IPB
 

Welcome Guest ( Log In | Register )

Buy Eagles tickets or search for cheap sports tickets including all playoff games and super bowl tickets @ LetsGetTickets.com
Posts in this topic
- nephillymike   The detailed analysis of 4th and 1, 4th and 8   Sep 30 2017, 09:07 PM
- - Zero   I didn't follow the links, I also didn't s...   Oct 1 2017, 05:11 AM
|- - nephillymike   QUOTE (Zero @ Oct 1 2017, 05:11 AM) I did...   Oct 1 2017, 06:34 AM
|- - Zero   QUOTE (nephillymike @ Oct 1 2017, 07:34 A...   Oct 1 2017, 06:54 AM
||- - nephillymike   QUOTE (Zero @ Oct 1 2017, 06:54 AM) This ...   Oct 1 2017, 09:30 AM
|- - The Franchise   QUOTE (nephillymike @ Oct 1 2017, 07:34 A...   Oct 5 2017, 03:52 PM
|- - Rick   QUOTE (The Franchise @ Oct 5 2017, 04:52 ...   Oct 5 2017, 04:31 PM
|- - The Franchise   QUOTE (Rick @ Oct 5 2017, 04:31 PM) Becau...   Oct 5 2017, 04:35 PM
|- - Rick   QUOTE (The Franchise @ Oct 5 2017, 05:35 ...   Oct 6 2017, 05:16 AM
|- - The Franchise   QUOTE (Rick @ Oct 6 2017, 06:16 AM) You k...   Oct 6 2017, 09:44 AM
|- - Rick   QUOTE (The Franchise @ Oct 6 2017, 10:44 ...   Oct 7 2017, 07:07 AM
|- - The Franchise   QUOTE (Rick @ Oct 7 2017, 07:07 AM) I kno...   Oct 7 2017, 11:56 AM
|- - Rick   QUOTE (The Franchise @ Oct 7 2017, 12:56 ...   Oct 7 2017, 03:46 PM
|- - The Franchise   QUOTE (Rick @ Oct 7 2017, 04:46 PM) I hav...   Oct 7 2017, 04:11 PM
|- - nephillymike   QUOTE (The Franchise @ Oct 7 2017, 04:11 ...   Oct 7 2017, 04:59 PM
||- - The Franchise   QUOTE (nephillymike @ Oct 7 2017, 04:59 P...   Oct 7 2017, 05:19 PM
||- - Rick   QUOTE (The Franchise @ Oct 7 2017, 06:19 ...   Oct 7 2017, 06:51 PM
||- - nephillymike   QUOTE (The Franchise @ Oct 7 2017, 05:19 ...   Oct 7 2017, 09:09 PM
||- - The Franchise   QUOTE (nephillymike @ Oct 7 2017, 10:09 P...   Oct 7 2017, 10:47 PM
||- - Pila   QUOTE (The Franchise @ Oct 8 2017, 03:47 ...   Oct 8 2017, 12:53 AM
||- - Rick   QUOTE (Pila @ Oct 8 2017, 01:53 AM) Forec...   Oct 8 2017, 08:02 AM
||- - The Franchise   QUOTE (Pila @ Oct 8 2017, 01:53 AM) Forec...   Oct 8 2017, 11:45 AM
||- - Pila   QUOTE (The Franchise @ Oct 8 2017, 04:45 ...   Oct 8 2017, 03:15 PM
|- - nephillymike   QUOTE (The Franchise @ Oct 7 2017, 04:11 ...   Oct 7 2017, 05:00 PM
- - Rick   Excellent analysis of the numbers. That's the ...   Oct 1 2017, 08:18 AM
|- - Zero   QUOTE (Rick @ Oct 1 2017, 09:18 AM) Excel...   Oct 1 2017, 09:36 AM
|- - nephillymike   QUOTE (Zero @ Oct 1 2017, 09:36 AM) I...   Oct 1 2017, 09:52 AM
|- - Rick   QUOTE (nephillymike @ Oct 1 2017, 10:52 A...   Oct 2 2017, 05:06 AM
|- - Pila   QUOTE (Rick @ Oct 2 2017, 10:06 AM) Well-...   Oct 2 2017, 08:51 AM
