for display only
Big Blue Interactive The Corner Forum  
Back to the Corner

Archived Thread

It is looking more and more like JJ is too old school

NoGainDayne : 10/11/2021 12:57 pm
this has been a concern of mine despite him being sharp and having good leadership skills.

4th and 5 from the 37 to open the game. Jones looking like he could gash the defense for runs if needed. Offense on a roll from last week. Kicking a 54 yard field goal. That wasn't a good "feel" for the game. You have to know Dallas is going to march it up and down the field on your team. Your only shot is a confident offense with it's foot on the gas for 60 mins. That "feel" for the game BS is garbage. There was no feel. JJ played it too conservatively which is not a rarity for him and we watched another season collapse as a result.

Bad call. Bad analytics. Bad "feel." Even Mike McCarthy has this shit figured out. You see more and more teams converting 4th downs in their OWN territory and we can't even do it in the situations teams have understood for years.

The idea that the Giants "get" numbers when the numbers so often add up against them is ridiculous.

Sadly this case is becoming more and more clear that JJ wasn't brought in to change things but convince us that we were in good hands so Jints central could stay stuck in the past.
Pages: 1 2 3 <<Prev | Show All |
RE: RE: RE: Well I was making the case of given where the team was  
giantstock : 10/11/2021 7:45 pm : link
In comment 15409443 NoGainDayne said:
Quote:
In comment 15409405 giantstock said:


Quote:


In comment 15409197 NoGainDayne said:




I actually completely agree with this. And have been saying for years the Giants need to staff up their tech team to build a model of this level of sophistication. But they don't seem to have anyone with remotely sufficient technical chops to do this and they continue to get more basic things like timeouts wrong which leads me to believe if they have models producing results for the basic stuff they aren't good ones.

Not only does the big math support this as I pointed out. I'd say there are a variety of specific factors that would point to this being a good idea.

The Dallas defense gives up the 4th most yards per play at 6.3 which lends well to a conversion probability for 5 yards.

The Giants offense actually ranks 9th in yards per play at 6.1 lending even further to that conversion over the base.

Given the larger math favoring this play and Dallas being quite bad in yards per play allowed, the Giants being relatively good at yards per play and the fact that Dallas is #1 in points per play add that to the Giants 23rd ranking in points per play allowed all of those are pointing in the same direction actually that the adjustments to that base model pointing further and further that going for it is a good idea.


But the data you are giving me - what does it prove when you cite things like "Gives up 4th most yards per play?"

The 1st two games Dallas played Tampa Bay and the Chargers. That's a different animal.

And the two games that they played Philly and Carolina they were up 41-14 and 36-14 respectively until both teams put up meaningless 4th quarter points/ 4th quarter drives.
You can pick it apart all day long like that  
NoGainDayne : 10/11/2021 8:02 pm : link
but you can pick yourself right out of the larger patterns that have formed that way and that really isn't a strong thing to do with math. My way isn't perfect but it's better than doing it like that
RE: RE: RE: Eric did you read the article above?  
Eric on Li : 10/11/2021 8:36 pm : link
In comment 15409407 NoGainDayne said:
Quote:

The larger point to me seems to be, if the higher level results aren't working for you like when you look at your record? And you continuing to lose close games. Why not start going with the higher math more?


I think they need to be more aggressive and play for touchdowns when the odds are in their favor. 4th and 6 there, with jones so far running for his life on the first few plays, imo was not in their favor. 4th and 1 or 2 and I'd have felt differently, and had it been more like a 55-60 yarder I'd have gone for it. 54 seems comfortably in Gano's range and 6 yards to go i'd guess is a less than 50% probability.

I would almost always go for it in near the 50 instead of punting because that's just yardage.

I would almost always use timeouts at the end of quarters and take shots to get into fg range as long as there is time for 2 plays.

The latter 2 are areas where I hope the coaching staff learns to be aggressive.

As a road dog with a kicker who had been close to automatic im taking an early lead and points on the board though. Play it out if Gano hits it. The game is probably 13-10 when Jones got hurt and they are 1 defensive stop away from a lead at halftime and the ball. Obviously you can't anticipate the injuries, though sadly we can now anticipate defensive collapse pre-halftime, but even still if Dallas took it down and goes up 17-13, I'd be a lot more pissed they wasted those 36 seconds and 3 time outs not taking a few shots even with Glennon to try to get it to 17-16 than not going for a 4th and 6 that was still a long way from being turned into a td even if they got the 6 yards.

imo 36 seconds, fresh set of downs, and multiple time outs to get 20-30 yards for an attempt to get +3 points when the offense just started clicking

>>>>>> better odds than

getting a do or die 4th and 6, AND then covering the 30+ yards and converting in a tough part of the field for a TD to get the extra +4 points.
defensive, ball control, field position football  
Producer : 10/11/2021 8:56 pm : link
is dead. You can't go through a season playing that way. Sure maybe a big game or two, it helps, but as the fundamental philosophy of the team, it is obviously out of step.
Would not go as far to say it is dead. But ball control and  
chick310 : 10/11/2021 9:04 pm : link
defensive mindset is clearly not what drives WINS.

Score early and often. Make the other team take unnecessary risks to hold serve or risk falling behind. Let weather, key injuries or some other factor dictate aggressiveness, but looking to just keep games close and pull it out in the end is no longer winning football.

look at all the teams going  
Producer : 10/11/2021 9:07 pm : link
on 4th down, Or foregoing a FG on fourth. They don't respect defenses. Teams can move it 60 yards in an instant. Field position is becoming meaningless.
RE: You can pick it apart all day long like that  
giantstock : 10/11/2021 9:32 pm : link
In comment 15409499 NoGainDayne said:
Quote:
but you can pick yourself right out of the larger patterns that have formed that way and that really isn't a strong thing to do with math. My way isn't perfect but it's better than doing it like that


I'm picking it apart because it doesn't make a valid case that the Giants should have gone for it on 4th down. The dat you're proving is not taking all factors into account, thus it's incomplete.
RE: Would not go as far to say it is dead. But ball control and  
giantstock : 10/11/2021 9:42 pm : link
In comment 15409585 chick310 said:
Quote:
defensive mindset is clearly not what drives WINS.

Score early and often. Make the other team take unnecessary risks to hold serve or risk falling behind. Let weather, key injuries or some other factor dictate aggressiveness, but looking to just keep games close and pull it out in the end is no longer winning football.


Not arguing but in 2019 how high powered were the San Fran passing game and Tenn passing game? They were 13 and 21 respectively.

And what if you don't have an elite passing team?

Do you just give up? San Fran and Tenn - should they have just changed their style and pretend like they are KC and Green Bay?

Because San Fran and Tenn were 2nd and 3rd respectively in rushing yards.
RE: RE: Would not go as far to say it is dead. But ball control and  
chick310 : 10/11/2021 9:49 pm : link
In comment 15409704 giantstock said:
Quote:
In comment 15409585 chick310 said:


Quote:


defensive mindset is clearly not what drives WINS.

Score early and often. Make the other team take unnecessary risks to hold serve or risk falling behind. Let weather, key injuries or some other factor dictate aggressiveness, but looking to just keep games close and pull it out in the end is no longer winning football.




Not arguing but in 2019 how high powered were the San Fran passing game and Tenn passing game? They were 13 and 21 respectively.

