Okay, what are your theories for difference?

[mwald]

Then along came a real man, picked up the deer, tossed it in the back of his pick-up and said "Thanks."
The three statisticians were left to diddle with their digits.

Easily the best post in this thread, including mine.

Real men with the courage to step outside the ivory tower. I said as much, but this says it better.

[BernardB]

Went back and read the Blanda and sacks thread and, from my perspective, Matt has taken up residence at an extremist position on stats.

Yet, in the sacks thread, I didn't see anyone take a valid position against what I was saying (some eventually and reluctantly agreed) and more importantly I didn't see any great or even merely good counter arguments to what I was saying.

{Hubris just leaks out of these sentences. From my perspective, Webster had the far better of it.}


Not sure about extremist position, the most recent CC article I wrote contains stats.

Regardless, my position isn't extremist, it's consistent and born out of how I grew up in football.

{I too see my position as middle of the road and not as an extremist (I am not a metrics type). But, I suspect, one is hardly ever in position to see oneself in an objective light.}



Posts are getting way too long in this thread, so here's an overly simplified version of one of many opinions I have on stats - in general, not across he board.

What stats are and do is insufficiency measure nothing more than the end result. Missing out on the very essence of football, what happened on the field. Further, a vast majority of stats leave out a vast majority of the players ON the field (first ex., stats for OL?) ...

Example -
Stats: QB threw int. CB intercepted pass. 20 other players weren't even on the field, according to the stats.
Reality: Numerous and practically infinite possibilities as to how the play, played out. Then just as many possibilities as for the why.

I inserted two comments above in brackets. Your last two paragraphs may well explain part of your frustration with stats. It sounds like your major complaint about stats is that they are not a 1 to 1 model of reality. You are correct, stats do not and cannot do this. The only 1 to 1 model of reality is reality itself.

There are obviously all kinds of stats which have sundry purposes. One potential purpose is to spot and correct misperceptions about reality which creep in due to our various biases.

If what you want is a 1 to 1 model, I understand your frustration with stats. But you are complaining against a false idea of what stats are attempting to achieve.

[Reaser]

I understand your frustration with stats.

I doubt it, but a large part is with how people use stats: without context, out of context, and putting stock in stats that very basic football knowledge says have no stock.

It's also not frustration, it's more wanting the collective football knowledge to improve (including my own) and/or not wanting to let the errant go unchecked so as others don't repeat flawed conclusions based off of flawed statistics. That's just me though, I like to learn from others and I like to help others that may not know exactly what they're talking about: "Bobby Layne sucks, look how many ints he threw and he didn't even complete half his passes" says the stats experts (and that was essentially posted on the old PFRA forum by your genius 'stats guys' - I'm sure others remember, especially since it actually happened twice in a couple year span.) People that know football a bit more than just looking up stats? They know better than that. That's the difference between those that depend on stats to tell the story and those that don't.

Keep fighting the good fight, though. You're very convincing, wait, no one has agreed with anything you've said, though I'm sure your excellent perception skills say otherwise.

[BernardB]

I understand your frustration with stats.

I doubt it, but a large part is with how people use stats: without context, out of context, and putting stock in stats that very basic football knowledge says have no stock.

It's also not frustration, it's more wanting the collective football knowledge to improve (including my own) and/or not wanting to let the errant go unchecked so as others don't repeat flawed conclusions based off of flawed statistics. That's just me though, I like to learn from others and I like to help others that may not know exactly what they're talking about: "Bobby Layne sucks, look how many ints he threw and he didn't even complete half his passes" says the stats experts (and that was essentially posted on the old PFRA forum by your genius 'stats guys' - I'm sure others remember, especially since it actually happened twice in a couple year span.) People that know football a bit more than just looking up stats? They know better than that. That's the difference between those that depend on stats to tell the story and those that don't.

Keep fighting the good fight, though. You're very convincing, wait, no one has agreed with anything you've said, though I'm sure your excellent perception skills say otherwise.

I want to celebrate what the likes of Bob Carroll, Pete Palmer, and John Thorn accomplished, but you are all twisted up ranting about what some bad stat types did. You are what could be called a lumper, you want to lump the good and the bad all together.

Well at least you did not regale me with your HS quarterback career.

[Reaser]

I want to celebrate what the likes of Bob Carroll, Pete Palmer, and John Thorn accomplished, but you are all twisted up ranting about what some bad stat types did. You are what could be called a lumper, you want to lump the good and the bad all together.

"The Hidden Game of Football" is literally sitting on my desk and I've listed it as - not my favorite but up there as - one of my favorite books on more than one occasion on PFRA forums.

"What makes a team win? Well, obviously, scoring more points than its opponents, and because football is divided into offense and defense, there are two ways to accomplish this: either score more points than the enemy or keep the enemy from scoring as many points as you do."

