Any statisticians around? Warning: Math!

[George_Calfas]

PSSSSS

I dont know which is worse, the bad Math or hofautos 12,000 post on this topic :lol:

 

[Guest]

You simply don't understand what I am saying.
I am saying that pWAR is pWAR, it is not a good measurement of greatness.
Greatness is a subjective word.
There is no way to calculate, to suggest that any pitcher is GREATER than any CATCHER, because GREATNESS is SUBJECTIVE.
Don't tell me I can't say that BENCH is one of the GREATEST PLAYERS of all time. Just about EVERY sports writer will agree with ME, not you.

So you say they added points for catcher, obviously the method they used is flawed, because of it's inability to name a great catcher. Again, I ask, how was it able to identify so many "great" pitchers?! My guess is that the author found there was no magic number that they could add to catchers, without putting too many catchers too high. The author was simply not clever enough to seperate out the great catchers. HOF monitor was able to, in creating their formulas. He may be making 6 digits (yawn), but he gave up too soon.

pWAR (which I named and developed) is a tool designed to measure only a player's prime. "Prime" is defined as their five best WAR seasons, as it was found that five seasons was the largest sample size (including HOFers and non-HOFers) to produce the desired result of at least 50 players with an average WAR of 8.0+.

8.0+ WAR was chosen because the develope of WAR set this as the level at which a player reaches MVP Quality.

I feel that if a player was capable of winning "at least" 5 MVP awards they were a 'great' players. Others have different measures, but that is mine.

No catcher in the entire history of the major league has thusfar shown himself to be able to sustain MVP Quality numbers. That is the fault of the player, not the statisticians.

You simply can't randomly inflate a catcher's numbers (they've aleady been scientifically adjusted at the WAR) level to state that they were doing something (winning 5 MVPs) when they obviously were not.

Yes. Johnny Bench was the best at his position, but that by itself does not elevate him above someone lower on the depth chart at anothe position.

If your only complaint with my pWAR construct is that no catcher fit the criteria of the result, then perhaps you need to go back and re-educate yourself on the catcher's of histoy. Instead of listening to stoies about how good someone is ... look at the numbers.

Neither you, nor I, nor anyone in history is capable of looking at a player and saying "He's great. He's good. He's average." We're flawed individuals and a number of flawed, bias, and uninformed beliefs form those opinions. The judgement must be left to the unbias, unvarnished numbers. It is the fair way. It is the correct way.

which again is why a 5 year pWAR is not a defiinitive or conclusive in defining one's greatness.
I bet the author of pWAR would agree as well. He never intended that a 5 year pWAR be used to conclude one's greatness, that was only your false conclusion.
I do like pWAR the more I see it, but only when used for it's intended purposes.
it's biggest fault that I see is in it's inability to rank catchers. They had a seperate formula for pitchers, so the author should create an updated pWAR(v2) that better assesses catchers.

Then to better define greatness (which is SUBJECTIVE), a combination of 5,7,10 pWAR could be used along with bonus points for records and other characteristics that are generally accepted as a measure of one's greatness, in a similar manner as the HOF Monitor does. pWAR 5 is SIMPLY not the answer.


pWAR 5 simply measures the best 5 years with subjective weights given for offense and defense,
and that may have a decent mechanism to measure pitchers, but does not have a good mechanism to measure catchers.

You've made a number of errors in this post and you sound very confused. I'd suggest you re-read the posts more carefully, and take a deep breath before just wildly replying.

The creator of WAR designed it as a way of clearly illustrating a player's value (in terms of wins added) to a team in a given season. These seasonal WARs can be totalled to see how many wins a player added over his career. It has been argued that since the objective is to win the game, the number of wins added by one particular player can be used to measure his quality. 8.0+ has been defined as MVP Quality, 5.0+ has been defined as All-Star Quality, 2.0+ has been defined as a typical Starter, 0.0-2.0 as a typical Reserve player, and a negative number a a typical Replacement player.

pWAR was constructed by me because a single season WAR was too small a sample size to compare players with, and career WAR was invalid because career length is a variable. Therefore, I set about testing an average of a player's seasons, eventually concluding that 5 seasons provided the most desirable results for the reasons posted above.

