Grantland: Arizona has a Numbers Problem

[WildBB]

I'd like to see how they out perform his numbers crunching with Palmer healthy a whole season. Then throw them all out the window, imo. It's mostly about QB play and your D. Turnovers, third down efficiency and TOP are huge factors.

 

[kerouac9]

I'd like to see how they out perform his numbers crunching with Palmer healthy a whole season. Then throw them all out the window, imo. It's mostly about QB play and your D. Turnovers, third down efficiency and TOP are huge factors.

Part of Barnwell's argument is that Palmer's torrid pace to begin the season is/was probably unsustainable. His 1.3 interception percentage is less than half his career average (including his time here). Even Drew STanton's 2.1 INT% was less than Palmer's career average.

There's very little chance that Palmer would've been able to keep up the pace he established at the beginning of the season; and not much chance he would have outperformed Stanton's interception rate, anyway.

Yes, if you believe that Palmer would've sustained his pace and thrown 29 TDs against 8 INTs, the Cards would've been better. But that wasn't going to happen.

 

[Arizona's Finest]

I really like Barnwell but he is almost exclusively numbers/analytics based. Every supposition he has is probablity and % based which on a big enough sample size across multiple years - yes you can absolutely draw inferences. But there is too much nuance that comes into play and there is error. Its wjhy all these teams who are analytics heavy dont win all championships every time. Probabilities suggest likelihood not absolute. ANd many times its a fraction of a %. Which again over a long enough time line means you win more then you lose. But its nothing certain. His partner in crime Robert Mays is the opposite and all tape/eye test based.

There usually is a happy medium. And yeah hes right we may not be as "lucky" this year. Well just be good. I believe Palmer stays healthy and we win at a 75% clip if that happens. Our defense may regress a bit (but i dont think much especially if Washington is back at any point) but the offense should make a nice jump just through better health and additions. There is literally no reason you would see the offense regressing as we kept everyone, added to it, and then its most guys 3rd year in system. that is HUGE.

 

[Arizona's Finest]

Part of Barnwell's argument is that Palmer's torrid pace to begin the season is/was probably unsustainable. His 1.3 interception percentage is less than half his career average (including his time here). Even Drew STanton's 2.1 INT% was less than Palmer's career average.

There's very little chance that Palmer would've been able to keep up the pace he established at the beginning of the season; and not much chance he would have outperformed Stanton's interception rate, anyway.

Yes, if you believe that Palmer would've sustained his pace and thrown 29 TDs against 8 INTs, the Cards would've been better. But that wasn't going to happen.

I unequivicaly disagree with the importance of that ratio and what it means towards winning. Ok lets say it goes up. What is the argument that it would go back to his career average or above? He has as good or better weapons, the best oline, the best offensive playcaller hes had, plays in a dome, best defense, and most importantly 3 years in the system.

Lets say it goes up close to 100% but no near his career average. Thats 2.6%. Oh man thats terrible.

Oh wait

2005 32 TD and 12 INTS 2.4%
2006 28 TD and 13 INTS 2.5%

Umm pretty sure we would take both of those years and be contenders if he put up those lines. So yeah maybe 1.3 isnt sustainable (although things seem pretty perfect in terms of everything set up for best year of career) but so what? We can still win and win big. And my guess is hes closer to the 1.9 or 2.0 area which would be phenomenol.

 

[MadCardDisease]

There is only one stat that matters in the NFL.

Wins vs Losses

 

[NeverSayDieFan]

GOOD teams find ways to WIN....

I've been a sports fan for 48+ years now and that's always been the law of the land. Case in point, my Cardinals in baseball. Do they have decent talent? YES! But, the MOST talented team? NO! They do, however, have exceptional strength of will and determination. They simply refuse to give up on a game and they play one game at a time. NO looking down the road. MAX effort now...NOT the next day.

In the end it's NOT how much you win by...it's simply whether you win or lose.

So, get ready for another nail-biter season...and WINS!!!!

Mark

 

[kerouac9]

I unequivicaly disagree with the importance of that ratio and what it means towards winning. Ok lets say it goes up. What is the argument that it would go back to his career average or above? He has as good or better weapons, the best oline, the best offensive playcaller hes had, plays in a dome, best defense, and most importantly 3 years in the system.

Lets say it goes up close to 100% but no near his career average. Thats 2.6%. Oh man thats terrible.

Oh wait

2005 32 TD and 12 INTS 2.4%
2006 28 TD and 13 INTS 2.5%

Umm pretty sure we would take both of those years and be contenders if he put up those lines. So yeah maybe 1.3 isnt sustainable (although things seem pretty perfect in terms of everything set up for best year of career) but so what? We can still win and win big. And my guess is hes closer to the 1.9 or 2.0 area which would be phenomenol.