And what if you don't have an elite passing team?

Do you just give up? San Fran and Tenn - should they have just changed their style and pretend like they are KC and Green Bay?

Because San Fran and Tenn were 2nd and 3rd respectively in rushing yards.


No, but there has to be some realization of who you are at the present time and go play ball, and what you need to do with your investment opportunities for the future (cap space and draft picks).

Good luck continuing to invest in the "pound the rock" mentality. You can be good with it (even great), and still be subject to being a dinosaur before the clock winds down.
RE: RE: RE: Would not go as far to say it is dead. But ball control and  
giantstock : 10/11/2021 10:16 pm : link
In comment 15409718 chick310 said:
Quote:
In comment 15409704 giantstock said:


Quote:


In comment 15409585 chick310 said:




No, but there has to be some realization of who you are at the present time and go play ball, and what you need to do with your investment opportunities for the future (cap space and draft picks).

Good luck continuing to invest in the "pound the rock" mentality. You can be good with it (even great), and still be subject to being a dinosaur before the clock winds down.


Huh? How did we get on the subject of future cap? I agree with you 100% about future cap.

But as of right now at this moment you agree with if you don't have an elite passing team and your best alternative is grind-and-pound football / keeping score close rather than pretend you are KC or Green Bay, then that's the bets way to go for this year, right?

Otherwise, was the San Fran and Tenn season's failures in 2019?
RE: RE: You can pick it apart all day long like that  
NoGainDayne : 10/12/2021 12:56 am : link
In comment 15409677 giantstock said:
Quote:
In comment 15409499 NoGainDayne said:


Quote:


but you can pick yourself right out of the larger patterns that have formed that way and that really isn't a strong thing to do with math. My way isn't perfect but it's better than doing it like that



I'm picking it apart because it doesn't make a valid case that the Giants should have gone for it on 4th down. The dat you're proving is not taking all factors into account, thus it's incomplete.


No you aren't seeing the numbers and how they work in a modeling perspective.

The baseline is go for it.

I am telling you numbers like the Cowboys allow the 4th most yards per play in the league. What do you want to argue it down to on a larger sample? Below league average? Because that's where it would have to get for it not to be a bigger reason to go for it. That tells me sorry, you don't really know how to use numbers because you are trying to explain away a sample of 5 games almost a full 3rd of a season. I'd bet a lot that the Cowboys D finishes the season averaging more than the league average yards per play.

Same thing for the Giants Offense. 5 games isn't insignificant that they are above league average in yards per play. You can try to explain it away all you want on that number too, but again, the numbers are the numbers. And both point in the direction of higher than the baseline. Not going to speculate on that because the offense they had at the time of the conversion attempt I expect to be better than the one they have subsequently this season. Either way 5 games and a rank of 9 isn't something to be scoffed at. Especially with the injuries they've had.

You are misunderstanding. It is you that is trying to make a case that the numbers should be below the baseline. Yet you have no numerical backing for that.

You want to say I don't have the data, great. But back it up for gods sake, with data. Not your little stories about why data you see isn't real. That's the opposite of sound data analysis.
RE: RE: RE: You can pick it apart all day long like that  
giantstock : 10/12/2021 2:20 am : link
In comment 15410051 NoGainDayne said:
Quote:
In comment 15409677 giantstock said:


Quote:


In comment 15409499 NoGainDayne said:


Quote:


but you can pick yourself right out of the larger patterns that have formed that way and that really isn't a strong thing to do with math. My way isn't perfect but it's better than doing it like that



I'm picking it apart because it doesn't make a valid case that the Giants should have gone for it on 4th down. The dat you're proving is not taking all factors into account, thus it's incomplete.



No you aren't seeing the numbers and how they work in a modeling perspective.

The baseline is go for it.

I am telling you numbers like the Cowboys allow the 4th most yards per play in the league. What do you want to argue it down to on a larger sample? Below league average? Because that's where it would have to get for it not to be a bigger reason to go for it. That tells me sorry, you don't really know how to use numbers because you are trying to explain away a sample of 5 games almost a full 3rd of a season. I'd bet a lot that the Cowboys D finishes the season averaging more than the league average yards per play.

Same thing for the Giants Offense. 5 games isn't insignificant that they are above league average in yards per play. You can try to explain it away all you want on that number too, but again, the numbers are the numbers. And both point in the direction of higher than the baseline. Not going to speculate on that because the offense they had at the time of the conversion attempt I expect to be better than the one they have subsequently this season. Either way 5 games and a rank of 9 isn't something to be scoffed at. Especially with the injuries they've had.

You are misunderstanding. It is you that is trying to make a case that the numbers should be below the baseline. Yet you have no numerical backing for that.

You want to say I don't have the data, great. But back it up for gods sake, with data. Not your little stories about why data you see isn't real. That's the opposite of sound data analysis.


No I am not misunderstanding. You are misrepresenting the data. For example, you're not taking into account when Dallas is up by 20-30 points in the 4th quarter that the defense most likely is not playing with the same urgency. And can you account for starters vs subs especially when it's 20+ point lead? Instead you are throwing this and a few other factors.

You're just leaving out too many factors that have enormous implications. Sure over the course of the season Dallas may allow a lot of yards per play. But that very well could be because in our division for example, they've had lots of blowouts. SO with blowouts, what makes you think they aren't going to allow more underneath passing just to run down the clock and end the game they've comfortably won?

I don't need to go through the effort to back up a fish story when I hear one. And you're telling a fish story by providing incomplete data. Because you don't have the necessary data don't get all upset with me.

Just look at the Phily Game. Dallas goes up 27-7 in early 3rd qtr. then Philly gets a TD with 7 minutes left in 3rd to make it 27-14. Dallas then scores two TDs in 4th qtr to make it 41-14. You can’t figure out for yourself that maybe Dallas at 41-14 might be playing less hard on D? I'm not supposed to mention this to you because I don't have some sort of model? You're deliberately not counting this type of information and a lot of other stuff and you're getting mad at me?

No see that's not it  
NoGainDayne : 10/12/2021 3:01 am : link
you are basically trying to tell me why I should just toss out numbers because you have anecdotes about why they don't hold.

I'm using larger representations of numbers to give examples of why you would actually have a reason to go over the baseline (which I've shown you says you go for it there)

Even if you could pick apart my examples of why you'd go above the baseline. Your anecdotes aren't really reasons to go below the baselines at all. You'd need larger numbers for that, of which you have none.

Pointing out that models don't explain everything is not a reason to not go with math representing larger patterns. I'd be very happy to believe what you are saying if you had any math to adjust the equation I shared in a larger way instead of you basically saying:

1. Getting this calc precisely right is not something either of us have the full math correct on therefore we should throw out the NYT calculation essentially

2. "My eyes and gut tell me the NYT calc doesn't apply here"

It actually completely disregards that the NYT calculation actually IS using math that is beyond what either of us can bring to bear on this problem and therefore is going to be right more often than either of our brains.

The computer is right more often because it is using far, far more data and processing power than either of our heads can contain. You pointing out individual examples is actually the exact reason these models were created.

Quote:
You are misrepresenting the data. For example, you're not taking into account when Dallas is up by 20-30 points in the 4th quarter that the defense most likely is not playing with the same urgency. And can you account for starters vs subs especially when it's 20+ point lead? Instead you are throwing this and a few other factors.