- - Zero   RE: The detailed analysis of 4th and 1, 4th and 8   Oct 1 2017, 10:07 AM
- - mcnabbulous   Cool stuff, Mikey.   Oct 1 2017, 11:22 AM
- - mcnabbulous   Don’t be offended, Mike. He’s just an asshole.   Oct 7 2017, 09:43 PM
|- - The Franchise   QUOTE (mcnabbulous @ Oct 7 2017, 10:43 PM...   Oct 7 2017, 10:44 PM
- - Aquila   QUOTE (mcnabbulous @ Oct 8 2017, 03:43 AM...   Oct 8 2017, 07:04 AM
- - nephillymike   http://www.phillyvoice.com/dislodge-yourse...-8-ca...   Oct 8 2017, 07:16 AM
|- - nephillymike   QUOTE (nephillymike @ Oct 8 2017, 07:16 A...   Oct 8 2017, 04:51 PM
|- - The Franchise   I didn't see that earlier. What can I say? I...   Oct 8 2017, 05:39 PM
|- - Rick   QUOTE (The Franchise @ Oct 8 2017, 06:39 ...   Oct 9 2017, 05:15 AM
|- - Zero   QUOTE (Rick @ Oct 9 2017, 06:15 AM) OMG...   Oct 9 2017, 05:51 AM
|- - Pila   QUOTE (Zero @ Oct 9 2017, 10:51 AM) I don...   Oct 9 2017, 10:05 AM
||- - Zero   QUOTE (Pila @ Oct 9 2017, 11:05 AM) No. H...   Oct 9 2017, 12:08 PM
||- - The Franchise   QUOTE (Pila @ Oct 9 2017, 10:05 AM) No. H...   Oct 9 2017, 01:30 PM
||- - nephillymike   QUOTE (The Franchise @ Oct 9 2017, 01:30 ...   Oct 9 2017, 09:44 PM
|||- - The Franchise   QUOTE In your 18 at bats data of hitter vs pitcher...   Oct 10 2017, 12:52 PM
|||- - nephillymike   QUOTE (The Franchise @ Oct 10 2017, 12:52...   Oct 10 2017, 08:20 PM
|||- - The Franchise   QUOTE (nephillymike @ Oct 10 2017, 08:20 ...   Oct 11 2017, 12:58 AM
||- - Rick   QUOTE (The Franchise @ Oct 9 2017, 02:30 ...   Oct 10 2017, 05:22 AM
||- - The Franchise   QUOTE Look at the Yankees/Indians game last night....   Oct 10 2017, 12:59 PM
||- - Rick   QUOTE (The Franchise @ Oct 10 2017, 01:59...   Oct 11 2017, 04:54 AM
||- - The Franchise   QUOTE (Rick @ Oct 11 2017, 05:54 AM) Sorr...   Oct 11 2017, 12:45 PM
||- - Rick   QUOTE (The Franchise @ Oct 11 2017, 01:45...   Oct 11 2017, 04:09 PM
||- - The Franchise   QUOTE (Rick @ Oct 11 2017, 05:09 PM) We...   Oct 11 2017, 04:35 PM
||- - Rick   QUOTE (The Franchise @ Oct 11 2017, 05:35...   Oct 11 2017, 06:42 PM
||- - The Franchise   QUOTE (Rick @ Oct 11 2017, 06:42 PM) And ...   Oct 11 2017, 06:56 PM
|- - Rick   QUOTE (Zero @ Oct 9 2017, 06:51 AM) ...   Oct 10 2017, 05:14 AM
- - Pila   That's right.   Oct 9 2017, 12:18 PM
- - The Franchise   While 4th and 8 attempts have been pleasantly remo...   Nov 5 2017, 04:32 PM
- - Pila   QUOTE (The Franchise @ Nov 5 2017, 10:32 ...   Nov 5 2017, 05:10 PM
- - The Franchise   QUOTE (Pila @ Nov 5 2017, 06:10 PM) Team ...   Nov 5 2017, 05:21 PM
- - Pila   QUOTE (The Franchise @ Nov 5 2017, 11:21 ...   Nov 5 2017, 05:22 PM
> The detailed analysis of 4th and 1, 4th and 8
nephillymike
post Oct 8 2017, 07:16 AM
Post #1


