Bill James' Win Shares Book

[Tyrus4189Cobb]

I hope to keep this out of the "baseball literature" forum because I'd rather hear feedback from the statheads. I, myself, am a self-proclaimed stathead novice. Always looking to gobble up more baseball info, especially on defense, how is James' Win Shares book? Is it worth the read? Do you ever use Win Shares, or do you dislike the metric?

[SamtheBravesFan]

I have it. It's some essays, some formulas, and a lot of results tables. It's kind of a dry read. However, it is important in that this was one of the biggest steps forward in evaluating baseball players. Generally, I'm good with numbers, but it seems those who really delve into them would enjoy this book the most.

Its' best use, if you don't like the advanced numbers, is as a companion book to the New Bill James Historical Baseball Abstract.

[BigRon]

I have it. It's some essays, some formulas, and a lot of results tables. It's kind of a dry read. However, it is important in that this was one of the biggest steps forward in evaluating baseball players. Generally, I'm good with numbers, but it seems those who really delve into them would enjoy this book the most.

Its' best use, if you don't like the advanced numbers, is as a companion book to the New Bill James Historical Baseball Abstract.

I echo Sam's coments. It was an extremely important work in advancing evaluation of players- possibly THE most important.

Concerning defense- who knows? Personally, I'm extremely skeptical of WAR's defensive evalautions. In general, James' evaluations seem to be somewhat more in line with what I've heard/read/seen. But, remember, this work was done more than 10 years ago so it is at least possible that some more of the recent metrics supercede it.

[Ubiquitous]

It's a dull book about an evolutionary dead end of a stat. If you want to read about Win Shares and have the book actually be interesting you're better off reading The New Historical Abstract.

Ubermetrics had been around for a very long time before Bill published his book and he really didn't bring anything new to the table with the book.

[Jackaroo Dave]

It's a dull book about an evolutionary dead end of a stat. If you want to read about Win Shares and have the book actually be interesting you're better off reading The New Historical Abstract.

Ubermetrics had been around for a very long time before Bill published his book and he really didn't bring anything new to the table with the book.

Ubi, could you elaborate?

In the Historical Abstract II, I got the impression that making all the individual contributions come out equal to all the team wins was an innovation. Also, if he brought nothing new to the table, how could he create a dead end in an evolving and growing body of knowledge? Do you see the paradox? I didn't express it clearly, but if Bill James is just going along with the flow, how come he wound up in a backwater?

It would seem he'd have to make some counterproductive innovations in order to reach a dead end.

(I would dearly love to see a conceptual timeline tracing the innovations and developments of sabremetrics, if anyone's not too busy.)

[Dude Paskert]

It's a dull book about an evolutionary dead end of a stat. If you want to read about Win Shares and have the book actually be interesting you're better off reading The New Historical Abstract.

Ubermetrics had been around for a very long time before Bill published his book and he really didn't bring anything new to the table with the book.

I'd recommend to read the long essay in the Historical Abstract first and then decide if you want to move on to the Win Shares book. I like reading dense stuff with a mathematical bent, but could never convince myself to read the Win Shares book...it looked that boring to me. I find the HA to be a brilliant book that I constantly pull off the shelf to browse through.
However, my impression was that the whole WAR thing grew out of Win Shares, at least in inspiration. Sounds like Ubi is saying I'm dead wrong on that.

[BigRon]

However, my impression was that the whole WAR thing grew out of Win Shares, at least in inspiration. Sounds like Ubi is saying I'm dead wrong on that.

I agree with that comment. Maybe there are those knowledgableenough to contradict it authoritatively, but I believe Win Shares was at the least the philosophical base for WAR, if not the mathematical base.

I have always contended that WAR became popular because it is FREE and ACCESSIBLE to users like BBFers. To use Win Shares- at least most of the historical stuff- you had to actually buy James' book and PAY MONEY! That's the kiss of death in this day and age.

[Ubiquitous]

WAR isn't based on how many wins the team has like Win Shares does. WAR is also based on what the average player does. Before win shares there was BPro's WARP which stands for Wins Above Replacement Player and WAR is basically WARP but with slightly different mechanics involved.

[EdTarbusz]

I tried reading the Win Shares book after reading the Second Historical Abstract and put it aside pretty quickly. I've never thought that baseball statistics were that interesting and I couldn't muster up the interest to keep reading Win Shares.

[Railsplitter]

To use Win Shares- at least most of the historical stuff- you had to actually buy James' book and PAY MONEY! That's the kiss of death in this day and age.

I hate to plug my own website, but it has all historical Win Shares data from 1871 - 2012. http://seamheads.com/baseballgauge/index.php

The numbers are slightly different than in the book, partly because we have more data now than we did in 2001. Also, I added a decimal point for more accuracy (I was never a fan of rounding to the nearest whole number).

[BigRon]

I hate to plug my own website, but it has all historical Win Shares data from 1871 - 2012. http://seamheads.com/baseballgauge/index.php

The numbers are slightly different than in the book, partly because we have more data now than we did in 2001. Also, I added a decimal point for more accuracy (I was never a fan of rounding to the nearest whole number).

Thanks very much. Your site is the first place- other than James' book- where I have seen all historical WS info.

[brett]

WAR isn't based on how many wins the team has like Win Shares does. WAR is also based on what the average player does. Before win shares there was BPro's WARP which stands for Wins Above Replacement Player and WAR is basically WARP but with slightly different mechanics involved.

Isn't win shares above bench just basically WARP with different mechanics?

[Jackaroo Dave]

Isn't win shares above bench just basically WARP with different mechanics?

If I understand the discussion so far, I don't understand how it could be:


1 WARP and WAR are essentially mechanical variants.

2. WAR does not take into account team wins, unlike win shares (and by inference win shares above bench).

3. So win shares above bench and WARP are more than mechanical variants (unless taking/not taking team wins as your baseline is considered a mechanical variation).

However. I recall reading on another thread that in the final summing up, team WAR works out to be the sum of individual players' WAR, so maybe they are mechanical variants after all?

Little help, please? Thanks!

[Railsplitter]

I would think that a team's total WAR (B-Ref's version at least) would be closer to a team's pythagorean wins than it would actual wins

[Dude Paskert]

I would think that a team's total WAR (B-Ref's version at least) would be closer to a team's pythagorean wins than it would actual wins

This doesn't sound right to me, as a team of replacement level players wouldn't win zero games.