|
The New Bill James Historical Baseball Abstract
|
List Price: $26.00
Our Price: $17.16
Your Save: $ 8.84 ( 34% )
Recent Price Activity
Availability: Usually ships in 24 hours
 *Eligible orders over $25 ship free
Manufacturer: Free Press
|
Average Customer Rating:  
|
|
Binding: Paperback
Dewey Decimal Number: 796
EAN: 9780743227223
ISBN: 0743227220
Label: Free Press
Manufacturer: Free Press
Number Of Items: 1
Number Of Pages: 1008
Publication Date: 2003-05-06
Publisher: Free Press
Studio: Free Press
|
|
|
|
|
|
Editorial Reviews:
|
|
When Bill James published his original Historical Baseball Abstract in 1985, he produced an immediate classic, hailed by the Chicago Tribune as the "holy book of baseball." Now, baseball's beloved "Sultan of Stats" (The Boston Globe) is back with a fully revised and updated edition for the new millennium. Like the original, The New Bill James Historical Baseball Abstract is really several books in one. The Game provides a century's worth of American baseball history, told one decade at a time, with energetic facts and figures about How, Where, and by Whom the game was played. In The Players, you'll find listings of the top 100 players at each position in the major leagues, along with James's signature stats-based ratings method called "Win Shares," a way of quantifying individual performance and calculating the offensive and defensive contributions of catchers, pitchers, infielders, and outfielders. And there's more: the Reference section covers Win Shares for each season and each player, and even offers a Win Share team comparison. A must-have for baseball fans and historians alike, The New Bill James Historical Baseball Abstract is as essential, entertaining, and enlightening as the sport itself.
|
|
|
Spotlight customer reviews:
|
Customer Rating: 
Summary: Insightful and Entertaining, but Win Shares needs an overhaul
Comment: I'm a long-time fan of Bill James and purchased this book several years ago. It has the analysis, insights, passion, reverence, and irreverence that are trademarks of his work. Recently I purchased the Win Shares book, admittedly years after it was published. Since the concept of Win Shares is the underpinning for the ratings in this book, I'd like to share some of my comments regarding this approach.
1. As others have mentioned, the insistence of James of allocating 52% of Win Shares to defense (pitching and fielding) and 48% to offense is arbitrary and yields distorted Win Shares values for the players. It would seem to me to be logical to assign the proportion of Win Shares to offense based on the relative value of the team's offense to the defense. For example, this formula could be used to evaluate the percentage of Win Shares that goes to the offense. A/(A+(2*B)-C) where A is runs scored (adjusted for park factor), B is average team runs scored for the league, and C is runs allowed (adjusted for park factor). Using this technique, the Blue Jays in 2008 have a formula that resolves to 721/(721 + (2*775)-616), or rounded to 44% . Therefore 44% of the Blue Jays Win Shares of 2008 would go to their offense and 56% to their defense. Using the same formula for the Texas Rangers of 2008 yields 61% of Win Shares that go to offense and 39% to defense. It seems reasonable to assign the values this way because the Blue Jays clearly won more games because of their pitching, and the Rangers won more games because of their hitting. Of course, this will make the Win Shares of individual players more accurate as well as the Ranger's hitters, for example, clearly deserve more than 48% of the team's Win Shares. James at one point mentions that he wouldn't want the percentage of pitching Win Shares to go above 58%, because the effect of that could cause some offensive players to have negative Win Shares. But that seems to be pointing out a flaw in the Win Shares offensive calculation rather than a rationale for keeping the pitching share at around 52%.
2. For middle infielders, a percentage of the evaluation is based on assists, which makes sense. There is an "expected number of assists" for a shortstop(or second baseman) that is based on a) the proportion of assists that typically go to a shortstop on a league basis, as well as b) an adjustment for the number of inning a left-handed pitcher was on the mound. Then the actual number of assists for a player is evaluated against the expected number giving the Assists Scale. This is good as a start, but the formula would seem to be inaccurate to the extent that your fellow fielders are either quite good or quite bad. For example, take 2 shortstops from two different teams with the exact same fielding ability. Player A has a great fielding second-baseman and third-baseman next to him. Player B's fielding counterparts are sub-par. Using the formula, it would seem to me that Player B would end up with a higher number on the Assists Scale merely because his second baseman and third baseman are waving at balls that Player B's fellow fielders would field safely. Player B would then have more opportunities to make plays himself. Therefore Player B would end up with a higher percentage of his team's assists and therefore look better on the Assists Scale, despite being no better than Player A. Fielding has a zero-sum aspect to it that makes it hard to evaluate a players' assists (or putouts) in isolation. If a system, such as this one, only look at plays made (such as assists), then we are trying to extract opportunity (plays that could be made) from outcome(actual assists), which is a futile task, it would seem to me. Using the "expected number of assists" described above goes part of the way to show opportunity, but does not take into account the fellow fielders, as I mentioned. I don't have a solution to this, but it is a limitation of the formula.