I was able to disprove this little theory of yours pretty fast:

1. Two teams have a larger margin of victory than the Cowboys. The Bills and Cardinals, they rank 1st and 14th respectively in yards allowed per play. Your little theory on junk time creating these higher yards per play allowed falls apart pretty quickly there.

2. You are really failing to understand this simple concept. The biggest numbers say go for it, if you want to say not to go with the math there and that "my data" is missing things, well you really need to have better data yourself than just trying to punch holes in anything I say. The position you are arguing commenced with the large mathematical hole.

I'm annoyed with you because you are trying to poke holes in my equation modifiers and really failing to address why the math that says go for it should be tossed out in the first place. And then you are trying to pick apart my equation modifiers and also really failing to understand those equation modifiers are being used because they represent larger samples of data which are inherently more reliable than smaller samples. It doesn't make them the truth or the gospel but it does make them better than anecdotes and saying "well this is unsolvable so what makes sense in my brain is better to go with." The process you are describing and laying out here is actually everything that is wrong with the Giants approach right now


This discussion  
madeinstars : 10/12/2021 5:04 am : link
is a great example of the difference of understanding analytics versus not getting it at all.

This is exactly the problem guys like Judge have. They just don't understand analytics on a fundamental level.
RE: RE: RE: RE: Also worth noting  
Gatorade Dunk : 10/12/2021 7:18 am : link
In comment 15409352 giantstock said:
Quote:
In comment 15409343 Enzo said:


Quote:


In comment 15409264 rsjem1979 said:


Quote:


In comment 15409254 Giantfan in skinland said:


Quote:


WFT is allowing 31 ppg to their opponents. So our 29 was below average against them.



Also worth noting that while 29 and 27 are well above average for the Giants, by modern NFL standards nobody would describe either as a "lot" of points.


agree. And let's not forge they only got in the endzone 2 times in that game.



Average league scoring is 23.7 - Giants SCORED 27 and 29. Now what they score doesn't count? That's how the narrative gets changed now?

Sure Ok. LOL.

Forget the league average; that has all the bad teams baked into it.

The average points per game for teams with a record above .500 so far this year is 26.675, and that includes the Bears' anemic offense.

And if we're talking about "a lot of points" then we should focus on the best offenses as the barometer. So far this season, the top 10 offenses in the league are averaging 30.26 ppg.
Completely agree with the OP  
Mike in Long Beach : 10/12/2021 10:33 am : link
I do agree he has the qualities of a leader. I also completely agree he's an un-evolved in-game play-caller (that's a lot of hyphens in a row).

Literally speaking, he not only doesn't even seem to understand how the math works, he's too stubborn to embrace even the idea of learning it. It will be a death sentence for this team, even if they do manage to field a decent roster.
RE: No see that's not it  
giantstock : 10/12/2021 1:18 pm : link
In comment 15410068 NoGainDayne said:
Quote:
you are basically trying to tell me why I should just toss out numbers because you have anecdotes about why they don't hold.

I'm using larger representations of numbers to give examples of why you would actually have a reason to go over the baseline (which I've shown you says you go for it there)

Even if you could pick apart my examples of why you'd go above the baseline. Your anecdotes aren't really reasons to go below the baselines at all. You'd need larger numbers for that, of which you have none.

Pointing out that models don't explain everything is not a reason to not go with math representing larger patterns. I'd be very happy to believe what you are saying if you had any math to adjust the equation I shared in a larger way instead of you basically saying:

1. Getting this calc precisely right is not something either of us have the full math correct on therefore we should throw out the NYT calculation essentially

2. "My eyes and gut tell me the NYT calc doesn't apply here"

It actually completely disregards that the NYT calculation actually IS using math that is beyond what either of us can bring to bear on this problem and therefore is going to be right more often than either of our brains.

The computer is right more often because it is using far, far more data and processing power than either of our heads can contain. You pointing out individual examples is actually the exact reason these models were created.



Quote:


You are misrepresenting the data. For example, you're not taking into account when Dallas is up by 20-30 points in the 4th quarter that the defense most likely is not playing with the same urgency. And can you account for starters vs subs especially when it's 20+ point lead? Instead you are throwing this and a few other factors.



I was able to disprove this little theory of yours pretty fast:

1. Two teams have a larger margin of victory than the Cowboys. The Bills and Cardinals, they rank 1st and 14th respectively in yards allowed per play. Your little theory on junk time creating these higher yards per play allowed falls apart pretty quickly there.

2. You are really failing to understand this simple concept. The biggest numbers say go for it, if you want to say not to go with the math there and that "my data" is missing things, well you really need to have better data yourself than just trying to punch holes in anything I say. The position you are arguing commenced with the large mathematical hole.

I'm annoyed with you because you are trying to poke holes in my equation modifiers and really failing to address why the math that says go for it should be tossed out in the first place. And then you are trying to pick apart my equation modifiers and also really failing to understand those equation modifiers are being used because they represent larger samples of data which are inherently more reliable than smaller samples. It doesn't make them the truth or the gospel but it does make them better than anecdotes and saying "well this is unsolvable so what makes sense in my brain is better to go with." The process you are describing and laying out here is actually everything that is wrong with the Giants approach right now



You’re deliberately not using the data I used and using other data. The Bills and Cards are different than the Cowboys. Address the cowboy’s vs Philly.

This is why I am telling you you do9dn't understand what data you are using. My point which you don’t seem to understand that is YOU CNA"T LUMP THE BILLS WITH THE COWBOYS.

The have completely different circumstances. Each game each team has completely different ramifications. For example, injuries across the board affect how a team plays. You need to add in so much more to draw any conclusions.

You don't use partial data to draw conclusions. That's what you're doing.
RE: RE: No see that's not it  
Gatorade Dunk : 10/12/2021 6:16 pm : link
In comment 15410776 giantstock said:
Quote:
In comment 15410068 NoGainDayne said:


Quote:


you are basically trying to tell me why I should just toss out numbers because you have anecdotes about why they don't hold.

I'm using larger representations of numbers to give examples of why you would actually have a reason to go over the baseline (which I've shown you says you go for it there)

Even if you could pick apart my examples of why you'd go above the baseline. Your anecdotes aren't really reasons to go below the baselines at all. You'd need larger numbers for that, of which you have none.

Pointing out that models don't explain everything is not a reason to not go with math representing larger patterns. I'd be very happy to believe what you are saying if you had any math to adjust the equation I shared in a larger way instead of you basically saying:

1. Getting this calc precisely right is not something either of us have the full math correct on therefore we should throw out the NYT calculation essentially

2. "My eyes and gut tell me the NYT calc doesn't apply here"

It actually completely disregards that the NYT calculation actually IS using math that is beyond what either of us can bring to bear on this problem and therefore is going to be right more often than either of our brains.

The computer is right more often because it is using far, far more data and processing power than either of our heads can contain. You pointing out individual examples is actually the exact reason these models were created.



Quote:


You are misrepresenting the data. For example, you're not taking into account when Dallas is up by 20-30 points in the 4th quarter that the defense most likely is not playing with the same urgency. And can you account for starters vs subs especially when it's 20+ point lead? Instead you are throwing this and a few other factors.



