Mike Wallace being shopped by Miami Dolphins

The statistics don't account for it, but I think we have to ask ourselves whether that's a meaningful absence of data in terms of the issue being discussed. For example, if 29% of Ryan Tannehill's downfield passes to Mike Wallace in 2014 were deemed catchable, and only 25% of such passes by Ben Roethlisberger in 2012 were deemed catchable, does it matter by how much the other balls were "uncatchable"?

It may in fact matter if what we're doing is evaluating the two quarterbacks in general, in terms of how well they throw downfield overall, but if we're evaluating them with regard to how well they threw downfield to Mike Wallace when functioning in similar offenses, I believe the binary classification of "catchable versus uncatchable" works just fine, without any more precise evaluation.

In other words, I'm not sure the desire on your part to, as you said above, "to see not only a pass/fail tally but what actually happened to cause a play to succeed or fail" has any significance in this case. With other issues we might discuss, that might indeed be a meaningful absence of data, and the statistics may be woefully inept in answering the question we're addressing, but I'm not sure it is in this case.

 

I'm finding it more and more impossible to discuss the topic at hand with you. I'm sorry.

 

I actually think your stats would be more meaningful if we were talking about someone other than Mike Wallace.

I've said this before, but my recollection is that Tannehill had many more "egregious" underthrows on deep balls to Wallace than deep balls to other receivers, say Hartline. And by egregious, I mean the underthrows clearly allowed the DB to catch up, making the catch more difficult than need be. I really think it is something specific to Wallace with Tannehill. He doesn't seem to be able to estimate where Wallace will be after he makes his move, and so he either throws the ball with improper velocity or throws it too late (I think throwing it too late is more likely for deep balls).

Putting aside stats for a moment, do you and others agree with that observation?

If Tannehill has a deep ball problem that's specific to Wallace, then we need better input data here, because I don't remember such "egregious" underthrows by Roethlisberger on such a frequent basis for his deep balls, either to Wallace or to others (Roethlisberger seems to be far more content in just throwing it out there where the receiver can go get the ball).

In other words, it suggests the distribution of throws as a function of distance from the receiver actually might matter here.

 

why do I have a different take, is it just how we look at things or tolerate things, or put into context..

I look at it like this, Tannephins is the boards stat guy, if it goes with what I see then it's all good, win, win, if it doesn't I argue against the stats as he presents them, I don't get upset with him for having his vision of the game vrs mine and presenting his argument on behalf of the numbers, I don't see him telling us that we dont have a clue or couldn't possibly be right, I see folks being a jerk to him because he argues on behalf of the numbers and they get frustrated, then he will take his shot, just like any man would do..

When he presents numbers it makes me think about my own evaluation, if I look deeper and think that his numbers are not taking into account certain variables that may favor his argument, I tell him so..

His point that folks take personally is we are here on a board, we're not in the league..it's up to us to show that we can disprove the numbers with solid consistent evaluations where the record shows that our ratio of identifying talent is well above the curve..

We all need to find a balance and have some perspective..for example, CK posted some numbers about our offense lead by lazor relative to the years before, they were incredible, so here I am pissed at lazor a bit for not using Ryan enough in the way that best utilizes his skillset ( which is still something I stand by), but it was his first year and perspective is in order, that numbers checked me so I have to look at it differently.

Jim, I know you spend the time, I'm aware of your record on players, and I know Tannephins spends the time researching his side..it's up to all of us to fuse it all together without insulting one another..

I would say the same to Tannephins, some of the eyes on this board are very good, and you have to accept that your numbers may be deceiving you at times, now I know that hard for you to accept, but you simp,y cannot take numbers and run a football operation..if you did, I'd guarantee you that you would lose the battle of building a roster..

Now if you want to accept some sort of challenge over the long haul I'm in..I'm actually doing this with another member who's gonna be roboscout..

 

But just very simply speaking, what meaning do you ascribe to the fact that the percentage of catchable balls Tannehill threw to Wallace in 2014 was higher than the percentage of them Roethlisberger threw to him in 2012, when both QBs were in similar offenses?