3. In evaluating player ratings, James evaluates players by a combination of total Win Shares and Win Shares during a player's peak years. A Win Share value for a player of (for example) 30 consists of perhaps the first 20 points consisting of value below replacement level, then perhaps the next 5 points are above replacement but below average, and the final 5 would be above average. A player who plays a longer career would tend to have more of their Win Shares consisting of value below average and below replacement value points. Therefore if you compared two similar players on lifetime Win Shares, the one with the longer career would have more Win Shares even if he had less value above replacement value for his career. James also uses Win Shares during a player's peak years on the theory that we tend to evaluate players on their best years rather than their total career. He says that by using a percentage (I forget what it is) of the peak Win Shares value with a percentage of the lifetime Win Shares for a player, we get the best of both worlds. But I think that using a formula that includes the peak Win Shares merely mitigates some of the problem with total Win Shares, specifically the fact that players with longer careers get overvalued. I think if there was a way to extract the portion of Win Shares that is above replacement level, then that value could be totaled for each season, and the new statistic of Total Win Shares Above Replacement Level (TWSARL?) could be used as the player rating. The combination of total Win Shares and peak win shares is not as coherent to me, as it combines two different things, as well as having the limitation I mentioned.
I'm still getting to understand the Win Shares approach, but these are my initial impressions. Win Shares is an ambitious, worthy idea, and James' implementation and formulas are quite impressive. As he says, it turns the usual method of player evaluation upside-down, and puts players contributions in context of the team, which has been more naturally understood in other team sports like football and basketball. I think, however, that his approach caused some unexpected compromises and rationalizations to be made to pull it all together, which is implied by my comments.
I must say again that the New Historical Baseball Abstract is another ground-breaking and essential book from Bill James , despite my reservations on Win Shares
Customer Rating:
Summary: The essential baseball book
Comment: If a baseball fan were to be stuck on a deserted island with only one book, this should be it. James here is at his best, with history, statistics and analysis presented in his unique manner. This book can be read for long stretches, but its' format makes it perfect for grabbing a few minutes here and there. The only problem is that is was published in 2001; it would be wonderful to have updated player rankings based on what has occurred since then.
Customer Rating:
Summary: An awesome book to have on your shelf
Comment: I have probably read this book 100 times. I still have the original hard cover and it is looking pretty worn, but it is one of my favorite things on my shelf. So, be warned, if you don't like books that you'll thumb through constantly, this book isn't for you.
James goes through the history of baseball in a decade-by-decade format, listing the best teams, players, and lots of interesting tidbits. Then he goes into the player rankings, #1 through #100 at each position. It will make more sense if you've read Win Shares, but honestly, even if you haven't, you'll be fine. Sometimes he doesn't give any explanation for why a player ranks where he does, other than their stats at the end of the section for that position. But even so, it's a great read.
James also includes a (controversial) section on the top 100 players of all time, with explanations for why each player ranks where they did.
If you like learning about the history of this great game, or just want to discover some players you've never heard of, this book is for you.
Customer Rating:
Summary: Bill James Has Completely Lost It.
Comment:
Bill Jame's 100 greatest players the NEW list starts around page 358 and reaches peak idiocy on page 360 where he explains that Lou Gehrig wasn't in the top ten because if he and Ruth were so good why did they only win 4 pennants in the 10 years they were teammates. What a NIMROD.
Then HE DOESN'T PICK ONE CATCHER IN THE TOP 40.
Even worse and probably the most heinous change was the move of Warren Spahn from TENTH, now get this, to 36th. Spahn won his first game at the age of 25. He won 363 games. He won 20 games 13 times.
__________________
Customer Rating:
Summary: the man and the work that put sabermetrics on the map
Comment: Bill James is famous for his ability to collect, publish and analyze statistics about baseball. This is the second edition of his history book covering through the entire 20th century. But as James says in his preface this is more than just an update. In reviewing the first book he found that he didn't like a number of things that he did and so he has changed. Some may think for the better others for the worse but in my case I never read his 1980s edition so I have no basis for comparisons.
James is not a professional statistician but has good statistical intuition and is respected by professional statistician who specialize in sports statistics.
James covers the rules of the game and is very detialed about the players and the rule changes and strategy changes. What I enjoyed most about the book was his lists of the all time top 100 players at each position. This is something sports statisticians think about often and using statistical adjustment techniques and Bayesian methods professional statistician like Schell and Berry have written articles and in Schell's case a book on how to do this. Schell's book includes a list of the all time greatest hitters with Tony Gwynn at the top. The book tells you how the list is constructed and teaches statistical methods along the way.
James has no formal statistical method for constructing his lists. At each position he ranks the top 100 players and does a good job of mixing the old timers with the present day players. Though subjective, this is a difficult task for anyone and James is one of the few who knows enough detail of the history and players in baseball to be up to the task. I may not agree with all of his rankings but that is part of what makes talking about baseball fun. James provides descriptions of the players on his list that may be thought of as justification for their inclusion or rank.
The list of number 1 players by position is as follows:
1. catcher - Yogi Berra
2. pitcher - Walter Johnson
3. 1st base - Lou Gehrig
4. 2nd base - Joe Morgan
5. shortstop - Honus Wagner
6. 3rd base - Mike Schmidt
7. left field - Ted Williams
8. center field - Willie Mays
9. right field - Babe Ruth
The American Statistical Association formed a section SIS (Statistics in Sports). I am a member and so are many other statisticians including Carl Morris, Hal Stern, Mike Schell, Jim Albert, Jay Bennett and Scott Berry. We all have the common ground of interest in sports (particularly baseball). The introduction of true statistical methods in sports has turned sports partly intp a science. Mike Schell wrote a statistics book about statistical adjustment of individual player statistics based on the effect of the home ball park. Albert and Bennett have also contributed books. Efron and Morris long before this movement was in full force wrote a major statistical paper for the Journal of the American Statistical Association that used predicting baseball player averages using Stein shrinking estimator (an Empirical Bayes estimator).
It is books like this that amass large amounts of baseball data and use baseball knowledge and common sense ot look at the game in a differnt way.
|
|
|
|
|
|
|
|