Do you have the ability to show the top 50 regardless if WAR is less than 8?

I want to see the best 10 year averages regardless if the average is 6 or 7.
Also, are these request easy, i don't mean to ask for a lot of your time if they are not easy for you to spit out.
If easy, i would like to see a 7 year list as well.

Yes. It is possible to generate a "Top 50" list of players WAR-averages for 7 and 10 years. However, I have not yet done so, as my focus has been on 5 years.

Before I go any further I need clarification of something. Did you create the term "pWAR" or did the author of WAR create it?

....so whenever i see pWAR, that is ONLY the best 5 year WAR,
so the "best 10 years" lists you gave are not a 10 year pWAR? they are just the best 10 year WAR.

If you created the term "pWAR", then I would simply suggest a "pWARv2" which would look similar to:

4x(best 5 years WAR) + 2x(next best 2 years WAR) + next best 3 years WAR + magic number for catchers.


PS - I agree, I have not spent any time with WAR prior to this thread. I will investigate and learn about it more, I am very interested.

Yes. I created the term pWAR and the formula used to construct it using the data published by the creato of WAR.

Whenever you see pWAR, it only repesents a 5 year average.

The ten year lists I've povided are a legitimate 10 year average of WAR, however there's no term (such as pWAR) that I've coined for a 10 year average. It was merely posted, by request, for comparison. The complete data on my computer actually has averages for all possible years, as before I could determine the best year for pWAR I had to sample the results.

There is no scientific or mathmatical basis fo the 'formula' you describe above. And there is no "magic number" for catchers. It is not possible for you, or I, or anyone to 'fudge the numbers' to make Johnny Bench look like a legitimate MVP candidate for at least five years.

I want to make this clear because I genuinely don't think you understand. My list of "great players" is a list of playes who were capable of winning at least 5 MVP awards at the peak of their careers. Johnny Bench does not fit that criteria. Joe Mauer is halfway there, and may get there.

PSS, i just read where the p in PWAR means prime, so if you are the author of "pWAR", maybe don't create a "pWARv2", but create a "gWAR", g for greatness.

gWAR = 4x(pWAR) + 2x(next best 2 years WAR) + next best 3 years WAR + magic number for catchers.
(i personally would add bonus points for other characteristics, but i can do that in my head)

or if those numbers would take a lot of work, maybe since career war is already available, try something like this:
gWAR = 4x(pWAR) + 2x(career WAR) + magic number for catchers.

Again. You're just making up numbers with no basis for the reason you're choosing them.

 

[hofautos]

>>I feel that if a player was capable of winning "at least" 5 MVP awards they were a 'great' players. Others have different measures, but that is mine.

Agreed, it is "your" measure of greatness, and others have different measures.

And you even suggest comparitively mediocre pitchers and outfielders are "greater" than the best catchers...even further, you not only suggest who is great, you suggest you can accurately state who is greater than who in a numerical value without any consideration for career numbers, longevity, records set, and without any appreciation of replacement for the catcher.

 

[hofautos]

There is no scientific or mathmatical basis fo the 'formula' you describe above. And there is no "magic number" for catchers. It is not possible for you, or I, or anyone to 'fudge the numbers' to make Johnny Bench look like a legitimate MVP candidate for at least five years.
ME>> I am not trying to make Bench look like a MVP candidate for at least 5 years, I am trying to define greatness based on different criteria.

I want to make this clear because I genuinely don't think you understand. My list of "great players" is a list of playes who were capable of winning at least 5 MVP awards at the peak of their careers. Johnny Bench does not fit that criteria. Joe Mauer is halfway there, and may get there.
ME>> I do understand. I understand your definition of great does not equal my definition of great.