Token Axis Member


Group: Members
Posts: 14454
Joined: 23-April 04



http://www.phillyvoice.com/dislodge-yourse...-8-call-people/


https://www.pro-football-reference.com/play...;order_by=yards


So I went back and checked the article again and low and behold, the guy mentions the 18% comment of another reporter. But then, as most reporters won't do, he does his own detailed NFL only research and tailors a pro football reference report that details all 4th and 8 situations as far back as he could and gets a conversion rate of 32.8%. This is from a reporter who isn't the biggest Pedey fan. The link of the report is above, if it doesn't work in this thread, simply click the link in the article. He used a sample of 314 NFL 4th and 8's since the early 90's.
Go to the top of the page
 
+Quote Post
 
Start new topic
Replies
nephillymike
post Oct 8 2017, 04:51 PM
Post #2


Token Axis Member


Group: Members
Posts: 14454
Joined: 23-April 04



QUOTE (nephillymike @ Oct 8 2017, 07:16 AM) *
http://www.phillyvoice.com/dislodge-yourse...-8-call-people/


https://www.pro-football-reference.com/play...;order_by=yards


So I went back and checked the article again and low and behold, the guy mentions the 18% comment of another reporter. But then, as most reporters won't do, he does his own detailed NFL only research and tailors a pro football reference report that details all 4th and 8 situations as far back as he could and gets a conversion rate of 32.8%. This is from a reporter who isn't the biggest Pedey fan. The link of the report is above, if it doesn't work in this thread, simply click the link in the article. He used a sample of 314 NFL 4th and 8's since the early 90's.


Franchise ??
Go to the top of the page
 
+Quote Post
The Franchise
post Oct 8 2017, 05:39 PM
Post #3


Hall of Famer


Group: Members
Posts: 1339
Joined: 23-November 15



I didn't see that earlier.

What can I say? It's refreshing that he's actually using NFL data instead of college, but it still is flawed data, as it isn't tailored to the teams on the field when making your decision. We simply aren't going to get past that here.

The vast majority of the attempts are in the 4th quarter when down significantly. In many situations the defense is likely playing soft, allowing 8 yards to be picked up to keep the clock running. It's still too simplistic to say with any confidence that it's a '1 in 3' chance, but like I said, at least it's NFL data being used.



--------------------
"If it came down to both teams were even, talent-wise, I think the opponent's team would win if it came down to coaching. Andy Reid got out-coached in a lot of games, man, a lot of big games."
Go to the top of the page
 
+Quote Post
Rick
post Oct 9 2017, 05:15 AM
Post #4


Hall of Famer


Group: Members
Posts: 1527
Joined: 15-September 10



QUOTE (The Franchise @ Oct 8 2017, 06:39 PM) *
I didn't see that earlier.

What can I say? It's refreshing that he's actually using NFL data instead of college, but it still is flawed data, as it isn't tailored to the teams on the field when making your decision. We simply aren't going to get past that here.

The vast majority of the attempts are in the 4th quarter when down significantly. In many situations the defense is likely playing soft, allowing 8 yards to be picked up to keep the clock running. It's still too simplistic to say with any confidence that it's a '1 in 3' chance, but like I said, at least it's NFL data being used.

OMG!!! You will never admit there is something to the numbers no matter how much people show you. Just give it up already.....
Go to the top of the page
 
+Quote Post
Zero
post Oct 9 2017, 05:51 AM
Post #5


Soldier #1654 of the Eagle Empire


Group: Members
Posts: 13908
Joined: 24-April 04



QUOTE (Rick @ Oct 9 2017, 06:15 AM) *
OMG!!! You will never admit there is something to the numbers no matter how much people show you. Just give it up already.....
"There are three kinds of lies: lies, damned lies, and statistics." Numbers can only tell us probability, not guarantee an outcome. If the likelihood of success is 50% then the likelihood of failure is also 50%. When your job is dependent on the outcome I'd think the tendency would lean more to a safer, conservative approach. That said, the benefit of statistics (analytics) may be to expand the definition of "safe" depending on circumstances. I don't think any of the numbers consider either the emotional or physical state of the teams on the field and how that may affect the potential for success or failure on a given play.
Go to the top of the page
 
+Quote Post
Pila
post Oct 9 2017, 10:05 AM
Post #6


Eusbio


Group: Members
Posts: 5623
Joined: 28-April 04



QUOTE (Zero @ Oct 9 2017, 10:51 AM) *
I don't think any of the numbers consider either the emotional or physical state of the teams on the field and how that may affect the potential for success or failure on a given play.
No. How would one go about replicating the exact present conditions if such data can't exist until post mortem? Even in controlled experiments we can't get a droplet of humidity to freeze in the same exact pattern.