I was able to disprove this little theory of yours pretty fast:

1. Two teams have a larger margin of victory than the Cowboys. The Bills and Cardinals, they rank 1st and 14th respectively in yards allowed per play. Your little theory on junk time creating these higher yards per play allowed falls apart pretty quickly there.

2. You are really failing to understand this simple concept. The biggest numbers say go for it, if you want to say not to go with the math there and that "my data" is missing things, well you really need to have better data yourself than just trying to punch holes in anything I say. The position you are arguing commenced with the large mathematical hole.

I'm annoyed with you because you are trying to poke holes in my equation modifiers and really failing to address why the math that says go for it should be tossed out in the first place. And then you are trying to pick apart my equation modifiers and also really failing to understand those equation modifiers are being used because they represent larger samples of data which are inherently more reliable than smaller samples. It doesn't make them the truth or the gospel but it does make them better than anecdotes and saying "well this is unsolvable so what makes sense in my brain is better to go with." The process you are describing and laying out here is actually everything that is wrong with the Giants approach right now





You’re deliberately not using the data I used and using other data. The Bills and Cards are different than the Cowboys. Address the cowboy’s vs Philly.

This is why I am telling you you do9dn't understand what data you are using. My point which you don’t seem to understand that is YOU CNA"T LUMP THE BILLS WITH THE COWBOYS.

The have completely different circumstances. Each game each team has completely different ramifications. For example, injuries across the board affect how a team plays. You need to add in so much more to draw any conclusions.

You don't use partial data to draw conclusions. That's what you're doing.

You really don't understand how data works.

The purpose of data is to remove anecdotal noise and give yourself an agnostic view of a situation. You're bending over backwards to insert anecdotal noise into the data in an effort to distort what the larger dataset might suggest.

If you're using data properly, you analyze the full data, and then contextualize against the more limited scenario. You seem to want to approach it the other way around, and that's simply not the way it works.

You seem to want to examine each individual scenario on its own, and doing that provides absolutely no insight. If you're doing that, you're operating entirely by gut. And over time, aggregate will win out over gut, even if gut wins out occasionally along the way.

And if we're really being honest, JJ's gut hasn't exactly been infallible thus far.
Best part is even the anectdotal noise doesn't actually hold up....  
Giantfan in skinland : 10/12/2021 6:57 pm : link
What Nogaindayne said:
The Dallas defense gives up the 4th most yards per play at 6.3 which lends well to a conversion probability for 5 yards.

What giantstock said:
The 1st two games Dallas played Tampa Bay and the Chargers. That's a different animal.

And the two games that they played Philly and Carolina they were up 41-14 and 36-14 respectively until both teams put up meaningless 4th quarter points/ 4th quarter drives.

What the data shows:
In the fist quarter of games, Dallas is actually allowing MORE yards (6.5) than their overall average. So no, that data does not appear to be inflated due to garbage time soft defense.

Between their own 21-50 yard line, they allow 7.06 yards per play.

When the game is tied, they allow 6.16 yards per play.

All of this supports the basic point NGD made to you - high level data of all NFL teams suggests going for it and this is only strengthened when you look at the specifics of the situation (you can do the same thing with the Giants O as above - for example, so far this year, when they need 4-6 yards, they are averaging 7.15).

So again, if you want to answer the data, have the data to answer it. Just pointing out that they won games by large margins doesn't actually prove anything (and, in fact, the actual data shows that the yardage per play went down as the game became more lopsided).
Lol...  
Giantfan in skinland : 10/12/2021 7:02 pm : link
just realized I was only looking at rushing yardage....the averages allowed/gained are actually even higher with passing factored in.
A lack of introspection plus an excess of hindsight  
Grizz99 : 10/13/2021 12:07 pm : link
Has he gone for it and failed. This thread and the very same postrts would be bellyaching about that
RE: A lack of introspection plus an excess of hindsight  
Gatorade Dunk : 10/13/2021 12:35 pm : link
In comment 15412016 Grizz99 said:
Quote:
Has he gone for it and failed. This thread and the very same postrts would be bellyaching about that

Hey Grizz?

Remember when you got banned? I'm looking forward to that happening again.
RE: RE: RE: No see that's not it  
giantstock : 10/13/2021 1:21 pm : link
In comment 15411305 Gatorade Dunk said:
Quote:
In comment 15410776 giantstock said:


Quote:






You’re deliberately not using the data I used and using other data. The Bills and Cards are different than the Cowboys. Address the cowboy’s vs Philly.

This is why I am telling you you do9dn't understand what data you are using. My point which you don’t seem to understand that is YOU CNA"T LUMP THE BILLS WITH THE COWBOYS.

The have completely different circumstances. Each game each team has completely different ramifications. For example, injuries across the board affect how a team plays. You need to add in so much more to draw any conclusions.

You don't use partial data to draw conclusions. That's what you're doing.


You really don't understand how data works.

The purpose of data is to remove anecdotal noise and give yourself an agnostic view of a situation. You're bending over backwards to insert anecdotal noise into the data in an effort to distort what the larger dataset might suggest.

If you're using data properly, you analyze the full data, and then contextualize against the more limited scenario. You seem to want to approach it the other way around, and that's simply not the way it works.

You seem to want to examine each individual scenario on its own, and doing that provides absolutely no insight. If you're doing that, you're operating entirely by gut. And over time, aggregate will win out over gut, even if gut wins out occasionally along the way.

And if we're really being honest, JJ's gut hasn't exactly been infallible thus far.


No I understand how data works. I work with it and have for 30 years and analyze it and make decisions from it every day and have helped build other processes.

You seem to want to ignore that in order to use the data the nodayne wanted in terms of yards per play in order to make a sole decision, it will always be bogus. That stat doesn't account for other factors. You need more data to make that decision.

You and others don't seem to understand that if you are going to ride hard on one stat, then that one stat needs to be near infallible. Sorry, but that's how analyzing data works.

It's not "analytics" otherwise. You need inputs form mnay factors as you can.
RE: Best part is even the anectdotal noise doesn't actually hold up....  
giantstock : 10/13/2021 1:29 pm : link
In comment 15411329 Giantfan in skinland said:
Quote:
What Nogaindayne said:
The Dallas defense gives up the 4th most yards per play at 6.3 which lends well to a conversion probability for 5 yards.

What giantstock said:
The 1st two games Dallas played Tampa Bay and the Chargers. That's a different animal.

And the two games that they played Philly and Carolina they were up 41-14 and 36-14 respectively until both teams put up meaningless 4th quarter points/ 4th quarter drives.

What the data shows:
In the fist quarter of games, Dallas is actually allowing MORE yards (6.5) than their overall average. So no, that data does not appear to be inflated due to garbage time soft defense.

Between their own 21-50 yard line, they allow 7.06 yards per play.

When the game is tied, they allow 6.16 yards per play.

All of this supports the basic point NGD made to you - high level data of all NFL teams suggests going for it and this is only strengthened when you look at the specifics of the situation (you can do the same thing with the Giants O as above - for example, so far this year, when they need 4-6 yards, they are averaging 7.15).

So again, if you want to answer the data, have the data to answer it. Just pointing out that they won games by large margins doesn't actually prove anything (and, in fact, the actual data shows that the yardage per play went down as the game became more lopsided).