In other words, why focus on the factors revolving around the uncatchable balls instead of the focusing on the percentage of catchable ones? Regardless of what happened on the uncatchable balls, Roethlisberger threw a lower percentage of catchable ones to Wallace than Tannehill did.

 

this is creepy Jim.

 

I think what you've done is equate for as much as you can with the available data. You're basically showing the effect (ideally) of all factors other than the QB's and Wallace on the QB-Wallace deep ball. So, you're saying the average effect of different coverages, play calls, etc.. on the QB-Wallace deep ball is X% catchable vs. uncatchable, as a binary variable. In that case, the fact it's a binary variable doesn't matter.

Things change when you try to infer something about the QB on the QB-Wallace connection. If one can point out what seems to be an important difference one hasn't (for the moment, one can't) corrected for with the stats, then the claims are less justified.


OK, so what about the catchable balls thing? You know, the very first thing we really need here, and maybe you can find this out, is to see what "definition" they are using of "catchable". Is it what the refs use when calling (or not calling) a penalty? I doubt it because that radius is HUGE. Is it what analysts describe as "catchable" after seeing say the combine results? I also doubt that because most of the time they talk about "catchable" as a ball only the receiver can catch, and often can keep running with afterwards. So the first thing is to try and find out what "catchable" really means in this context.

Personally, I think they are using "catchable" as the ball being within some radius where the coverage may be decent but not so good that it's really impossible for the WR to catch the ball (just a guess here..). If something that like is called catchable, then take the explanation I gave about Wallace having to backtrack for Tannehill's balls. Many of those balls are still catchable, but far more difficult to catch than they should be. This is something that a more sophisticated analysis should penalize the QB for (putting the WR in a position where the coverage got WORSE because of the throw). I doubt that happens as often with Roethlisberger.

Once again, that's an educated guess that depends on what "catchable" really means here.

 

I still don't know why anyone uses stats to judge individual players in football. It is absolutely the worst sport to use stats other than for team performance. This isnt baseball where the game revolves around a one on one matchup primarily throughout the game and therefore stats are revealing. This is a sport where 22 players interact on every play and most have some impact on whether the play succeeds or not. if you give players only a simplified grade of positive, negative, neutral on every play you are left with over 10,000 possibilities of who impacted the play. if you give them a more realistic 5 potential grade (great, good, neutral, poor, wtf) you are left with over 5 million possibilities.there arent any stats or algorithms out there that account for any of this. you're just finding meaning in nonsense

 

Assuming the definition of "catchable" is being reliably applied across the two QBs (Tannehill and Roethlisberger) when throwing to Wallace, why does it matter how it's defined?

Also consider that from 2009 to 2011, the percentage of catchable downfield balls Roethlisberger threw to Wallace was much higher (50%, compared to the 25% in 2012), and in 2012 the percentage of catchable downfield balls Tannehill threw to Brian Hartline (48%) was much higher than the percentage of catchable downfield balls he threw to Mike Wallace in 2014 (29%).

Essentially what you have here is all of the evidence's pointing to how, when Mike Wallace is in a timing-oriented pass offense, he "disables" his quarterback in terms of the quarterback's typical downfield performance.

And again, all of this is consistent with how Mike Wallace was viewed coming out of college, when he was a mere 3rd-round pick despite running an ultra-fast 4.33 40 at the combine. He just wasn't viewed as having the traits of a successful wide receiver at that time, and I don't believe he displays those traits now, either, unless the precise mix of ingredients are present around him in terms of the quarterback and the offensive scheme.

 

Using that logic, we would consider Dan Marino's 420 career touchdowns to be meaningless in terms of his individual greatness, since we'd be using a statistic to judge an individual player in football.

 

Yeah, that's what I was trying to point out. The distributions within both the "catchable" and "uncatchable" conditions could be different (plausibly so.. e.g. having to backtrack for a catchable ball, and a higher percent of highly uncatchable balls), and that a more sophisticated analysis would likely lower Tannehill's overall rating because of this.