But data science when collected and applied with diligent standards can serve as a guard against an array of cognitive follies to an informed application, most specifically in the correlative illusion (correlation vs causation).

Statistics themselves are not necessarily inherently flawed. The inherent flaws lie on the absolutism of the interpreter - i.e. the iron belief that one has a 32% chance of success going on 4th and 8 because the statistics collected tells you that. What you can reasonably deduce from those statistics, even collected with the most prudent standards is that its historic success have been 1/3 probable, and thus not entirely insane to think there's a reasonable chance (relatively speaking, say vs a hail-mary), but that it cautions against it. Ultimately that information can be used in one's own process of intuition without railing off into delusion. And that is ultimately the real purpose of data science.


--------------------
[img]http://dreadart67.files.wordpress.com/2008/03/benfica-2008-ii.jpg[/img]
Go to the top of the page
 
+Quote Post
Zero
post Oct 9 2017, 12:08 PM
Post #7


Soldier #1654 of the Eagle Empire


Group: Members
Posts: 13908
Joined: 24-April 04



QUOTE (Pila @ Oct 9 2017, 11:05 AM) *
No. How would one go about replicating the exact present conditions if such data can't exist until post mortem? Even in controlled experiments we can't get a droplet of humidity to freeze in the same exact pattern.

But data science when collected and applied with diligent standards can serve as a guard against an array of cognitive follies to an informed application, most specifically in the correlative illusion (correlation vs causation).

Statistics themselves are not necessarily inherently flawed. The inherent flaws lie on the absolutism of the interpreter - i.e. the iron belief that one has a 32% chance of success going on 4th and 8 because the statistics collected tells you that. What you can reasonably deduce from those statistics, even collected with the most prudent standards is that its historic success have been 1/3 probable, and thus not entirely insane to think there's a reasonable chance (relatively speaking, say vs a hail-mary), but that it cautions against it. Ultimately that information can be used in one's own process of intuition without railing off into delusion. And that is ultimately the real purpose of data science.
That's pretty much what I said albeit the other side of the coin. Statistics are data, information that adds to personal experience and intuition. An expansion of our potential resources, never a directive.
Go to the top of the page
 
+Quote Post
The Franchise
post Oct 9 2017, 01:30 PM
Post #8


Hall of Famer


Group: Members
Posts: 1339
Joined: 23-November 15



QUOTE (Pila @ Oct 9 2017, 10:05 AM) *
No. How would one go about replicating the exact present conditions if such data can't exist until post mortem? Even in controlled experiments we can't get a droplet of humidity to freeze in the same exact pattern.

But data science when collected and applied with diligent standards can serve as a guard against an array of cognitive follies to an informed application, most specifically in the correlative illusion (correlation vs causation).


I feel I've been completely consistent in my opinion, and I don't think we really disagree on much here. It's more a couple other people demanding I genuflect before a study I've determined isn't gospel.

To your point here, you are correct. Going back to baseball for a second; today, Yu Darvish is pitching against the D-Backs. Chris Iannetta has faced him 18 times. He is 2-16, with 2 walks, and 8 strikeouts. He has no extra base hits, and an OPS of .347. One can easily assume that Iannetta will have a bad day at the plate, and probably shouldn't even get the start.

On the other hand, when looking at the 4th and 8 data used, we're throwing in scores of scenarios where teams are down by a lot late in the 4th, and are successfully converting because defenses are playing soft. We have no examples of this Eagles team taking on that Giants team in a hard fought down. There are ways of getting and applying more relevant data, but that isn't what happened. The problem is exacerbated when throwing in equally flawed data from numbers geeks, such as passing and running efficiency, expected points, etc.


--------------------
"If it came down to both teams were even, talent-wise, I think the opponent's team would win if it came down to coaching. Andy Reid got out-coached in a lot of games, man, a lot of big games."
Go to the top of the page
 
+Quote Post
nephillymike
post Oct 9 2017, 09:44 PM
Post #9


Token Axis Member


Group: Members
Posts: 14454
Joined: 23-April 04



QUOTE (The Franchise @ Oct 9 2017, 01:30 PM) *
I feel I've been completely consistent in my opinion, and I don't think we really disagree on much here. It's more a couple other people demanding I genuflect before a study I've determined isn't gospel.

