football too)

and lol.. i still iphone it.. even thought i can't type on the damn thing

Right, but go back to Bergen's point. How can I take a "dropped interception" seriously when they do not account for INTs that were thrown directly to a WR that then had the ball wrestled away? I happened on more than 10% of his ACTUAL INTs. Isn't that "statistically significant" to you numbers monkeys? I remember there was a stat about some turnover machine having x amount of picks off of the hands of WRs. That would have to be considered too. It is at least as significant as dropped INTs and either one without considering the other pretty much useless.

[CTM]

Right, but go back to Bergen's point. How can I take a "dropped interception" seriously when they do not account for INTs that were thrown directly to a WR that then had the ball wrestled away? I happened on more than 10% of his ACTUAL INTs. Isn't that "statistically significant" to you numbers monkeys? I remember there was a stat about some turnover machine having x amount of picks off of the hands of WRs. That would have to be considered too. It is at least as significant as dropped INTs and either one without considering the other pretty much useless.

I'm sure those occurances happenned to every player. shEli seems to be the victim of it often, maybe not rip out of hands, but the WR has the ball bounce of his chest and then picked.

And sure, a stat like that would be interesting (particularly like in the dropped INT case, where one player is such an outlier!), but to say one is meaningless without the other is silly. I'm sure the Jets coaches are interested in the fact that Sanchez had 15 potential INT's dropped whereas Henne had 1. Jets fans should be too, as it means if Sanchez comes out next season as the same player, we can expect 20+ INT's next season for sure.

[#27TheDominator]

I'm sure those occurances happenned to every player. shEli seems to be the victim of it often, maybe not rip out of hands, but the WR has the ball bounce of his chest and then picked.

And sure, a stat like that would be interesting (particularly like in the dropped INT case, where one player is such an outlier!), but to say one is meaningless without the other is silly. I'm sure the Jets coaches are interested in the fact that Sanchez had 15 potential INT's dropped whereas Henne had 1. Jets fans should be too, as it means if Sanchez comes out next season as the same player, we can expect 20+ INT's next season for sure.

It's interesting, but meaningless. They watched every play - enough to determine what INTs were blatant drops - but don't mention that some 10-20% of the guys ACTUAL INTs were bogus? I call BS. If you are watching every play, then BS like the refs saying those were INTS or that Calvin Johnson didn't catch that TD in week 1 should be taken into consideration.

[JiFtheOracle]

Jets fans should be too, as it means if Sanchez comes out next season as the same player, we can expect 20+ INT's next season for sure.

And if this happens, I'm sure if we wanted to we could make a case that Sanchez's high INT numbers were because defenders got lucky and Sanchez was unlucky.

[CTM]

It's interesting, but meaningless. They watched every play - enough to determine what INTs were blatant drops - but don't mention that some 10-20% of the guys ACTUAL INTs were bogus? I call BS. If you are watching every play, then BS like the refs saying those were INTS or that Calvin Johnson didn't catch that TD in week 1 should be taken into consideration.

They probably have a stat like that too. Like I said, if Sanchez had 2 and Eli had 3 and Brady 1, that's not very meaningful to discuss in terms of Sanchez. This one is interesting cause Sanchez was 67% higher then the guy with the 2nd most dropped INT's and 15x greater then the lowset (henne). Add in the fact that Sanchez is a premier name on a premier team, and I don't understand why it's so confusing that someone would write about this, and why some people would find it interesting.

[CTM]

And if this happens, I'm sure if we wanted to we could make a case that Sanchez's high INT numbers were because defenders got lucky and Sanchez was unlucky.

I'm sure if Henne had Sanchez's switched spots and Henne had 15 dropped INT's and Sanchez had Henne's 1, this would be a very different conversation.

[#27TheDominator]

They probably have a stat like that too. Like I said, if Sanchez had 2 and Eli had 3 and Brady 1, that's not very meaningful to discuss in terms of Sanchez. This one is interesting cause Sanchez was 67% higher then the guy with the 2nd most dropped INT's and 15x greater then the lowset (henne). Add in the fact that Sanchez is a premier name on a premier team, and I don't understand why it's so confusing that someone would write about this, and why some people would find it interesting.

The whole purpose of the stat is to show how many balls that were thrown *should* have been picked. I'm not talking about the stat geeks theory that 3 isn't enough to have statistical significance. I don't find it confusing that somebody would write about it, I find it silly to break things down to such a minute level and then ignore other pathetic minutiae with the same point.

[CTM]

The whole purpose of the stat is to show how many balls that were thrown *should* have been picked. I'm not talking about the stat geeks theory that 3 isn't enough to have statistical significance. I don't find it confusing that somebody would write about it, I find it silly to break things down to such a minute level and then ignore other pathetic minutiae with the same point.

It's not pathetic minutiae when a single person is that far outside the norm. This article wouldn't have been written is Sanchez had 7 dropped INT's. Just like it was written last year when he was in the middle of the pack. You're two int's that were but shouldn't have been stat IS pathetic minutiae if most QB's have a similiar numbers.