PSS, i just read where the p in PWAR means prime, so if you are the author of "pWAR", maybe don't create a "pWARv2", but create a "gWAR", g for greatness.

gWAR = 4x(pWAR) + 2x(next best 2 years WAR) + next best 3 years WAR + magic number for catchers.
(i personally would add bonus points for other characteristics, but i can do that in my head)

or if those numbers would take a lot of work, maybe since career war is already available, try something like this:
gWAR = 4x(pWAR) + 2x(career WAR) + magic number for catchers.

Again. You're just making up numbers with no basis for the reason you're choosing them.

ME>> I am not making up numbers any more than you are, and there is more logic in my basis than yours.
In your first message in the thread you defined the greatest 52 players based on criteria without any consideration for career numbers.

I am agreeing that 5 years is a good variable to use for an individuals best years, and I am suggesting carreer numbers cannot be overlooked.
I am suggesting that your pvalue is TWICE as important as carreer numbers, but career numbers cannot be overlooked in defining greatness, and to add some value for the deficiency in WARs inability to resolve properly for catchers.

2x PWAR + fraction of Carreer war (such that pwar carries twice this weight) + catcher variable MAKES A WHOLE LOT MORE SENSE than pwar in defining "greatness".

You can say i am making it up, but your making up of "pwar" makes no effort in giving any value for career numbers in defining greatness and WAR is defunct for its bad estimation for catchers.

I think I understand enough now, to agree to disagree.

You stand on PWAR to define greatness, and I stand on it needing adjustment for career numbers and for catchers before you can suggest it can define the greatest.

/thread imho

 

[hofautos]

PSS for Califas

Chevy, do me one last favor since you have this data already.
TOP 100, provide pWAR values, and I will create a gWAR using your pWAR.

I will use a derivative of what the HOF monitor uses for catchers:
(For tough defensive positions, 60 for 1800 games as a catcher, 45 for 1,600 games, 30 for 1,400, and 15 for 1,200 games caught).

(I will use a derivative of career WAR such that pWAR is roughly twice the value of career WAR.)

 

[packbusta]

I don't have much to add other than it has been entertaining to follow this thread. I like the idea of this pWAR, but would like to see a critical analysis on it before I fully recognize it as a significant stat. That being said, I do agree with the concerns of many posters who are saying that a player's greatness, fame if you will, goes beyond just stats.

 

[Guest]

>>I feel that if a player was capable of winning "at least" 5 MVP awards they were a 'great' players. Others have different measures, but that is mine.

Agreed, it is "your" measure of greatness, and others have different measures.

And you even suggest comparitively mediocre pitchers and outfielders are "greater" than the best catchers...even further, you not only suggest who is great, you suggest you can accurately state who is greater than who in a numerical value without any consideration for career numbers, longevity, records set, and without any appreciation of replacement for the catcher.

I don't think there are any players above Bench whom you could define as 'mediocre.'

The goal of baseball is to win games. WAR is an attempt to quantify how many wins each individual player is responsible for.

My pWAR states this in terms of five years. Every player on my list would win you at least one game a season more than Bench. Therefore, I am comfortable placing those players ahead of Bench on a five year ranking.

 

[Guest]

PSS for Califas

Chevy, do me one last favor since you have this data already.
TOP 100, provide pWAR values, and I will create a gWAR using your pWAR.

I will use a derivative of what the HOF monitor uses for catchers:
(For tough defensive positions, 60 for 1800 games as a catcher, 45 for 1,600 games, 30 for 1,400, and 15 for 1,200 games caught).

(I will use a derivative of career WAR such that pWAR is roughly twice the value of career WAR.)

The problem you're making is that you're forming an opinion first.

You state "Johnny Bench is great." "I cannot except statistics that state he is not." "Therefore, I must add a 'magic number' to inflate Bench's numbers to make him appear great in the statistics."

It does not wok like that. You must do the hard math first and view the results, and based on those results form opinions on a player.

You're drawing a conclusion with no statistic merit, and then attempting to force the results.

This is bad. lol

 

[ronfromfresno]

PSS for Califas

Chevy, do me one last favor since you have this data already.
TOP 100, provide pWAR values, and I will create a gWAR using your pWAR.