To your point here, you are correct. Going back to baseball for a second; today, Yu Darvish is pitching against the D-Backs. Chris Iannetta has faced him 18 times. He is 2-16, with 2 walks, and 8 strikeouts. He has no extra base hits, and an OPS of .347. One can easily assume that Iannetta will have a bad day at the plate, and probably shouldn't even get the start.

On the other hand, when looking at the 4th and 8 data used, we're throwing in scores of scenarios where teams are down by a lot late in the 4th, and are successfully converting because defenses are playing soft. We have no examples of this Eagles team taking on that Giants team in a hard fought down. There are ways of getting and applying more relevant data, but that isn't what happened. The problem is exacerbated when throwing in equally flawed data from numbers geeks, such as passing and running efficiency, expected points, etc.


In your 18 at bats data of hitter vs pitcher, I would suggest there is valuable information to be had by using leaguewide data of RHP vs LH batter and vice verse that gives context to the likelihood of success in the matchup you are trying to predict.

The same is true in football.

For example, if I have a data sample of 315 flips of a coin, and I do five other 315 coin flip samples, I may get a range of heads % from 45-55%. Now if you and I go head to head of 18 coin flips and you win 12 of 18, you would be off if you were expecting a 67% success rate going forward. Knowing that a historical likely range in the 45-55% would be valuable information. Same type of deal.
Go to the top of the page
 
+Quote Post
Rick
post Oct 10 2017, 05:14 AM
Post #10


Hall of Famer


Group: Members
Posts: 1527
Joined: 15-September 10



QUOTE (Zero @ Oct 9 2017, 06:51 AM) *
"There are three kinds of lies: lies, damned lies, and statistics." Numbers can only tell us probability, not guarantee an outcome. If the likelihood of success is 50% then the likelihood of failure is also 50%. When your job is dependent on the outcome I'd think the tendency would lean more to a safer, conservative approach. That said, the benefit of statistics (analytics) may be to expand the definition of "safe" depending on circumstances. I don't think any of the numbers consider either the emotional or physical state of the teams on the field and how that may affect the potential for success or failure on a given play.

And this is exactly what I've been saying, the numbers don't lie but they are not perfect. However, they've been using numbers in baseball for DECADES with great success. Why the NFL has resisted looking at numbers for so long has always been a mystery to me.

They are not perfect, they don't accurately predict the outcome but, they can help provide more data to help make a more informed decision.
Go to the top of the page
 
+Quote Post
Rick
post Oct 10 2017, 05:22 AM
Post #11


Hall of Famer


Group: Members
Posts: 1527
Joined: 15-September 10



QUOTE (The Franchise @ Oct 9 2017, 02:30 PM) *
I feel I've been completely consistent in my opinion, and I don't think we really disagree on much here. It's more a couple other people demanding I genuflect before a study I've determined isn't gospel.

To your point here, you are correct. Going back to baseball for a second; today, Yu Darvish is pitching against the D-Backs. Chris Iannetta has faced him 18 times. He is 2-16, with 2 walks, and 8 strikeouts. He has no extra base hits, and an OPS of .347. One can easily assume that Iannetta will have a bad day at the plate, and probably shouldn't even get the start.

On the other hand, when looking at the 4th and 8 data used, we're throwing in scores of scenarios where teams are down by a lot late in the 4th, and are successfully converting because defenses are playing soft. We have no examples of this Eagles team taking on that Giants team in a hard fought down. There are ways of getting and applying more relevant data, but that isn't what happened. The problem is exacerbated when throwing in equally flawed data from numbers geeks, such as passing and running efficiency, expected points, etc.

And again, you totally miss what we've been saying.

In your baseball example, that is exactly what we AREN'T saying. Nobody said Ianetta shouldn't get the start. However, what the numbers say is, he has a good probability of having a bad day against Darvish. But it does not mean he WILL have a bad day.