So he's going to go back to 8 years ago and perform at that level again? It's possible, but is it likely? He's going to repeat the best two years of his career a decade later?

 

[Arizona's Finest]

So he's going to go back to 8 years ago and perform at that level again? It's possible, but is it likely? He's going to repeat the best two years of his career a decade later?

I mean why not? Hes in what can be considered a QB's prime (30-35). He has as good or better weapons. He actually has an offensive coach. He plays half his games ina dome and almost everywhere else in warm weather. His defense is better then its been and gets more turnovers. So yeah i do. This is the best circumstances he has ever played in. I dont think its unreasonable that he plays much closer to 2005 level then his years with the Raiders or his average because of it? Do you?

You say 8 years ago like thats a long time. But if you look season to season this is his best chance to replicate 2005/2006 or get close based on everything i listed above. In fact I would make the argument he could/should surpass it.

But even if not? He has shown the capability to be at that level. His getting older isnt going to be that dramatic a difference as A) hes a QB and B) you take health out of equation which is what I am assuming.

ALso once you have seen the superstar ceiling its much easier to get back then to try and achieve when you are 34+. See Warner, Kurt.

Well see. I expect a great year and a very good team. But its the NFL who knows? SEA is a Wilson injury from being 8-8. Dont they have every right to expect to be great too?

 

[Mr. Boldin]

I dont see what the problem is with the article. The Pythagorean expectation is a key tool in every sport and does a great job of measuring impact via point differential.

Im a basketball analytics guy, but this article makes a ton of sense.

Anyone expecting Palmer to throw for nearly 30 and 9 last year or this year is thinking too far outside the box.

What I would propose is that, like Barnwell suggests, Arians is an elite coach who is capable of teaching his systems to players and game planning with the best. If we outperform expectations again this year, then there is an outlier in the analytics that has more to do with Arians and Keim than number can quantify.

Keep in mind, that can still be the case even if we win 9 games. I think we have great synergy between GM and coach and that has proven to be as important as any organisational/player role in professional sports to sustain success.

 

[TJ]

The Cardinals never played the Seahawks with Palmer is 2014.

They lost those games a combined 9-54 or -45

That is the vast majority of the point differential issues.

In 2013 with Palmer they split the series and were 39-44 or -5

It's a dumb article to be honest

Agree. Junk stats, TBH. We also got manhandled by Denver when Stanton was the QB and when he suffered a concussion. We also lost a game against Atlanta we had no business losing.

 

[juza76]

they need to go back to school,maths is not an opinion

 

[Russ Smith]

So he's going to go back to 8 years ago and perform at that level again? It's possible, but is it likely? He's going to repeat the best two years of his career a decade later?

HE might not this year but he did just that for the 6 games he played and it's not like was playing out of his mind good.

He certainly wouldn't be the first veteran QB to learn the value of protecting the ball as got later into his career, or the first INT prone QB who improved in that area because he was on a better team and realized it was ok to throw the ball away or take a sack instead of always trying to make a big play.

I have no idea how his season turned out if he played the whole year, but it's not like he was Nick Foles outlier stats good for those 6 games last year.

 

[Russ Smith]

Agree. Junk stats, TBH. We also got manhandled by Denver when Stanton was the QB and when he suffered a concussion. We also lost a game against Atlanta we had no business losing.

We actually were in the Denver game until the injuries happened. Hell we were in the game after the fluke TD by Thomas until the wheels predictably fell off.

 

[WildBB]

2005 32 TD and 12 INTS 2.4%
2006 28 TD and 13 INTS 2.5%

Umm pretty sure we would take both of those years and be contenders if he put up those lines. So yeah maybe 1.3 isnt sustainable (although things seem pretty perfect in terms of everything set up for best year of career) but so what? We can still win and win big. And my guess is hes closer to the 1.9 or 2.0 area which would be phenomenol.

I think with those numbers we're pretty guaranteed to be at least a 10-12 win season Team depending if the D raises up.

 

[MaoTosiFanClub]

We have top 10 talent and good (when healthy of course) quarterback play. That's all it takes in the NFL, we'll be fine.

 

[Hollywood]

Analytics are stupid. His main argument is that we went 4-1 in games decided by 7 points or less. We have already matched that this season.

This was sent from my cell pjone so please excise any typeos or other errors.

 

[Cbus cardsfan]

Part of Barnwell's argument is that Palmer's torrid pace to begin the season is/was probably unsustainable. His 1.3 interception percentage is less than half his career average (including his time here). Even Drew STanton's 2.1 INT% was less than Palmer's career average.

There's very little chance that Palmer would've been able to keep up the pace he established at the beginning of the season; and not much chance he would have outperformed Stanton's interception rate, anyway.

Yes, if you believe that Palmer would've sustained his pace and thrown 29 TDs against 8 INTs, the Cards would've been better. But that wasn't going to happen.