[jason423]

Right, but go back to Bergen's point. How can I take a "dropped interception" seriously when they do not account for INTs that were thrown directly to a WR that then had the ball wrestled away? I happened on more than 10% of his ACTUAL INTs. Isn't that "statistically significant" to you numbers monkeys? I remember there was a stat about some turnover machine having x amount of picks off of the hands of WRs. That would have to be considered too. It is at least as significant as dropped INTs and either one without considering the other pretty much useless.

The interception lost by the receiver is just dumb luck. Manning of the Giants seemed to have a ton of those. They would bounce off the guys hands. Off the guys chest. I dont think Sanchez had a ton of them. He had the two freaky ones against GB (I think?) that never should have been called picks, but was anything else out of the norm? I dont think in his case that would be considered statistically significant. My guess is he was right around the norm. The dropped interception thing is almost totally in his control and is so far beyond the norm that you have to think similar play will lead to more picks the following year. My guess is if someone charted Brett Favre there would be something that shows alot of ups and down in the dropped picks/lots of picks numbers.

I had run something last year on all the rookie QBs, and specifically tracked Sanchez for what would have been the average development for someone who started around where he was as a rookie. The statline would have been:

58% 210 YPG 18 Tds 19 INTs 71-73 QBR

Clearly the main areas where he was off were interceptions and completion percentage. When you factor both in together I think you have to say that the dropped interceptions played a big role in his number of picks being so low to the norm. If he was completing 64% of his passes that would be a different story, but there is no way to deny that accuracy is an issue. Now I do think part of the issue is the Jets offensive scheme which I dont think really does a QB any favors when there is a small mistake made on the field, but he does need to improve further next year.

FWIW his overall improvement level did track very highly, but not in that elite level that almost every superstar tracked in. Since we are talking about Sanchez more than FOs opinions, here is what I wrote about Mark when comparing him to the players who improved on the high end and I think its a valid conclusion about his ceiling and what the Jets have to do to protect him.

Now the question becomes what is his ceiling. He did fall short of the high end projections, primarily because he continued to battle accuracy problems all season long. Had he hit even the 58% mark he probably would have tracked with the best. His QB rating jump from 63 to 75.3 represented a 19.5% jump which is excellent. That is above the 17% threshold and puts him on track with the multi time pro bowlers who started as true rookies in the NFL.

Before we all get too excited it should be noted that the 19.5% improvement is in the low end, where there are a handful of guys who ended up busting. The 19.5% improvement puts Sanchez in the middle of the Trent Edwards, Bernie Kosar, Troy Aikman, Rodney Peete, Billie Joe Tolliver grouping. Of those names you have one Hall of Famer, one solid starter and three busts. The next leap would have been in the plus 27% category which really limit's the amount of busts.

One area that sets Sanchez apart from some of the poor players is the fact that his YPG spiked by so much. Tolliver and Peete both regressed because the staffs took responsibility away to make the game safer with them in. His YPG increase is also higher than the other three names on the list. His decrease in interceptions per attempt was better than anyone else in the list and his yardage per game increase was right in the middle. The reasons he fell way short in the QBR is because of the lack of touchdowns and poor growth in completion percentage. The only QB to see such a poor completion percentage improvement was Rodney Peete, who fell from 52.8 to 52.3%. The next closest to Sanchez' 1.8% growth was Neil Lomax at 5.5% and Troy Aikman at 7.1%. There was no QB who fell harder in touchdown passes per attempt. While the TD is somewhat of a fluky stat, Sanchez' numbers fell a dramatic 22% over his rookie year. The next closest would be Aikman who fell around 10%. Kosar and Donovan McNabb were the only other players who were worse in this category in year two, but both were less than 1% worse.

I think two years into it, it is safe to say that Sanchez is going to have an excellent chance at being a solid starter in this league. I would say at this stage now he has reduced his bust chance to about 10% and at the very least will suit up for 1 Pro Bowl as a Jet. The ceiling is probably not as high on others on the list. Based on where he places and looking at his categories of improvement there is probably a 35 to 40% chance that he will be an elite level player for any consistent period if time. Most likely he is going to be an above average starter that can start in the NFL for at least a decade.

Based on his numbers you would like to say that the best case scenario for him is probably to build a team around him the way the Dallas Cowboys did Troy Aikman in the early 1990's. They surrounded Aikman with a great offensive line, a high end receiver, a bailout tight end, a very good running back, and a tremendous defense. Aikman ended up in the Hall of Fame, which he probably would not have done if he was asked to put the team on his back the way a Peyton Manning or John Elway was asked to do. There is nothing wrong with that and that is the approach the Jets should be taking from this point forward with Sanchez to give him the best chance of success.

[#27TheDominator]

It's not pathetic minutiae when a single person is that far outside the norm. This article wouldn't have been written is Sanchez had 7 dropped INT's. Just like it was written last year when he was in the middle of the pack. You're two int's that were but shouldn't have been stat IS pathetic minutiae if most QB's have a similiar numbers.