Look at the Yankees/Indians game last night. The Indians brought back Bauer on 3 days rest to face the Yankees. Bauers' numbers against the Yankees have been ridiculously-good (.9 ERA, etc...). However, his numbers on 3 days rest have also been suspect. The Yankees knock him out of the game early and go on to win. So, which numbers should they have gone with? Obviously, they went with the first set of numbers. They were wrong in this situation.

Again, and I'll say it slowly this time, the numbers don't lie, however, they are NOT going to actually predict the outcome of a specific situation, however, they WILL give you an idea of the probability of an outcome in a given situation. Just so you know, the definition of probability is the likelyhood of something happening or being the case. Note, it does NOT say it is what will actually happen.

The point many of us have been trying to make yet you still continue to argue about is, the numbers are another useful tool, which can (and should) be used to help make a decision. Obviously, other factors should also be used to make a decision in each situation.
Go to the top of the page
 
+Quote Post
The Franchise
post Oct 10 2017, 12:52 PM
Post #12


Hall of Famer


Group: Members
Posts: 1339
Joined: 23-November 15



QUOTE
In your 18 at bats data of hitter vs pitcher, I would suggest there is valuable information to be had by using leaguewide data of RHP vs LH batter and vice verse that gives context to the likelihood of success in the matchup you are trying to predict.


In this case, that would be completely irrelevant as you have a large enough sample size to show individuals against each other. If a pitcher were making his debut, it would matter somewhat if facing a heavy lefty or righty lineup, but since major league pitchers have to face everyone it wouldn't be enough to cancel his start. In a pinch hit situation, everyone in baseball knows opposite hitters have a better chance of picking up breaking balls, so opposite side hitters are used almost every time.

And look! Iannetta didn't get the start, the D-Backs must read this page. He pinch-hit later on when Darvish was out.

QUOTE
The same is true in football.


No it isn't.

QUOTE
For example, if I have a data sample of 315 flips of a coin, and I do five other 315 coin flip samples, I may get a range of heads % from 45-55%. Now if you and I go head to head of 18 coin flips and you win 12 of 18, you would be off if you were expecting a 67% success rate going forward. Knowing that a historical likely range in the 45-55% would be valuable information. Same type of deal.


Every individual coin flip is 50%. The more flips we have, the more towards 50% each it will go. I honestly have no idea why you would use coin flips as an example.


--------------------
"If it came down to both teams were even, talent-wise, I think the opponent's team would win if it came down to coaching. Andy Reid got out-coached in a lot of games, man, a lot of big games."
Go to the top of the page
 
+Quote Post
The Franchise
post Oct 10 2017, 12:59 PM
Post #13


Hall of Famer


Group: Members
Posts: 1339
Joined: 23-November 15



QUOTE
Look at the Yankees/Indians game last night. The Indians brought back Bauer on 3 days rest to face the Yankees. Bauers' numbers against the Yankees have been ridiculously-good (.9 ERA, etc...). However, his numbers on 3 days rest have also been suspect. The Yankees knock him out of the game early and go on to win. So, which numbers should they have gone with? Obviously, they went with the first set of numbers. They were wrong in this situation.


Bauer has had good numbers against the Yankees this year, but over his career against several of their players they are batting over.300 - and no pitcher will consistently do well on 3 days rest. It's worth pointing out he still didn't give up any earned runs last night. The Indians obviously wanted their ace, Kluber, to have the decisive game in Cleveland if Bauer couldn't get it done. He has rock solid numbers against almost all of their hitters, and will likely turn in a great start tomorrow.

The equivalent of what you two have been saying would be if we went back 15 years and looked at data that showed how Jose Mesa did against Derek Jeter, or how many times the '02 Indians won at home against the '02 Yankees. It would be irrelevant, and would contribute nothing to intellectually honest probability.



--------------------
"If it came down to both teams were even, talent-wise, I think the opponent's team would win if it came down to coaching. Andy Reid got out-coached in a lot of games, man, a lot of big games."
Go to the top of the page
 
+Quote Post
nephillymike
post Oct 10 2017, 08:20 PM
Post #14


Token Axis Member


Group: Members
Posts: 14454
Joined: 23-April 04



QUOTE (The Franchise @ Oct 10 2017, 12:52 PM) *
Every individual coin flip is 50%. The more flips we have, the more towards 50% each it will go. I honestly have no idea why you would use coin flips as an example.