I think there's a good chance it happens. I'd even go as far to say 31 TD's against 9 Ints by week 13.

 

[crisper57]

I think there's a good chance it happens. I'd even go as far to say 31 TD's against 9 Ints by week 13.

Classic bump. Wow! He is on another level in this offense.

 

[Chaz]

I unequivicaly disagree with the importance of that ratio and what it means towards winning. Ok lets say it goes up. What is the argument that it would go back to his career average or above? He has as good or better weapons, the best oline, the best offensive playcaller hes had, plays in a dome, best defense, and most importantly 3 years in the system.

Lets say it goes up close to 100% but no near his career average. Thats 2.6%. Oh man thats terrible.

Oh wait

2005 32 TD and 12 INTS 2.4%
2006 28 TD and 13 INTS 2.5%

Umm pretty sure we would take both of those years and be contenders if he put up those lines. So yeah maybe 1.3 isnt sustainable (although things seem pretty perfect in terms of everything set up for best year of career) but so what? We can still win and win big. And my guess is hes closer to the 1.9 or 2.0 area which would be phenomenol.

Good call.
Through week 13: 9 INT 453 ATT = 2%

 

[TJ]

I unequivicaly disagree with the importance of that ratio and what it means towards winning. Ok lets say it goes up. What is the argument that it would go back to his career average or above? He has as good or better weapons, the best oline, the best offensive playcaller hes had, plays in a dome, best defense, and most importantly 3 years in the system.

Lets say it goes up close to 100% but no near his career average. Thats 2.6%. Oh man thats terrible.

Oh wait

2005 32 TD and 12 INTS 2.4%
2006 28 TD and 13 INTS 2.5%

Umm pretty sure we would take both of those years and be contenders if he put up those lines. So yeah maybe 1.3 isnt sustainable (although things seem pretty perfect in terms of everything set up for best year of career) but so what? We can still win and win big. And my guess is hes closer to the 1.9 or 2.0 area which would be phenomenol.

2015 (13 games in)--31 TDs 9 INT 2.0%

AZ Finest = prophet

 

[CardsFan88]

Good thing analytics isn't the universe, describes the past, and can't predict the future.

The old adage is, that's why you play the games.

A tool yes. Solid predictor, no. Just ask Wall Street about 2008.

 

[football karma]

I didn't have a big problem with the analysis in Aug: it implictly assumed that Palmer wouldn't play a full season (or even a majority), or, if he did, his numbers would be closer to Stantons than what he did in the six games he did play. If those things they assumed happened, 8-8 would be expected.

Instead -- quite the opposite. Not only has Palmer come back healthy, he took the time to refine his mechanics and footwork. His third year in the offense has elevated his game. Stuff like that is hard to model mathematically.

 

[Bert]

I think there's a good chance it happens. I'd even go as far to say 31 TD's against 9 Ints by week 13.

Love it!!!!!

I would love to run some analytics on some members prediction success rates.


Sent from my iPad using Tapatalk

 

[CardsFan88]

I didn't have a big problem with the analysis in Aug: it implictly assumed that Palmer wouldn't play a full season (or even a majority), or, if he did, his numbers would be closer to Stantons than what he did in the six games he did play. If those things they assumed happened, 8-8 would be expected.

Instead -- quite the opposite. Not only has Palmer come back healthy, he took the time to refine his mechanics and footwork. His third year in the offense has elevated his game. Stuff like that is hard to model mathematically.

The inherent problem with models are, they don't represent reality, especially not in the present. They are incomplete guesses based on the past since models don't include all the variables of the universe. The past is not prologue.

The second problem is given that so much was up in the air, and Palmer has always had some great tools and did indeed play great last year, it was kind of funny they would even decide to go down this road an attempt one and assume so much.

One problem compounds with another and leads to being completely and utterly wrong. But they decided to go there. Human error coupled with mathematical error can lead to just about the worst analysis and decisions possible.

It's a tool to see if you missed something out of left field, not a predictor. Sadly, most use it as a shortcut to prediction so they don't have to think. It can take the human brain/creativity element out of decision making, and provides a false sense of confidence which in of itself introduces new opportunities for errors and expands the range of error.

Finally, when wrong, instead of realizing the futility of it to predict, they generally just try to re-tinker with the model, and assume it will be right, until proven wrong once again... rinse and repeat. It's hard to learn from your mistakes when you think the mistake was merely you got a solvable problem slightly wrong, instead of realizing the methodology can't solve it to begin with.

 

[itlnsunsfan]

The inherent problem with models are, they don't represent reality, especially not in the present. They are incomplete guesses based on the past since models don't include all the variables of the universe. The past is not prologue.

As my engineering professor liked to say, "All models are wrong. Some are useful."