I actually think your stats are more meaningful for this argument (how Wallace influences deep ball productivity in different situations) than when you used these stats to argue against Jim1's "Tannehill throws a lousy deep ball" argument. That's when I thought the stats don't really make your case.

 

i do. watching him throw the ball was what made him great. when you watched him you knew you would probably never see another one like him in your lifetime. that's what made it so special

 

And that is in no way mutually exclusive with a view that his 420 touchdown passes had meaning in terms of his individual greatness. In other words, one can easily say Dan Marino was great based on what one saw when watching him play, and Dan Marino was great because of his 420 touchdown passes.

You don't have to believe just one or the other of those. You can believe both.

Deeming Marino's 420 touchdown passes to be meaningless, however, in my opinion suggests some pretty shoddy thinking in the effort of portraying statistics to be meaningless. At that point, in the effort to portray statistics as meaningless, one has reached a "the sun revolves around the earth" position in my opinion. To each his own, however, as always.

 

sure but it doesnt really mean much. stats are just numbers. how many of those touchdowns can you attribute to dwight stephenson's blocking or don shula recognizing the talent and letting marino wing it. how many more could marino have had if he had duper and clayton type receivers his whole career. how many more tds would he have had if he played in today's game with its flag football rules. is a 90 yd pass worth less if the receiver gets tackled on the 1 yd line vs a 5 yd td pass to a TE

 

When you consider that the "rare air" Marino is in with his 420 TDs is nowhere near fully explained by the Dolphins' record during that period, which was nowhere near as different from the average team's, it isn't difficult to arrive at the conclusion that his individual ability, rather than his surroundings, was primarily responsible for his performance in that regard.

 

disagree. alot of factors lead to it. if dwight stephenson is replaced by the 25th ranked center for all those years, i guarantee the numbers would have gone down. if they had a running back the numbers would have gone down but there would be a ring on his finger. he could have thrown for 100 tds and he still would have been the best qb i ever saw play. you know the best qb of the 60s was? greg cook. but you'll never know that because injuries derailed his career after only his third start. and he basically only has a year of playing to go by but that one year he was winging it like marino. stats dont tell the story in football. in baseball yes. somewhat in basketball. more difficult in hockey and completely innefectual in football unless you're using stats to judge the entire offense as a collective but thats a different story altogether. when you have between 10,000 and 5,000,000 possible interactions on every play you would need an eisntein level math genius combined with a vince lombardi football expert to come up with algorithims that could be useful for determining individual worth isolated from the other 21 players

 

You've taken a position in which the statistics customarily attributed to individual ability are meaningless for you. Aaron Rodgers's highest QB rating in league history, as an indicator of his individual ability? Meaningless. Emmitt Smith's highest number of rushing yards in league history, as an indicator of his individual ability? Meaningless. Brett Favre's (and soon Peyton Manning's) highest number of passing yards in league history, as an indicator of his individual ability? Meaningless.

 

yup and i think emmitt smith is a good example. also terry bradshaw's four rings. smith really wasnt that great a back, he was good but there were so many better than him, but he played behind one of the all time great offensive lines. bradshaw was really horrible at times but still ended up with four rings. there was a time when bradshaw was considered an all time great but thankfully the last few decades have seen his reputation slip considerably

 

Well, there's a difference between that statement (bolded above) and saying Smith's highest number of rushing yards in league history are meaningless as an indicator of his individual ability. Certainly you must see some indication of Smith's ability in that statistic, or do you believe just anybody with any level of ability can amass that number of yards, simply because he has one of the all-time great offensive lines?

 

well i'm saying that emmitt was good but he isnt in the conversation for all time greats. its a great example of how the stats would lead you to believe he is an all time great but an eyeball test will tell you he was good but not great.

 

We'll have to agree to disagree that Emmitt Smith isn't an all-time great. Certainly the Hall of Fame induction panel disagrees with you, as well.

 

What makes the case in that instance is that Tannehill wasn't any less accurate downfield than the average QB in 2014, and in 2012, when he wasn't handicapped by Mike Wallace downfield -- and as a rookie no less -- he was actually somewhat above the league average in that area. Those are pretty simple statistics.