I would yield on this - if you could make mention on how many yards per play was Dallas giving up after they had the big leads of 35-14/ 41-14 etc vs the two teams. I said you (he) needs to look at ALL the data. THAT:S HOW YOU MKAE DECSIOSN WHEN YOU ARE USING THE TYPE OF STAT HE REFERENCED.

You got data to back that up then I'm on board. But the point to nodayne was you can't ignore it, right? Then I'm on board. I'm not dug in if you got it?

So back it up otherwise from what I recalls Philly for example had 75 yard drive and 90 yard drive when game wasn't much in doubt. Was number of yards per play low?

But this is the type of data you also need when making a decision to, right? These type of stats MATTER, don't they?
RE: A lack of introspection plus an excess of hindsight  
giantstock : 10/13/2021 1:33 pm : link
In comment 15412016 Grizz99 said:
Quote:
Has he gone for it and failed. This thread and the very same postrts would be bellyaching about that


That was my point on this thread or another.
Except the NYT calculator isn't one stat  
NoGainDayne : 10/13/2021 1:50 pm : link
it is a machine learning algorithm built on tons of advanced data built to help with this problem.

I have done nothing but show you data that would increase that baseline. And to this point you have shown no data to explain why you should decrease it, only anecdotes. And you are also seeming to miss the point that even if you can move the numbers they'd have to move a whole lot to move past the league average for the modifiers I'm proposing.

Talk about your work with data all you'd like but I don't really care how much experience anyone has if they can't use data to make the points they are trying to make. You just are focusing how I didn't prove every little point but I think the bigger issue is that you didn't prove your point at all.

On an almost 1/3rd use of season sample I don't even think that type of thinking is productive unless you are going to have a machine do all the thinking. I brought in Arizona and Buffalo as an example because I disagreed with the assertion that you were trying to make that you can grade down yards per play allowed because Dallas had a few blowouts and played good offenses. In fact I looked into it for you and Dallas' five opponents this year Tampa, NY, Philly, LAC and Carolina averaged 5.96 yards a game which is only 4.5% above the league median this year at 5.7. (I'm linking to it because the only reason I didn't do average was i'd have to type all those numbers lol) Or 4.5% above the median value.

Anyway, my point is I'd classify your exercise here as a complete waste of time. Because Dallas ranks 29th in the league in giving up yards and moving them below league average isn't going to work with flimsy anecdotal logic like you proposed. That's what I don't understand about this whole line of thinking. The NYT calculator plus data supporting a GREATER reason to go for that is 1/3rd of a season sample should be plenty of information to say ok this was a bad move.

Your whole process with this is off, sorry. It's not the job of an analysis to answer every single last question and exception that your mind can come up with.
NFL Yards Per Play - ( New Window )
RE: RE: A lack of introspection plus an excess of hindsight  
Gatorade Dunk : 10/13/2021 3:18 pm : link
In comment 15412167 giantstock said:
Quote:
In comment 15412016 Grizz99 said:


Quote:


Has he gone for it and failed. This thread and the very same postrts would be bellyaching about that



That was my point on this thread or another.

You two can talk yourselves through an inability to process data.
RE: RE: RE: RE: No see that's not it  
Gatorade Dunk : 10/13/2021 3:20 pm : link
In comment 15412141 giantstock said:
Quote:
In comment 15411305 Gatorade Dunk said:


Quote:


In comment 15410776 giantstock said:


Quote:






You’re deliberately not using the data I used and using other data. The Bills and Cards are different than the Cowboys. Address the cowboy’s vs Philly.

This is why I am telling you you do9dn't understand what data you are using. My point which you don’t seem to understand that is YOU CNA"T LUMP THE BILLS WITH THE COWBOYS.

The have completely different circumstances. Each game each team has completely different ramifications. For example, injuries across the board affect how a team plays. You need to add in so much more to draw any conclusions.

You don't use partial data to draw conclusions. That's what you're doing.


You really don't understand how data works.

The purpose of data is to remove anecdotal noise and give yourself an agnostic view of a situation. You're bending over backwards to insert anecdotal noise into the data in an effort to distort what the larger dataset might suggest.

If you're using data properly, you analyze the full data, and then contextualize against the more limited scenario. You seem to want to approach it the other way around, and that's simply not the way it works.

You seem to want to examine each individual scenario on its own, and doing that provides absolutely no insight. If you're doing that, you're operating entirely by gut. And over time, aggregate will win out over gut, even if gut wins out occasionally along the way.

And if we're really being honest, JJ's gut hasn't exactly been infallible thus far.



No I understand how data works. I work with it and have for 30 years and analyze it and make decisions from it every day and have helped build other processes.

You seem to want to ignore that in order to use the data the nodayne wanted in terms of yards per play in order to make a sole decision, it will always be bogus. That stat doesn't account for other factors. You need more data to make that decision.

You and others don't seem to understand that if you are going to ride hard on one stat, then that one stat needs to be near infallible. Sorry, but that's how analyzing data works.

It's not "analytics" otherwise. You need inputs form mnay factors as you can.

No, you really don't.

You continue to ask for anecdotal variables to be factored in without any understanding of aggregated data.

Whoever is paying you to work with their data should ask for a refund.
RE: RE: RE: RE: RE: No see that's not it  
giantstock : 10/13/2021 4:18 pm : link
In comment 15412343 Gatorade Dunk said:
Quote:
In comment 15412141 giantstock said:


Quote:





No, you really don't.

You continue to ask for anecdotal variables to be factored in without any understanding of aggregated data.

Whoever is paying you to work with their data should ask for a refund.


You'd be fired after the 1st day.

Making decisions with incomplete data is not analysis and knowing that it's incomplete and not accepting is fireable. It's hubris. It's wrong. There's no place for it.
No giantstock  
NoGainDayne : 10/13/2021 4:28 pm : link
you are talking about analysis paralysis. Where you want to disregard the results of something because you can fathom examples where it doesn't work. It doesn't mean you toss it out. It means you situate it, properly with other data based arguments. You have not engaged in this process, and that is your problem.

We bump up against this stuff sometimes and the Giants seem to think this way too. "Oh well this is a complicated problem so the computer can't solve it, let's just go with what I think." That's not really the proper way to leverage or grow with technology, which is to say what does it know better than me? What adjustments do I need to make? Can I make those adjustments with data? DATA not individual examples. Or generalizations like I've heard on this thread like "you gotta put points on the board" it's like yeah you do, touchdowns, that's why more and more teams are going for it.

There is consistently this weird dichotomy when this topic comes up where people want to talk about how complex these problems are to solve, and to do it right it involves a massive project that does take in data sets like weather and team tendencies and plays etc. But the Giants clearly don't get the even more basic stuff and more importantly just because you can't rely on an output all the time, or it might be missing key data points in the ultimate non-linear equation does not mean you just toss it out or you attempt to challenge it with unsubstantiated claims.

RE: Except the NYT calculator isn't one stat  
giantstock : 10/13/2021 5:11 pm : link
In comment 15412191 NoGainDayne said:
Quote:
it is a machine learning algorithm built on tons of advanced data built to help with this problem.

