Wednesday, August 26, 2015

Review: Hardball Retrospective (2015)

By Derek Bain

I wonder if Derek Bain has seen the sun in the past few years.

He obviously has put a lot of work into his book, "Hardball Retrospective." It's 435 large pages, and he apparently didn't get much help putting it all together. It's obviously quite a piece of work, and he deserves a great deal of credit for his persistence.

As for the book itself, it's unique ... and will take some explaining first.

Bain went back through baseball history and assigned players to the team that originally signed them. That could include the draft or signings as free agents. Today that would mean international players; in the old days it would be straight signings or purchases from minor league teams. Rob Neyer once wrote a book that had the all-time best signings by position for each team in the majors; it's an interesting list.

Then, Bain magically outlaws trading in baseball and creates major-league rosters for each year. In other words, Nolan Ryan was a Met at the start, so he's placed on the Mets' roster - with the same statistics as he had elsewhere - for the next couple of decades plus. The rosters, by the way, aren't included here.

Then it's time to get out the calculators and computers. Win shares and wins above replacement are totaled for each team, and Bill James' Pythagorean records are calculated. Eventually, Bain came up with season records for each team from 1901 to 2013. I told you it was a lot of work. For example, if only signed players were included, the best team in the American League in 2004 would have been ... the Cleveland Indians. The Boston Red Sox would have gone only 87-75, perhaps because Manny Ramirez was an Indian and David Ortiz would have been a Mariner, instead of the actual 98-64.

Bain also takes the time to look at how teams have drafted over the years, and what clubs do a better job in drafting early and late.

Add it up, and there's plenty of interesting data here. The issue comes with how it is all presented. If you don't pay attention to advanced baseball statistics, this may send you running in the other direction. The book has many tables, numbers and anagrams.

If you can jump past that hurdle, there are several questions that come up along the way. The biggest is how it is presented. Much of the book is a team-by-team breakdown of the numbers, with comments of the highlights of some seasons. (Just asking: Why weren't the teams put in alphabetical order?) For example, Roger Maris' MVP seasons of 1960 and 1961 are mentioned, but in the Cleveland section since that's where he started his career. A question follows - would Maris have duplicated those statistics in Cleveland, in a different ballpark and with different teammates? Tough to say. If he didn't hit 61 homers in 1961, as seems likely, he probably wouldn't have been the MVP either.

That made many of the team comments a little irrelevant, even if they are designed as a "what if?". I think there was a better way to structure the book, by concentrating on the "revised" year-by-year standings. It seems like you could have a lot more fun with that. Interestingly, Bain has started to write up some of those very yearly reviews for websites; a quick search will turn them up.

The other possible flaw with this is that we don't see a year-by-year roster, so we don't know what goes into the year-by-year statistics. Does every player who had more than a cup of coffee in the majors in a given year get assigned to a team? What happens if a roster had 30 such contributors in a given year? (There is a bottom in terms of plate appearances and batters faced, so that teams who don't have many players contributing to the numbers - think expansion teams - aren't counted.) And what happens if, say, five San Francisco Giants outfield "originals" all have big years in the same season? It wouldn't seem fair to credit them all to the Giants, since they couldn't all play at once and some statistics would have to suffer. This issue should have been explained.

"Hardball Retrospective," then, is a difficult book to rate. The concept and effort are fine; it could have been executed in a more interesting way. On the other hand, there's a simple test to determine whether you should buy it if you are a baseball fan. Take a look at the publication, or at least check out the online articles. If you have an interest in it, you'll know right away. My guess is that this will fit a small but appreciate niche in the baseball-loving audience.

Three stars

Learn more about this book.

Be notified of new posts via Twitter @WDX2BB.

1 comment:

  1. Budd,

    Thank you for the review. I would like to respond to several comments and attempt to clarify some of the questions that you posed in the article.

    Regarding the rosters, I wanted to include them but decided to omit them due to space constraints (I retained the “All-Time” Rosters for each team instead). I am planning to include the rosters on my website. I created a Supplemental Statistics, Charts and Graphs page online for this purpose:

    http://www.tuatarasoftware.com/statistics.html

    “Just asking: Why weren't the teams put in alphabetical order?”

    The teams are in alphabetical order by team nickname (Angels, Astros, Athletics, Blue Jays, etc.) I grew up sorting my baseball cards in this manner and never gave it another thought!

    “The other possible flaw with this is that we don't see a year-by-year roster, so we don't know what goes into the year-by-year statistics. Does every player who had more than a cup of coffee in the majors in a given year get assigned to a team? What happens if a roster had 30 such contributors in a given year? (There is a bottom in terms of plate appearances and batters faced, so that teams who don't have many players contributing to the numbers - think expansion teams - aren't counted.) And what happens if, say, five San Francisco Giants outfield "originals" all have big years in the same season? It wouldn't seem fair to credit them all to the Giants, since they couldn't all play at once and some statistics would have to suffer. This issue should have been explained.”

    In the Methodology chapter I attempted to explain how I handled the variance in plate appearances and innings pitched among the teams.
    ... The “original” won-loss record was calculated by normalizing the runs scored and runs allowed, against the league average for that season. This was necessary due to the distribution of plate appearances and innings pitched across the “original” teams (the variance is much closer between “actual” teams, assuming each team played the same number of games during a season).

    Using your example of the Giants outfielders, I calculated the total number of plate appearances and batters facing pitcher for each “original” team, then computed the sum to arrive at league totals. Dividing by the number of teams in the league, I arrived at a league average for PA+BFP. The “original” teams’ WAR and Win Share values were then calculated using a ratio of their total PA+BFP against the league total.

    Sincerely,
    Derek Bain

    ReplyDelete