Lumper, you're just grasping at things now because no one buys what you're selling (see: other responses to your posts in this thread) but no, many people here know me and there's plenty of discussion on stats. I believe I even mentioned "winning stats" in this very thread (though in a different context) which you whined about that, also. You are what could be called a whiner, you want to whine about references to winning stats and the losing all together.

[BernardB]

I want to celebrate what the likes of Bob Carroll, Pete Palmer, and John Thorn accomplished, but you are all twisted up ranting about what some bad stat types did. You are what could be called a lumper, you want to lump the good and the bad all together.

"The Hidden Game of Football" is literally sitting on my desk and I've listed it as - not my favorite but up there as - one of my favorite books on more than one occasion on PFRA forums.

"What makes a team win? Well, obviously, scoring more points than its opponents, and because football is divided into offense and defense, there are two ways to accomplish this: either score more points than the enemy or keep the enemy from scoring as many points as you do."

Lumper, you're just grasping at things now because no one buys what you're selling (see: other responses to your posts in this thread) but no, many people here know me and there's plenty of discussion on stats. I believe I even mentioned "winning stats" in this very thread (though in a different context) which you whined about that, also. You are what could be called a whiner, you want to whine about references to winning stats and the losing all together.

Guess we put this thread rest, giving spite for spite.

[oldecapecod11]

by BernardB » Sat Jan 23, 2016 11:37 pm

"Guess we put this thread rest, giving spite for spite."

There was no spite on Reaser's part.
As you were going down to defeat, you grasped at straws and tried to invoke names you likely envy.
As children, there was Hans Brinker; or, the Silver Skates...
As adults, we have Bernard Brinker - the tarnished tongue.
Go in peace; you've already been torn to pieces.

[BernardB]

by BernardB » Sat Jan 23, 2016 11:37 pm

"Guess we put this thread rest, giving spite for spite."

There was no spite on Reaser's part.
As you were going down to defeat, you grasped at straws and tried to invoke names you likely envy.
As children, there was Hans Brinker; or, the Silver Skates...
As adults, we have Bernard Brinker - the tarnished tongue.
Go in peace; you've already been torn to pieces.

What you just wrote is the very definition of spite, although you will want to call it truth telling.

[bachslunch]

Oh what the hey. My FWIW on stats.

To a fair extent I agree with Reaser's concern about how stats are used. Context matters, and there are no shortage of people who blindly cite them. Many defensive stats in particular have context problems like sacks, interceptions, and tackles.

Contrary to what one might think given that I routinely cite postseason honors in HoF cases, I'm fine with using film study here. My problem is knowing who to trust in film breakdown given how differently people seem to interpret film, and having a comprehensive enough sample and some kind of reasonable ranking for comparison. I really like and appreciate what Ken Crippen's site offers in this regard and only wish there were more players on it -- though no question making it comprehensive would be a beyond-Herculean task. It has definitely changed how I see the HoF worth of some players (more positively re Winston Hill, Jim Ray Smith, Bob Skoronski, and Alex Karras, more negatively re Dick Schafrath and Dick Stanfel). It confirms my suspicions about what the low postseason honors tell us about guys like Pat Fischer and Mike Curtis. And it also puts Jerry Kramer's snub in better context; while I'd still likely put him in, it's clear his pass protection deficiencies provide some level of argument for keeping him out.

That said, I'm all in favor of attempts to apply sabermetric thinking to football when one can. Baseball lends itself better to this approach and is better developed in that sport, but there's no reason not to try in football. There are some useful baby steps here already. And even baseball sabermetric thinking continues to evolve, after all.

As to using sabermetrics to predict future performance, it works to some extent in baseball but there are misses also. In that case you're trying to figure out whether a player is a good signing risk or not, as well as how your team generally will do. And it's useful, but not infallible, and may remain that way indefinitely. But it's remarkably useful for quantifying the past regarding things like HoF cases and general trends, both very helpful things.

Fire away if you must.

[bachslunch]

That said, I respect Reaser's posts and knowledge of pro football. But not because he played high school or college football like millions of others.

Ditto.

This is a little similar to another forum where NFL history gets discussed from time to time in random threads. Years ago, somebody tried to rip me for even daring to have an opinion about some NFL matter that occurred before each of us were born. The other person contended I should not even have an opinion because the event happened before I was born. My response was something along the lines of, "Well, I guess nobody alive today should be able to discuss dinosaurs."

Yup. This is akin to folks who say unless you're in the movie industry you have no business writing film criticism, or unless you play the clarinet you have no business writing criticism about a clarinet recital. Sorry, but it's just not so, and in fact such observations can become a refuge for scoundrels (often by the person getting criticized).