I have done nothing but show you data that would increase that baseline. Play - ( New Window )


If it takes into account the Philly scenario then it's probably acceptable and I’ll be with you probably. I keep asking the data for it and you keep avoiding the question. I've asked something simple--

SHOW ME THE DATA FOR THIS PHILLY GAME. I PROVIDED YOU DATA FOR THIS GAME regarding the drives and scores. Instead of providing me data, you write me novel replies. So I’m doing the novel in return. But If you replied specifically to my question, great tell when you replied on this thread and I’ll read it. I’m not dug in but I’m not reading your entire novels for a simple question.

In addition, relating to 1st para above, I did also ask that in football analytics last I heard that every TD the teams should go for 2. Not sure if what I heard is right but if it is, then which teams do it I asked? I never got a reply from you. You blew right past that question just as you blew by my Philly question.

The problem is that you want to TELL ME that this algorithm is all encompassing instead of SHOWING ME providing facts OF THE PHILLY GAME which I’ve asked.

A further point is that you seem to think that because you put out data which is probably faulty which I have pointed out, that you’ve twisted this into that I NEED then to provide YOU some type of modified pinpoint system of my own to counter it? Bullcrap.

I’m not getting paid for any of this and I suppose neither are you. I just pointed out your faults and have asked you to prove it. That was YOUR DEFENSE.

You provided data and I said it doesn’t support the Philly game. PROVE TO ME THAT IT DOES. That's what I have asked/implied. If you have provided data regarding specifically the Philly game tell me which post you replied to. It still doesn’t address the Analytics of going for it after a TD but one issue at a time.

I’m not “dug in.” I just want you to prove what you say rather than tell me what you say is accurate.


No man  
NoGainDayne : 10/13/2021 5:22 pm : link
you tried to say the average was pumped up by junk time points and I made a very strong case with data why you shouldn't toss in random things like that. Because teams with larger margins of victory and statistically more junk time do not give up nearly as many yards per play.

The assumption that you should even consider factors of "junk time" in team averages is something that you'd actually have to prove a lot better for me not to be able to essentially do what I did and grab data to show that you hadn't even done basic homework to prove out the assertion of junk time being able to move the needle over a 5 game span.

If you feel I fully didn't address point it's because it wasn't a strong enough point to begin with and as I have been asking you to actually come to the table with data if you are going to tell someone their data is bad.
RE: No giantstock  
giantstock : 10/13/2021 5:33 pm : link
In comment 15412452 NoGainDayne said:
Quote:

We bump up against this stuff sometimes and the Giants seem to think this way too. "Oh well this is a complicated problem so the computer can't solve it, let's just go with what I think." T


No this is what you're not understanding. I'm not saying throw away the data.

I'm saying YOU are WAY TOO INFLUENCED by it as - how can I put this ' "I'm all in with just this data!"

The data you are providing should be used as tool among other tools. You seem to want to use it soley as being your sole guide to make decisions.

Am I wrong in these assumptions?
RE: No man  
giantstock : 10/13/2021 5:35 pm : link
In comment 15412522 NoGainDayne said:
Quote:
you tried to say the average was pumped up by junk time points a


Junk time points in the Philly game. Yes. You got the data for the Philly game? If you do- show me when game swa swon at 27-7 and 41-14.
I am saying that  
NoGainDayne : 10/13/2021 5:59 pm : link
one half of one game isn't enough to move that average significantly. Especially when they are ranked 29th in the league it's not enough to move them past the middle of the pack which is what would need to happen before you'd even consider calling it a downward modifier.

Either way there were other upward modifiers I made. Then there is the base calculation, again.

The larger point is even if you can pick apart my modifiers you haven't made one single good point about why you'd adjust the baseline calculation down.
RE: RE: No giantstock  
Gatorade Dunk : 10/13/2021 6:01 pm : link
In comment 15412536 giantstock said:
Quote:
In comment 15412452 NoGainDayne said:


Quote:



We bump up against this stuff sometimes and the Giants seem to think this way too. "Oh well this is a complicated problem so the computer can't solve it, let's just go with what I think." T




No this is what you're not understanding. I'm not saying throw away the data.

I'm saying YOU are WAY TOO INFLUENCED by it as - how can I put this ' "I'm all in with just this data!"

The data you are providing should be used as tool among other tools. You seem to want to use it soley as being your sole guide to make decisions.

Am I wrong in these assumptions?

You're really trying to claim that you've been working with data for 30 years but you're still on here making anecdotal arguments?
RE: RE: RE: RE: RE: RE: No see that's not it  
Gatorade Dunk : 10/13/2021 6:03 pm : link
In comment 15412433 giantstock said:
Quote:
In comment 15412343 Gatorade Dunk said:


Quote:


In comment 15412141 giantstock said:


Quote:





No, you really don't.

You continue to ask for anecdotal variables to be factored in without any understanding of aggregated data.

Whoever is paying you to work with their data should ask for a refund.



You'd be fired after the 1st day.

Making decisions with incomplete data is not analysis and knowing that it's incomplete and not accepting is fireable. It's hubris. It's wrong. There's no place for it.

Data is data.

You are trying to storytell with the data before it's even computed.

You're a fraud.
RE: I am saying that  
giantstock : 10/13/2021 6:10 pm : link
In comment 15412582 NoGainDayne said:
Quote:
one half of one game isn't enough to move that average significantly. Especially when they are ranked 29th in the league it's not enough to move them past the middle of the pack which is what would need to happen before you'd even consider calling it a downward modifier.

Either way there were other upward modifiers I made. Then there is the base calculation, again.

The larger point is even if you can pick apart my modifiers you haven't made one single good point about why you'd adjust the baseline calculation down.


Because the scenarios of each game are different. And every data point/ factor/ affects other data points/ factors.

You haven't provided data for the Philly game that I asked, have you? Yet you are telling me I'm wrong? It seems that either you don't have it or it doesn't prove your point so you don't want to share it.

And yet I'm supposed to say "Sure I believe you because you say so." C'mon.

You pushed our argument because you said/implied "My data is correct." SO from the beginning I spoke of the Philly game. So all I'm asking is "Show me. Don't Tell me."
RE: RE: RE: RE: RE: RE: RE: No see that's not it  
giantstock : 10/13/2021 6:12 pm : link
In comment 15412589 Gatorade Dunk said:
Quote:
In comment 15412433 giantstock said:


Quote:


In comment 15412343 Gatorade Dunk said:


Quote:


In comment 15412141 giantstock said:


Quote:





No, you really don't.

You continue to ask for anecdotal variables to be factored in without any understanding of aggregated data.

Whoever is paying you to work with their data should ask for a refund.



You'd be fired after the 1st day.

Making decisions with incomplete data is not analysis and knowing that it's incomplete and not accepting is fireable. It's hubris. It's wrong. There's no place for it.


Data is data.

You are trying to storytell with the data before it's even computed.

You're a fraud.


I'm doubting the data until it's shown to me.

I'm not like you that "believes" and then throws bombs at others because you're so callous.
RE: RE: RE: No giantstock  
giantstock : 10/13/2021 6:21 pm : link
In comment 15412588 Gatorade Dunk said:
Quote:
In comment 15412536 giantstock said:


Quote:


In comment 15412452 NoGainDayne said:


Quote:



We bump up against this stuff sometimes and the Giants seem to think this way too. "Oh well this is a complicated problem so the computer can't solve it, let's just go with what I think." T




No this is what you're not understanding. I'm not saying throw away the data.