I will use a derivative of what the HOF monitor uses for catchers:
(For tough defensive positions, 60 for 1800 games as a catcher, 45 for 1,600 games, 30 for 1,400, and 15 for 1,200 games caught).

(I will use a derivative of career WAR such that pWAR is roughly twice the value of career WAR.)

The problem you're making is that you're forming an opinion first.

You state "Johnny Bench is great." "I cannot except statistics that state he is not." "Therefore, I must add a 'magic number' to inflate Bench's numbers to make him appear great in the statistics."

It does not wok like that. You must do the hard math first and view the results, and based on those results form opinions on a player.

You're drawing a conclusion with no statistic merit, and then attempting to force the results.

This is bad. lol

But you did that when shrinking your sample size to 5 years to rightfully include Koufax in your mix. Again pWar is an interesting ranking system but can you seriously say that the top player at a position in all of the history of the game isn't great? What catcher do for a team can't be measured, you would want Gaylord Perry over Bench because he would win a couple more games but what would Bench's replacement do with handling pitchers and how many games would that lose? Can't measure it, so it gets thrown out.

Giambi had a nice stretch but to say he was great overall is off, like I said before your formula will inflate players that aren't great because of the average and small sample size. Sal Bando never lead the league in a meaningful catergory, never batted over .285, or rarely over .260, in a full season. His career average season was 19 HR, .254 BA and 83 RBI. He had almost no stolen bases and his defense was below average at best. He had one maybe two good WAR seasons and when averaged with lower seasons he becomes a top 10 3B. Having a 5 season span of 10, 7, 6, 5, 4 doesn't make a great player but it does average 8 so he must be an all-time great when he only had ONE great season.

 

[hofautos]

>>I feel that if a player was capable of winning "at least" 5 MVP awards they were a 'great' players. Others have different measures, but that is mine.

Agreed, it is "your" measure of greatness, and others have different measures.

And you even suggest comparitively mediocre pitchers and outfielders are "greater" than the best catchers...even further, you not only suggest who is great, you suggest you can accurately state who is greater than who in a numerical value without any consideration for career numbers, longevity, records set, and without any appreciation of replacement for the catcher.

I don't think there are any players above Bench whom you could define as 'mediocre.'

The goal of baseball is to win games. WAR is an attempt to quantify how many wins each individual player is responsible for.

My pWAR states this in terms of five years. Every player on my list would win you at least one game a season more than Bench. Therefore, I am comfortable placing those players ahead of Bench on a five year ranking.

that's funny since you can't win a game without a catcher :lol:
And you are also comfortable to suggest Gaylord Perry's and his 700 games make him MUCH "GREATER" than Johnny Bench and his 2200 games.
Please don't repeat that you are only considering 5 years as defining great. I have heard it, I understand it, I just disagree with it.
Your definition of GREAT is simply different than most.

I am content to agree to disagree. I think both of us at least understands the others position.
In closing, again, I like what this guy has to say.
http://sportsphd.wordpress.com/2010/03/ ... -catchers/

I still would appreciate the top 100 pwar values, as I do see value in them, and regardless if you provide them to me or not, I still thank you for this thread, as I do see great value in what you have offered.

 

[Guest]

PSS for Califas

Chevy, do me one last favor since you have this data already.
TOP 100, provide pWAR values, and I will create a gWAR using your pWAR.

I will use a derivative of what the HOF monitor uses for catchers:
(For tough defensive positions, 60 for 1800 games as a catcher, 45 for 1,600 games, 30 for 1,400, and 15 for 1,200 games caught).

(I will use a derivative of career WAR such that pWAR is roughly twice the value of career WAR.)

The problem you're making is that you're forming an opinion first.

You state "Johnny Bench is great." "I cannot except statistics that state he is not." "Therefore, I must add a 'magic number' to inflate Bench's numbers to make him appear great in the statistics."

It does not wok like that. You must do the hard math first and view the results, and based on those results form opinions on a player.

You're drawing a conclusion with no statistic merit, and then attempting to force the results.