Because in statistics, there is a regression towards the mean. In your Darvish example, the real long term numbers are not likely to be those lopsided ones of their statistically insignificant 18 encounters, but more towards the righty vs lefty and vice verse historical statistically significant numbers.
Go to the top of the page
 
+Quote Post
The Franchise
post Oct 11 2017, 12:58 AM
Post #15


Hall of Famer


Group: Members
Posts: 1339
Joined: 23-November 15



QUOTE (nephillymike @ Oct 10 2017, 08:20 PM) *
Because in statistics, there is a regression towards the mean.


In terms of coin flips, yes. But it's ridiculous to think that 18 plate appearances is statistically insignificant, especially when coming up with 2-16. Don't take my word for it - the guy was benched yesterday. Results of individual matchups are no doubt affected by lefty-righty alignments, such as Michael Brantley's abysmal 1-13 against Sabathia, but nothing can compare to individual stats, provided there's a few games worth of at-bats to go off of. He'll probably be used to pinch hit tomorrow as well.


--------------------
"If it came down to both teams were even, talent-wise, I think the opponent's team would win if it came down to coaching. Andy Reid got out-coached in a lot of games, man, a lot of big games."
Go to the top of the page
 
+Quote Post
Rick
post Oct 11 2017, 04:54 AM
Post #16


Hall of Famer


Group: Members
Posts: 1527
Joined: 15-September 10



QUOTE (The Franchise @ Oct 10 2017, 01:59 PM) *
Bauer has had good numbers against the Yankees this year, but over his career against several of their players they are batting over.300 - and no pitcher will consistently do well on 3 days rest. It's worth pointing out he still didn't give up any earned runs last night. The Indians obviously wanted their ace, Kluber, to have the decisive game in Cleveland if Bauer couldn't get it done. He has rock solid numbers against almost all of their hitters, and will likely turn in a great start tomorrow.

The equivalent of what you two have been saying would be if we went back 15 years and looked at data that showed how Jose Mesa did against Derek Jeter, or how many times the '02 Indians won at home against the '02 Yankees. It would be irrelevant, and would contribute nothing to intellectually honest probability.

Sorry but, again, you are ridiculous. Again, you'll never admit you might be wrong even though the whole sports world agrees with what we're saying here.


Pointless trying to explain how things work to you.
Go to the top of the page
 
+Quote Post
The Franchise
post Oct 11 2017, 12:45 PM
Post #17


Hall of Famer


Group: Members
Posts: 1339
Joined: 23-November 15



QUOTE (Rick @ Oct 11 2017, 05:54 AM) *
Sorry but, again, you are ridiculous. Again, you'll never admit you might be wrong even though the whole sports world agrees with what we're saying here.


Pointless trying to explain how things work to you.


You have offered literally nothing intellectual or useful in this entire thread. The 'entire sports world' agrees that going for it on 4th and 8 in that situation is blatantly stupid. It isn't my fault you don't understand basic sports strategy, or statistics, and their applications. Mikey at least comes across as knowing his stuff.


--------------------
"If it came down to both teams were even, talent-wise, I think the opponent's team would win if it came down to coaching. Andy Reid got out-coached in a lot of games, man, a lot of big games."
Go to the top of the page
 
+Quote Post
Rick
post Oct 11 2017, 04:09 PM
Post #18


Hall of Famer


Group: Members
Posts: 1527
Joined: 15-September 10



QUOTE (The Franchise @ Oct 11 2017, 01:45 PM) *
You have offered literally nothing intellectual or useful in this entire thread. The 'entire sports world' agrees that going for it on 4th and 8 in that situation is blatantly stupid. It isn't my fault you don't understand basic sports strategy, or statistics, and their applications. Mikey at least comes across as knowing his stuff.

No, you're back to the, "he shouldn't have gone for it," while we're explaining why he went for it. We've explained why--in sports--they actually look at the numbers and base decisions on them. But you, who is smarter than anyone, continues to claim how things are so different because it backs up what your ridiculous argument has been.

You're the one who doesn't seem to have a grasp of probabilities since you can't seem to get it through your head why they do it. They've been doing it for years in various sports and they're doing it even more in even more sports now. Why? Because it doesn't work?