I'm saying YOU are WAY TOO INFLUENCED by it as - how can I put this ' "I'm all in with just this data!"

The data you are providing should be used as tool among other tools. You seem to want to use it soley as being your sole guide to make decisions.

Am I wrong in these assumptions?


You're really trying to claim that you've been working with data for 30 years but you're still on here making anecdotal arguments?


I'm doubting the data beucase it is not factoring in what I've asked.

I made the comment asking to disprove it.

This isn't my job. SO when someone comes to me pushing data as the sole source ot use - I ask to prove it.

If you have any clue about data - when someone asks for proof of something specific - they should be able to provide.

What you seem to fail to understand is that I'm not arguing my point is right. I am arguing his sole point is wrong and I am asking him to prove the scenario I spiked out.

And I have ben very open if he can prove it- I'd probably yield. It's typical of posters like you. You don't want to address the specific question. Instead you trun it personal on me.

I take back what I said. You wouldn't last 5 minutes in our company.

SHOW ME THE DATA THAT I ASKED. It's that simple.
A. I have no idea what you want me to provide  
NoGainDayne : 10/13/2021 6:25 pm : link
B. Even if I did I wouldn't get it because the data from one half of one game isn't particularly relevant if it doesn't show up in larger patterns of data. Which I continue to ask you to show

Here is the brass tax. The teams Dallas have played average 5.96 yards per play overall including games with Dallas and Dallas allows 6.3 yards per play which is 5.7% more yards PER PLAY. That says that Dallas's defense by any way you slice it, isn't very good so you really have no reason to go below the baseline value which says you go for it there.

I am actually completely confused as to what point you are making about this Philly thing. If I could "get you your data" I would but it sounds like you just want to point at rocks for me to turn over so you can say you have a leg to stand on until I'm done turning the rocks over for you but your point isn't good and I'm not sure why you continue to argue it without any DATA.

Please stop with the individual examples, it is not helping your case that you actually understand this stuff. I'm talking about a 5 game sample as statistically viable and I think anything less than 3 and I couldn't possibly even take it seriously compared to patterns that hold over 5 games.

I'm just really failing to see what your larger point is in all this as well. I came with a NYT calculation showing the math says go for it and you haven't brought one remotely good piece of data to say why that math is wrong in this situation, not one.
RE: RE: RE: RE: RE: RE: RE: RE: No see that's not it  
Gatorade Dunk : 10/13/2021 6:50 pm : link
In comment 15412604 giantstock said:
Quote:
In comment 15412589 Gatorade Dunk said:


Quote:


In comment 15412433 giantstock said:


Quote:


In comment 15412343 Gatorade Dunk said:


Quote:


In comment 15412141 giantstock said:


Quote:





No, you really don't.

You continue to ask for anecdotal variables to be factored in without any understanding of aggregated data.

Whoever is paying you to work with their data should ask for a refund.



You'd be fired after the 1st day.

Making decisions with incomplete data is not analysis and knowing that it's incomplete and not accepting is fireable. It's hubris. It's wrong. There's no place for it.


Data is data.

You are trying to storytell with the data before it's even computed.

You're a fraud.



I'm doubting the data until it's shown to me.

I'm not like you that "believes" and then throws bombs at others because you're so callous.

The data is the data. It's fully available to you.

This isn't a matter of "believing" the data. The data exists. The fact that you need the data to be believable to you is proof positive that you have no idea how data actually works.
RE: RE: RE: RE: No giantstock  
Gatorade Dunk : 10/13/2021 6:54 pm : link
In comment 15412613 giantstock said:
Quote:
In comment 15412588 Gatorade Dunk said:


Quote:


In comment 15412536 giantstock said:


Quote:


In comment 15412452 NoGainDayne said:


Quote:



We bump up against this stuff sometimes and the Giants seem to think this way too. "Oh well this is a complicated problem so the computer can't solve it, let's just go with what I think." T




No this is what you're not understanding. I'm not saying throw away the data.

I'm saying YOU are WAY TOO INFLUENCED by it as - how can I put this ' "I'm all in with just this data!"

The data you are providing should be used as tool among other tools. You seem to want to use it soley as being your sole guide to make decisions.

Am I wrong in these assumptions?


You're really trying to claim that you've been working with data for 30 years but you're still on here making anecdotal arguments?



I'm doubting the data beucase it is not factoring in what I've asked.

I made the comment asking to disprove it.

This isn't my job. SO when someone comes to me pushing data as the sole source ot use - I ask to prove it.

If you have any clue about data - when someone asks for proof of something specific - they should be able to provide.

What you seem to fail to understand is that I'm not arguing my point is right. I am arguing his sole point is wrong and I am asking him to prove the scenario I spiked out.

And I have ben very open if he can prove it- I'd probably yield. It's typical of posters like you. You don't want to address the specific question. Instead you trun it personal on me.

I take back what I said. You wouldn't last 5 minutes in our company.

SHOW ME THE DATA THAT I ASKED. It's that simple.

I'll stand by my earlier post.

YOU'RE A FRAUD.
RE: A. I have no idea what you want me to provide  
giantstock : 10/13/2021 7:32 pm : link
In comment 15412616 NoGainDayne said:
Quote:
B. Even if I did I wouldn't get it because the data from one half of one game isn't particularly relevant if it doesn't show up in larger patterns of data. Which I continue to ask you to show

Here is the brass tax. The teams Dallas have played average 5.96 yards per play overall including games with Dallas and Dallas allows 6.3 yards per play which is 5.7% more yards PER PLAY. That says that Dallas's defense by any way you slice it, isn't very good so you really have no reason to go below the baseline value which says you go for it there.

I am actually completely confused as to what point you are making about this Philly thing. If I could "get you your data" I would but it sounds like you just want to point at rocks for me to turn over so you can say you have a leg to stand on until I'm done turning the rocks over for you but your point isn't good and I'm not sure why you continue to argue it without any DATA.

Please stop with the individual examples, it is not helping your case that you actually understand this stuff. I'm talking about a 5 game sample as statistically viable and I think anything less than 3 and I couldn't possibly even take it seriously compared to patterns that hold over 5 games.

I'm just really failing to see what your larger point is in all this as well. I came with a NYT calculation showing the math says go for it and you haven't brought one remotely good piece of data to say why that math is wrong in this situation, not one.


Let me ask you this. If you don't have enough data to differentiate between blowouts and close games, then what is it that you think you have? Unless you don’t believe that defensive teams play any softer with a lead that they most definitely won't lose?

Let me reiterate, I don’t know this for sure what I am about to say- and I admit I am more a hoop guy - but my point has been I believe defenses do. I'll reiterate I cannot support that. But it's something I have believed a long time. Keep in mind I believe this AS A FAN.

Are you suggesting to me that because I can't prove this, I can't ask the question of you when you are using the data in a manner that excludes this?

Let me put this another way, as an NBA fan; I would have doubted a small guard like Steph Curry being the combo of shooter/scorer could be so efficient as he has been for as long as he has been. I grew when the big men reigned. But now the data has shown yes Curry is what I never imagined a small player with such small stature could be. I’m just a fan.