This is bad. lol

But you did that when shrinking your sample size to 5 years to rightfully include Koufax in your mix. Again pWar is an interesting ranking system but can you seriously say that the top player at a position in all of the history of the game isn't great? What catcher do for a team can't be measured, you would want Gaylord Perry over Bench because he would win a couple more games but what would Bench's replacement do with handling pitchers and how many games would that lose? Can't measure it, so it gets thrown out.

Giambi had a nice stretch but to say he was great overall is off, like I said before your formula will inflate players that aren't great because of the average and small sample size. Sal Bando never lead the league in a meaningful catergory, never batted over .285, or rarely over .260, in a full season. His career average season was 19 HR, .254 BA and 83 RBI. He had almost no stolen bases and his defense was below average at best. He had one maybe two good WAR seasons and when averaged with lower seasons he becomes a top 10 3B. Having a 5 season span of 10, 7, 6, 5, 4 doesn't make a great player but it does average 8 so he must be an all-time great when he only had ONE great season.

I never said Giambi was 'great.' He does not meet my personal standard of greatness. Regarding Giambi, I only stated that my data has shown that his prime was better than public opinion would hold given the stigma on Giambi surrounding steroids. Further, I stated that his prime was ahead of that of McCovey and Cepeda, but when the data was expanded over time McCovey surpassed Giambi.

As for Sal Bando, I neve assigned the word 'great' to him. However, in his prime he was a good (All-Star) player at his position and ranks relatively high (top ten) on the depth chart at 3B using pWAR.

Regarding Koufax. The word 'great' was always thrown around when referred to Koufax, howeve I had no data to support that. Looking at Koufax in terms of WAR I saw that he had three 'great' seasons. These three 'great' seasons were enough to alter public perception of him (and his value) in the hobby significantly. I therefore decided to answer the following question: "As Sandy Koufax is judged by fans based solely on his prime seasons, how would Sandy Koufax' prime seasons compare against other player's prime seasons?"

Once I defined prime (not just using Koufax, but the entire Hall of Fame as detailed above) a player's prime was determined to be five seasons. When I entered the data, the results showed that Koufax' prime ranked in the top 10 among pitchers of the modern era (1901-present).

Therefore, with statistical evidence in hand, I was able to conclude that Koufax was, in fact, great in his prime.

That's all pWAR is when it comes down to it. A way of quantifying a player's prime.

If you believe a player should be judged on his best day, then pWAR is for you. If you've been asked, "If you could field a team for just one game..." I'd advise you to draft a pWAR team.

If you believe a playe should be judged by all his days, then pWAR is not for you, and I strongly recommend you consider career WAR as your barometer.

There are many people who choose to think of a player at his best. Many wish to ignore Willie Mays bumbling in the outfield for the Mets and Ken Griffey falling asleep in the Mariners' clubhouse. pWAR quantifies a player's prime for people who wish to focus on that aspect of a player's career.

 

[ronfromfresno]

But Chris your standard of greatness, to include in your collection, is to average an 8+ WAR for five seasons, but that average could include two great years of 10+ WAR that artificially inflates three 6 WAR seasons making that player great by your rankings. Earlier in the thread you stated that you chose 5 years because 7 moved Koufax too far down on the list. My suggestion is to move the average WAR rating higher or extend the time frame. As was shown with the Giambi comparison he was good for a moment but not a career so a long time frame is needed. And with a small sample size you need to increase WAR to 9+ or 10+ to eliminate players who go 12+, 10+, 6, 6, 5 two great years does not make a top ten players out of hundreds and hundreds of players in history.

 

[hofautos]

That's all pWAR is when it comes down to it. A way of quantifying a player's prime

I somewhat agree with above statement if you compare same position players. Definetely don't use it to compare catchers to any other position

If you believe a player should be judged on his best day, then pWAR is for you.
If you believe a player should be judged by all his days, then pWAR is not for you, and I strongly recommend you consider career WAR as your barometer.