Mikey knows far more about numbers than I do. I never claimed to be an expert. I just have a little knowledge about probabilities and why people use them to make decisions. You, on the other hand, keep spouting off about college numbers and other things. No matter how much info people provide you, you still keep insisting the numbers don't/can't work when, in fact, you are incorrect.

Then again, this is like EVERY time you've ever argued around here...
Go to the top of the page
 
+Quote Post
The Franchise
post Oct 11 2017, 04:35 PM
Post #19


Hall of Famer


Group: Members
Posts: 1339
Joined: 23-November 15



QUOTE (Rick @ Oct 11 2017, 05:09 PM) *
We've explained why--in sports--they actually look at the numbers and base decisions on them. But you, who is smarter than anyone, continues to claim how things are so different because it backs up what your ridiculous argument has been.

You're the one who doesn't seem to have a grasp of probabilities since you can't seem to get it through your head why they do it. They've been doing it for years in various sports and they're doing it even more in even more sports now. Why? Because it doesn't work?


And I've explained, many times over, why certain numbers matter more in a sport like baseball as opposed to football. Using historical numbers that have nothing to do with the teams on the field is intellectually dishonest and ill advised - such as going for it on 4th and 8 because it's '33%.' I've also pointed out that the numbers you demand I genuflect to would indicate that Tom Brady has the same chance against the Saints defense, as Matt Cassel does against the Chiefs defense - this makes your justification of those studies laughable.

I've also shown simple examples of how individual matchups in baseball make for far more accurate predictions of success or failure - I used Chris Iannetta's history against Yu Darvish as my first example, predicting he wouldn't get the start - and lo and behold, Iannetta was benched! I then used Michael Brantley's history against CC Sabathia as another example, suggesting he shouldn't get the start - and lo and behold, he isn't in the starting lineup tonight!!!

It's almost as if I know what I'm talking about.....

*Bonus example: Francisco Lindor and Carlos Santana both bat over .500 against Sabathia in their careers, though Lindor has a much smaller sample size. Let's see how they do!


--------------------
"If it came down to both teams were even, talent-wise, I think the opponent's team would win if it came down to coaching. Andy Reid got out-coached in a lot of games, man, a lot of big games."
Go to the top of the page
 
+Quote Post
Rick
post Oct 11 2017, 06:42 PM
Post #20


Hall of Famer


Group: Members
Posts: 1527
Joined: 15-September 10



QUOTE (The Franchise @ Oct 11 2017, 05:35 PM) *
And I've explained, many times over, why certain numbers matter more in a sport like baseball as opposed to football. Using historical numbers that have nothing to do with the teams on the field is intellectually dishonest and ill advised - such as going for it on 4th and 8 because it's '33%.' I've also pointed out that the numbers you demand I genuflect to would indicate that Tom Brady has the same chance against the Saints defense, as Matt Cassel does against the Chiefs defense - this makes your justification of those studies laughable.

I've also shown simple examples of how individual matchups in baseball make for far more accurate predictions of success or failure - I used Chris Iannetta's history against Yu Darvish as my first example, predicting he wouldn't get the start - and lo and behold, Iannetta was benched! I then used Michael Brantley's history against CC Sabathia as another example, suggesting he shouldn't get the start - and lo and behold, he isn't in the starting lineup tonight!!!

It's almost as if I know what I'm talking about.....

*Bonus example: Francisco Lindor and Carlos Santana both bat over .500 against Sabathia in their careers, though Lindor has a much smaller sample size. Let's see how they do!

And again, you've demonstrated you have very little understanding of probabilities.

Go to the top of the page
 
+Quote Post
The Franchise
post Oct 11 2017, 06:56 PM
Post #21


Hall of Famer


Group: Members
Posts: 1339
Joined: 23-November 15



QUOTE (Rick @ Oct 11 2017, 06:42 PM) *
And again, you've demonstrated you have very little understanding of probabilities.


I accept your surrender.



--------------------
"If it came down to both teams were even, talent-wise, I think the opponent's team would win if it came down to coaching. Andy Reid got out-coached in a lot of games, man, a lot of big games."
Go to the top of the page
 
+Quote Post

Reply to this topicStart new topic
4 User(s) are reading this topic (4 Guests and 0 Anonymous Users)
0 Members:
 
NFL Football Tickets    football flash games   
RSS Lo-Fi Version Time is now: 18th November 2017 - 10:50 AM