Similar to your data. I'm just a fan. So are you telling me your data doesn't or can't differentiate between when a game is a blowout vs when it is not? And as a result it can't differentiate for example between yards per play when the game is a blowout vs not? I'm surprised when you tell me you're not sure what I'm asking. In the Philly game vs Dallas at some point whether 27-7 or 34-14 or 41-14 the game was considered over. You are not understanding that I'm seeking that data on yards per play once the game is officially "done?"

Why is what I’m asking do you feel is so wrong? If you have created this amazing system then wouldn’t you want to differentiate between when a game is decided vs when it is not? And don’t factors such as injuries affect the data? Are you accounting for that? There are so many factors in football that affect behavior, correct? *****As the stats guy you are, wouldn't you want to include this data?
RE: RE: RE: RE: RE: No giantstock  
giantstock : 10/13/2021 7:50 pm : link
In comment 15412643 Gatorade Dunk said:
Quote:
In comment 15412613 giantstock said:


Quote:





I'll stand by my earlier post.

YOU'RE A FRAUD.


Yeah that means a lot. You "Stand by your comment." I can care less what a moron thinks-- And yes I mean you.
RE: RE: RE: RE: RE: RE: RE: RE: RE: No see that's not it  
giantstock : 10/13/2021 7:51 pm : link
In comment 15412637 Gatorade Dunk said:
Quote:
In comment 15412604 giantstock said:


Quote:





The data is the data. It's fully available to you.

This isn't a matter of "believing" the data. The data exists. The fact that you need the data to be believable to you is proof positive that you have no idea how data actually works.


This has to be one of the most dumbest posts I can recall. Thanks.
Ok I actually understand what you are asking so I can better respond  
NoGainDayne : 10/13/2021 8:57 pm : link
1. I don't have a system, I don't have access to the data or the time I'd need to build it out correctly even though I do have a plan that I've looked into partnering with a large AI company on (but they won't let me gamble with it so that stopped those convos for now)

2. The NYT calculator is actually working on expected points added per play so the math behind recommendation holds in lots of different situations or it wouldn't be that .37 we saw above would be a lot closer to 0

3. The way you are asking to consume data is very like "pivot table-y" and I wouldn't advocate for that at all. Nor would I dive down out of the equation that removes you from the expected points added calc embedded in the NYT algo IS accounting for a lot more variables and the work you are actually doing shrinks the sample size as something like you are suggesting would you have no guarantees you aren't overweighing factors already weighed.

Like here is the thing, as I said, I don't mind using yards per play at this point in the season because we have a sample of about 1K plays. Which actually isn't bad for a game like football.

When you start trying to hack that 1K into pieces though, you start to get into a zone of dangerous conclusions.

Anyway we can probably put this to bed because the 4th quarter has the lowest yards per play 5.44 (1st - 5.66, 2nd - 5.93, 3rd - 5.72) also average yards when leading is 6.04 vs. trailing at 5.48. These numbers suggest that score differential has very little affect on yardage as you'd also have to figure it gets larger as the game goes on.

Which really brings me to the crux of my frustration with what you were coming at me with which is. I didn't see any reason to think what you were suggesting about junk time was true, I didn't really think of it as a valid use of my time to investigate. You as the person that believes in that should really be the person hunting down the data. Which is also how it should go in the professional world unless your assertion is obvious to everyone around you.

And that's the larger point I'm making on data, if something with a large sample and proven math is something that you want to personally disagree with, the onus is really on you to shift the discussion to your perspective with some other large data set. Not an unsubstantiated assertion with a few examples. That's not really sound data analysis or collaborative processes at all.
2021 NFL Splits - ( New Window )
RE: Ok I actually understand what you are asking so I can better respond  
giantstock : 10/14/2021 4:29 am : link
In comment 15412798 NoGainDayne said:
Quote:


Anyway we can probably put this to bed because the 4th quarter has the lowest yards per play 5.44 (1st - 5.66, 2nd - 5.93, 3rd - 5.72) also average yards when leading is 6.04 vs. trailing at 5.48. These numbers suggest that score differential has very little affect on yardage as you'd also have to figure it gets larger as the game goes on.

Which really brings me to the crux of my frustration with what you were coming at me with which is. I didn't see any reason to think what you were suggesting about junk time was true, I didn't really think of it as a valid use of my time to investigate. You as the person that believes in that should really be the person hunting down the data. Which is also how it should go in the professional world unless your assertion is obvious to everyone around you.

And that's the larger point I'm making on data, if something with a large sample and proven math is something that you want to personally disagree with, the onus is really on you to shift the discussion to your perspective with some other large data set. Not an unsubstantiated assertion with a few examples. That's not really sound data analysis or collaborative processes at all. 2021 NFL Splits - ( New Window )


No. Let's end it here. I don't agree with how you're grouping the 1st paragraph. You have to account for blowouts vs close games. If I'm understanding right, you're not accounting for that because you don't have enough data to draw any meaningful conclusion.

Secondly, in the professional world- it's not true what you are suggesting. But possibly we're looking at it from different angles. This creating you made is in essence a proposal you want to integrate into our systems/processes. It doesn't have to be disproved. But it's upon you to come up with the answers to questions. And if what I think your saying is that the question I asked -you can't use because there isn't enough data so you chose to lump it in with other team figuring they are all relatable and thus you came up with an aggregate.

I think there are so many variables in football that affect so much of what's happening in the game which causes other variables within the game to change (such as injury and weather among other things). It doesn't sound like you can account for all these variables. Thus that makes the data you are presenting not as helpful as you think though still along with other data can be a useful tool used with other tools. I can't prove it. But without all the data I can be skeptical of it. Isn't that true in any business environment unless the business is desperate? You're telling me that there isn't enough data to answer my concern. How, if I were a business leader, would I accept that answer if I feel that there are so many variables that you can't account for?

Are you accounting for injuries and weather for example?
You did a great job with htis  
giantstock : 10/14/2021 8:16 pm : link
Great job. Good luck using it for any reason that you wish and hope you have greatest success and it shows what you want it to show.

I do believe there are other factors that can't be accounted for but this type of stuff is of high value.

Congrats.

Going ot post something similar on your other thread.

Great job.
RE: You did a great job with htis  
NoGainDayne : 10/15/2021 3:26 pm : link
In comment 15413856 giantstock said:
Quote:
Great job. Good luck using it for any reason that you wish and hope you have greatest success and it shows what you want it to show.

I do believe there are other factors that can't be accounted for but this type of stuff is of high value.

Congrats.

Going ot post something similar on your other thread.

Great job.


Honestly this isn't even the good part of my model. These are just standard underpinning of applying game theory (expected points per play) to play calling. And you are absolutely right, models doing that effectively would need to play out the conditional probabilities of each and every matchup. Right down to expected trajectories and OL matchups of like the possibility of blitz vs. DE vs. OT one on one.

A real solid expected points calculation would take real data as well as millions or hundreds of millions of simulated plays. For an interesting read on that look at the link below.

Actually the strength of my system is more on resource allocation and using incentive structures and budgeting combined with advanced projection technology to optimize for wins. Along these lines you could actually arbitrage bad seasons. If you'd like to email me I would be happy to send you my full proposal / explanation of how it works. You'd probably find it interesting. George@prospero.ai
Mu Zero by Google DeepMind - ( New Window )
Pages: 1 2 3 <<Prev | Show All |
Back to the Corner