Or if you believe as most people do, including myself, that players should be judged by both their prime and their career numbers, then you should use something in between the two or other resources. HOF MONITOR as most people know on this board, is an excellent source, but I personally believe that a combination of pWAR and other resources would be best.

 

[Guest]

But Chris your standard of greatness, to include in your collection, is to average an 8+ WAR for five seasons, but that average could include two great years of 10+ WAR that artificially inflates three 6 WAR seasons making that player great by your rankings. Earlier in the thread you stated that you chose 5 years because 7 moved Koufax too far down on the list. My suggestion is to move the average WAR rating higher or extend the time frame. As was shown with the Giambi comparison he was good for a moment but not a career so a long time frame is needed. And with a small sample size you need to increase WAR to 9+ or 10+ to eliminate players who go 12+, 10+, 6, 6, 5 two great years does not make a top ten players out of hundreds and hundreds of players in history.

Every player who posted a pWAR 8.0+ (5 years) has gone on to have at least a 5.0+ on the ten year model.

Are there players who posted a pWAR between 5.0-7.9 that fell below the 5.0+ theshold in the ten year model? Yes. Of course. However, the only conclusion pWAR has come to regarding those players is that they averaged at least five All-Star Quality seasons and are therefore 'good.' These players are neither a focus of the study, nor will they be part of my collection.

There is no statistical reason to consider the signifcance of pWAR 9.0+ or pWAR 10.0+. Unlike WAR 8.0+ and 5.0+, which were not randomly chosen, but instead were statistically identified by the developer of WAR when every player in the history of the game was profiled.

The 52 players thusfar identified to have a pWAR 8.0+ are great players.

 

[hofautos]

There is no statistical reason to consider the signifcance of pWAR 9.0+ or pWAR 10.0+. Unlike WAR 8.0+ and 5.0+, which were not randomly chosen, but instead were statistically identified by the developer of WAR when every player in the history of the game was profiled.

The 52 players thusfar identified to have a pWAR 8.0+ are great players.

Great, but not the greatest, and definetely not in the order suggested. Collect pWAR if you wish, but don't expect acceptance or following of those as the greatest. At minimum, anyone besides yourself that wishes a collection of the greatest will include Johnny Bench in their collection....

I will have Johnny Bench in place of your Robin Roberts.

The baseball Gods are disappointed in you.

 

[Guest]

That's all pWAR is when it comes down to it. A way of quantifying a player's prime

I somewhat agree with above statement if you compare same position players. Definetely don't use it to compare catchers to any other position

[quote="Chris Levy":2ywwxrzp]
If you believe a player should be judged on his best day, then pWAR is for you.
If you believe a player should be judged by all his days, then pWAR is not for you, and I strongly recommend you consider career WAR as your barometer.
.

Or if you believe as I and most people do, that players should be judged by both their prime and their career numbers, then you should use something in between the two or other resources. HOF MONITOR as most people know on this board is an excellent source, but I personally believe that a combination of pWAR and combination of other resources would be best.[/quote:2ywwxrzp]

If you read the fine print for HOF Monitor you'll see that it is not an accurate tool.

It assigns points for All-Star Game appearances. There were no All-Star Games before 1933, and voting is based on fan bias.
It assigns points for a MVP award. There was no continuous MVP award prior to 1922.
It assigns points for a Cy Young award. There was no Cy Young award prior to 1956.
It assigns points for a Rookie of the Year award. There was no Rookie of the Year award before 1947.
It assigns points for a Gold Glove. There were no Gold Glove awards before 1957.
It assigns points for League Championship Series. There was no League Championship Series before 1969.
It assigns points for Division Series. There was no Division Series before 1981.
It assigns points for leading the league in a statistical categoy. The league was expanded countless times, modifying the conditions of being able to lead the league in a statistical categoy each time.

A tool with this many innaccuracies just isn't reliable.

 

[Guest]

There is no statistical reason to consider the signifcance of pWAR 9.0+ or pWAR 10.0+. Unlike WAR 8.0+ and 5.0+, which were not randomly chosen, but instead were statistically identified by the developer of WAR when every player in the history of the game was profiled.