WHAT I AM TALKING ABOUT ISN'T A STAT! Football Outsiders is talking about it as a stat. My point is this - the reason dropped INTs matter is because they are a badly thrown ball, to the other team, that doesn't appear on the stat sheet. It's true that if other QBs had the same numbers it would be meaningless, but you are seperating two things that shouldn't necessarily be seperated.

Jason has a good analysis of Sanchez right here. I don't have anything against looking beyond the numbers (to even more numbers) I just prefer a more holistic approach. I'll concede that he could have had more INTs caught, though I would doubt the 15 figure. I don't think I could have missed that many balls bouncing off defenders chests. One thing I will say about Sanchez accuracy problems are that they don't always seem to lead to INTs. I think he plays tight and makes some truly pathetic throws trying to avoid INTs early in games. Those are the ones where he whiffs on dumps to the backs or can't hit a stationary Keller.

BTW, the fact that the INTs aaginst GB weren't in the control of Sanchez is precisely the point.

How about INTs on heaves at the end of halves or games with no time on the clock? Sure, eventually that stuff evens out, but you get statistically meaningless INTs. If I'm actually rating the player (or betting) I care about all that sh*t.

[CTM]

WHAT I AM TALKING ABOUT ISN'T A STAT! Football Outsiders is talking about it as a stat. My point is this - the reason dropped INTs matter is because they are a badly thrown ball, to the other team, that doesn't appear on the stat sheet. It's true that if other QBs had the same numbers it would be meaningless, but you are seperating two things that shouldn't necessarily be seperated.

Jason has a good analysis of Sanchez right here. I don't have anything against looking beyond the numbers (to even more numbers) I just prefer a more holistic approach. I'll concede that he could have had more INTs caught, though I would doubt the 15 figure. I don't think I could have missed that many balls bouncing off defenders chests. One thing I will say about Sanchez accuracy problems are that they don't always seem to lead to INTs. I think he plays tight and makes some truly pathetic throws trying to avoid INTs early in games. Those are the ones where he whiffs on dumps to the backs or can't hit a stationary Keller.

BTW, the fact that the INTs aaginst GB weren't in the control of Sanchez is precisely the point.

How about INTs on heaves at the end of halves or games with no time on the clock? Sure, eventually that stuff evens out, but you get statistically meaningless INTs. If I'm actually rating the player (or betting) I care about all that sh*t.

Dropped int's isn't meant to be a holistic stat. It's just a stat, like TD's, INT's, yards, sacks. All of which are equally vunerable to the same kind of what-if's you and others have posed here, what about garbage time yard's / td's, etc.. If you want a holistic stat, look at DVOR or DYAR those actually try and eliminate some of this noise... But people try and poke holes in them as well.

I really don't understand why you guys are holding a fringe stat to a standard you don't normally hold "official" stats to. It's just a number, a number that when put together with other numbers allows you to begin to paint a picture. The more numbers the better. This particular number simple says that Sanchez was very fortunate to have only 13 int's last year, and if he doesn't play better, luck most likely won't continue to be on his side and he'll be north of 20 next year. It makes no claim as to whether or not he'll play better, or doesn't attempt to grade him as a QB holistically, it simple provides SOME (not all) context around one of his stats, in this case INT's

WHAT I AM TALKING ABOUT ISN'T A STAT! Football Outsiders is talking about it as a stat.

on edit: All i'm taking about is the stat, a stat meant to shed light on a very specific thing, that when combined with other stats, lets you build a holistic view. It wasn't clear to me we weren't discussing the stat.

[#27TheDominator]

Dropped int's isn't meant to be a holistic stat. It's just a stat, like TD's, INT's, yards, sacks. All of which are equally vunerable to the same kind of what-if's you and others have posed here, what about garbage time yard's / td's, etc.. If you want a holistic stat, look at DVOR or DYAR those actually try and eliminate some of this noise... But people try and poke holes in them as well.

I really don't understand why you guys are holding a fringe stat to a standard you don't normally hold "official" stats to. It's just a number, a number that when put together with other numbers allows you to begin to paint a picture. The more numbers the better. This particular number simple says that Sanchez was very fortunate to have only 13 int's last year, and if he doesn't play better, luck most likely won't continue to be on his side and he'll be north of 20 next year. It makes no claim as to whether or not he'll play better, or doesn't attempt to grade him as a QB holistically, it simple provides SOME (not all) context around one of his stats, in this case INT's

on edit: All i'm taking about is the stat, a stat meant to shed light on a very specific thing, that when combined with other stats, lets you build a holistic view. It wasn't clear to me we weren't discussing the stat.

None of this is clear to me. I think I hate math!

The point where I was complaining about their specific stat method was that they were citing a percentage of passes that "should have been" INTs. At that point I think balls that are wrestled from WR or bounce of a TEs facemask into a CBs hands should be considered also. I agree there is something to take away from the idea of dropped INTs.