 

Right, which is logically consistent with Tannehill being the problem, Wallace being the problem, or both (when it comes to the Tannehill-to-Wallace deep ball). You acted like the stats (on their own) argued against Tannehill being the problem when they don't. Again, you need a richer dataset to do the analysis on.

 

The point is simply that Tannehill doesn't throw a lousy deep ball.

 

Your data doesn't prove that when he throws to Wallace, and that was the topic of discussion (not his deep balls to all other receivers).

 

Define "lousy".

 

Below average.

 

This is that particular part of the thread:

So in that part of the thread, the topic of discussion was Tannehill's deep ball ability in general, not in throwing to Wallace exclusively.

 

One would think you'd have to be out of the average range, at least, in the negative direction to qualify for the moniker "lousy."

 

I only debated Tannehill's deep balls to Wallace with you. If you exclude him, your stats are more meaningful.

 

I was referring to this statement on your part:

Again, this is what did that:

 

I entered the debate from post 165, which is a response to a post that talks about Wallace. And ever since then, I've stuck to the Tannehill-to-Wallace deep balls or just stats in general.

 

There is no denying that Tannehill's balls to Wallace have been not great. There is also no denying that the anti-Tannehill crowd has largely used his struggles to connect with Wallace as a reason to run around screaming "Tannehill throws a terrible deep ball!!!" The argument by us labeled "Tannehill lovers", or other fantastic monikers, has been quite consistently that he throws a bad ball right now to Wallace, not a bad deep ball in general. The question is, WHY is his deep ball bad to Wallace? I don't think there is a definitive answer, but I think that Wallace's shortcomings (route running, small catch radius, timing based offense) combined with Tannehills (no confidence in Wallace, too much pressure to hit wide open Wallace, poor blocking, etc) have created a terrible situation in regards to Tannehill to Wallace.

 

I think it's more a combination of exactly what you said about Wallace, with the fact that Tannehill is expected to engineer a timing-oriented pass offense, just like Ben Roethlisberger was in 2012 when he had even more difficulty than Tannehill throwing catchable downfield passes to Mike Wallace.

If you told Ryan Tannehill to forget about the timing pass offense he's supposed run, and told him to instead move around in the backfield and evade pass rushers until he saw Mike Wallace break free of coverage with his 4.33 speed, I don't think you'd see any problem with him connecting with Wallace downfield at a much greater rate.

Likewise, if you had Ryan Tannehill engineer a timing pass offense and paired him with a reciever with different traits -- and even one as mediocre perhaps as Brian Hartline -- then there would be no problem hitting that recevier downfield, either, just as there wasn't in 2012, when Tannehill was a rookie no less.

This is all about whether the downfield receiver has the appropriate traits to function within a timing pass offense.

 

Toss on top that most of the downfield passes to Wallace are ones that come directly over the top of him, on basically fly routes, instead of getting him on deep crossing patterns or corners.

 

Right, but that fits with the timing offense issue, as well. When he broke free of coverage while Roethlisberger was evading pressure in the non-timing offense they were both in from 2009 to 2011, he probably broke diagonally toward the middle of the field, which in turn supplied the highlight reel catches we all saw and assumed were deep posts or corner patterns.

Those highlights are possible in that sort of offense. In a timing pass offense, however, they're going to be the exception to the rule, unless the receiver achieves separation with route-running and shows himself to be open to the QB early in the route. Wallace doesn't have that skill, however.

So it's entirely possible that those sorts of routes were run in Miami just as much as they were in Pittsburgh, but because Wallace didn't appear open early during them, we never even saw many throws to him on those routes, let alone catches. Consequently we believe there were a meager number of them.

This is again what I mean about what the "film" doesn't show, and why a routine watching of games on TV can provide just enough information to "confirm" what you already believe.

 

part of the criticism of ryan is that he is not anticipating wallace beating his man early, which Wallace can do, and if thrown correctly and at the right time, wallace will continue to beat the man and run to the throw, seperating in the process..this is probably where the dissent comes into play between qb and receiver...your not going to find stats for this particular aspect of the game.