The 52 players thusfar identified to have a pWAR 8.0+ are great players.

Great, but not the greatest, and definetely not in the order suggested. Collect pWAR if you wish, but don't expect acceptance or following of those as the greatest. At minimum, anyone besides yourself that wishes a collection of the greatest will include Johnny Bench in their collection....

I will have Johnny Bench in place of your Robin Roberts.

The baseball Gods are disappointed in you.

Science and math have destroyed all gods.

Can you give me any statistical reason to indicate why you feel Johnny Bench was a great player? Thus far, you haven't. You are just making an illogical conclusion that the best at a given position must automatically be assumed to be 'great.' You have given no evidence to support such a conclusion. He was the best catcher, but not great.

As for Robin Roberts? What statistical reason do you have for dropping him? Given your past post history of blindly jumping to conclusions with no evidence, no research, and no undertanding I'd venture you have not. He was the best pitcher of his era. His numbers are simply off the chart.

 

[hofautos]

If you read the fine print for HOF Monitor you'll see that it is not an accurate tool.

at least it is reliable enough to put in the disclaimer for fools. If WAR doesn't have such a statement, it is even less reliable for not doing so.

It assigns points for All-Star Game appearances. There were no All-Star Games before 1933, and voting is based on fan bias.
A tool with this many innaccuracies just isn't reliable.

You really crack me up. I don't know all of either systems calculations, but i know It makes amends for older players in other ways.
Whatever formula HOF monitor uses, its pretty amazing how almost all players above a hof monitor 130 is in the hof and almost all players below 100 are not in...
I don't think WAR or pWAR in it's current state can predict such accuracy with such a small margin of difference.

 

[hofautos]

[quote="Chris Levy":1a24lalp]

There is no statistical reason to consider the signifcance of pWAR 9.0+ or pWAR 10.0+. Unlike WAR 8.0+ and 5.0+, which were not randomly chosen, but instead were statistically identified by the developer of WAR when every player in the history of the game was profiled.

The 52 players thusfar identified to have a pWAR 8.0+ are great players.

Great, but not the greatest, and definetely not in the order suggested. Collect pWAR if you wish, but don't expect acceptance or following of those as the greatest. At minimum, anyone besides yourself that wishes a collection of the greatest will include Johnny Bench in their collection....

I will have Johnny Bench in place of your Robin Roberts.

The baseball Gods are disappointed in you.

Science and math have destroyed all gods.

[/quote:1a24lalp]

Science and math?
Please explain to me the science in pWAR that you propose names the 52 greatest players, and how it can define several pitchers (who play only one in 5 games), but not one catcher. ::facepalm::

that is not science.

 

[Guest]

If you read the fine print for HOF Monitor you'll see that it is not an accurate tool.

at least it is reliable enough to put in the disclaimer for fools. If WAR doesn't have such a statement, it is even less reliable for not doing so.

[quote="Chris Levy":zmk1i263]
It assigns points for All-Star Game appearances. There were no All-Star Games before 1933, and voting is based on fan bias.
A tool with this many innaccuracies just isn't reliable.

You really crack me up. I don't know all of either systems calculations, but i know It makes amends for older players in other ways.
Whatever formula HOF monitor uses, its pretty amazing how almost all players above a hof monitor 130 is in the hof and almost all players below 100 are not in...
I don't think WAR or pWAR in it's current state can predict such accuracy with such a small margin of difference.[/quote:zmk1i263]

...

That is my initial reaction.

Do you know why "almost all" players with a HOF Monitor above 130 are in, and "almost all" players below 100 are not?

Because when they looked at the values produced by HOF Monitor they plotted all of the HOF players, and based on distribution set 100 as significant.

They selected a number that would reflect the maximum number of Hall of Famers on the positive side of the scale.

You have no idea how any of this works, do you? Honestly.

Do you really think someone just sat there and said, "Hmm. Yeah. 100 sounds 'good'?"

This makes my head hurt. LOL