Batter's Box Interactive Magazine Batter's Box Interactive Magazine Batter's Box Interactive Magazine
At noontime on Saturday a small crowd of baseball enthusiasts gathered in the theatre of the SkyDome for a book talk by Alan Schwarz. The crowd was diverse, ranging from SABR members to a man with his two young children to a Blue Jays employee who is also attending the 34th Society for American Baseball Research (SABR) convention. Schwarz’s book talk was sponsored by the Learning Annex and centered on his acclaimed book The Numbers Game. Schwarz also writes a weekly column for www.espn.com, a biweekly article for The New York Times and is a senior writer for Baseball America. The Numbers Game was his first book and received much acclaim, as it was named the best baseball book in 2004 and nominated for the 2004 Dave Moore Award, handed out by SABR to the “most important book on baseball” published in that year. It has recently been released in paperback.

The lecture was very informal, due in part to the revised start time of the game, which meant that some attendees did not arrive until 12:30. Schwarz casually talked with those who had arrived for 12:00 while he arranged his documents so he was organised for the talk. The discussion touched on several different topics, from Schwarz’s column the next day (June 19th) in the Times about whether star pitchers draw bigger crowds to Schwarz’s original design for the cover, which differed significantly from the hardcover cover design. Interestingly, Schwarz concluded that there were slightly discernable differences in attendance figures for some star pitchers, but he also speculated that this may not result in increased attendance overall, but rather that fans who planned to go to a certain game that week deliberately chose the one with the best pitching match up.

There was a significant amount of overlap between The Numbers Game and the content of the lecture, but with the informal setting Schwarz was able to expand upon some of the sections and respond to the mood and questions of the audience. The lecture began around 12:30 as Schwarz opened by talking about the earliest boxscores. With a quick mention about the myth of Abner Doubleday, Schwarz proceeded to explain how the earliest boxscores contained only two categories, runs and hands lost. Early baseball had connections to cricket and many of the earliest baseball pioneers were Englishmen, so this can be traced back to cricket, Schwarz explained. In cricket you either score a run or make an out each time up to bat (or you hit the ball and simply remain up to bat, choosing not to run), so in baseball the same categories were tracked. As you may have guessed, this completely ignores batters who are stranded on base after a base hit.

However, this did not last for too long. Henry Chadwick’s role in developing the earliest baseball statistics are detailed in the book’s first chapter. It’s particularly interesting to see how early baseball pioneers decided to quantify statistics, determining what events were worth keeping track of. The batting champion was then determined by Hits per Game, but within five years that was modified. People had quickly realised the inherent inequality of Hits per Game, as those at the top of the lineup got more at-bats, and thus more chances for bases hits, than those at the bottom. Interestingly, a citizen in Washington wrote to Chadwick and suggested that they divide hits by the number of hits by the number of at-bats. This change was adopted and this statistic (known as batting average) has remained one of the cornerstones of baseball ever since.

Schwarz used a interesting demonstration to explain some of the shortcomings of batting average, as he took a pile of coins and handed them a 10-year-old boy in the audience and asked him to count the money. The boy arrived at a total of $4 and change and Schwarz prompted him to make the connection to batting average and slugging percentage. The boy didn’t total the money by adding the number of coins, but rather he summed the value of the coins. Therefore, why did it make sense to value all hits equally when some are clearly more valuable than others? As Schwarz described how batting average neglected to measure one’s ability to get on base or hit for power he briefly described how on-base percentage and slugging percentage were adopted as official statistics.

Schwarz then went onto talk about a couple of Canadians who played large roles in the advance of statistics. The first was Allan Roth, whose debut with the Dodgers on April 15, 1947 was overshadowed by the man who started at second for the Dodgers that day. Roth was a Montreal native who tracked statistics for the Montreal Royals, the farm team for the Dodgers. Roth later kept track of statistics for the Dodgers, including RBI percentage (batting average with runners in scoring position) and their batters splits against right-handed and left-handed pitchers. Roth approached Branch Rickey with these numbers, which had never before been seen by major league front offices. Rickey was intrigued and saw how these advanced statistics could be of benefit to him in making transactions and hired Roth.

Roth introduced innovations such as charting each player’s spray zone, which allowed him to conclude that Dixie Walker’s bat speed was slowing done because he was hitting everything to left field. The Dodgers, abiding by Rickey’s maxim that it was better to trade someone a year too early rather than a year too late, traded Walker to the Pirates for Preacher Roe and Billy Cox. Roth spent several years as Rickey’s assistant, before Rickey left for Pittsburgh. Roth didn’t follow him, probably because he didn’t want to uproot his family. Unfortunately, Roth’s ideas were never taken very seriously by the new management after Rickey left and his influence quickly faded. He then went on to provide numerous statistics for up-and-coming baseball announcer Vin Scully and gave him access to all sorts of statistical information that other baseball announcers did not have.

Schwarz had a copy of an infamous article from Life magazine entitled “Goodbye to Some Old Baseball Ideas,” which described how Rickey and Roth were engaging in a statistical revolution never before seen in baseball. Schwarz circulated a photocopy of the article for the audience to browse through and read. At this point in history Life was an incredibly popular publication and they had an 11-page article devoted to baseball statistics, which spread the advancement of statistics in baseball decision-making to a whole new audience. The article’s feature picture showed Rickey pointing to a line of statistics which Schwarz termed “gobbledygook.” I’m not sure there’s even evidence Rickey took the formula very seriously. However, within the offensive part of the formula there was a section of the formula that equalled on-base percentage and another part that equalled slugging percentage without singles. The pitching part of the formula contained something that equalled opposing batting average, another which measured control and another which emphasised strikeouts. Whether or not Rickey ever used his complicated formula is one matter, but the fact that he was calculating what are recognised today as very important aspects of the game demonstrates how far Rickey and Roth’s Dodgers were ahead of their time with regards to statistical analysis.

Schwarz then spoke about another Canadian who made huge advances in baseball statistical analysis, even if very few people were ever exposed to his work. In fact, Schwarz called him his favourite character encountered while researching the book. George Lindsay, an Ottawa native, worked for the Canadian Department of National Defence modeling flight patterns and tactical strikes. However, he found combat too fluid to engage in serious analysis as the data was often incomplete. Lindsay turned his attention, at least in his free time, to another one of his loves, baseball, and decided to examine strategic decision-making in that field. Lindsay spent evenings stationed in bases in places from Quebec to Italy scoring baseball games, eventually scoring over 1,200 games. Lindsay spent countless hours interpreting the data from the games he scored, in addition to several hundred scored by his father. Lindsay reached several conclusions from his data, but was only able to publish the results in semi-obscure military journals. For example, in one study Lindsay was able to conclude that a batter facing a pitcher of the opposite-handedness had a higher batting average by 32 points, a fact which he believed would help managers making in-game decisions.

Lindsay also came up a run probabilities chart, which showed how many runs were likely to score in certain situations. For example, with the bases loaded and one out a team scored an average of 1.64 runs and with a runner on second and one out .67 runs usually scored. In a 24-page report that was published in the journal Operations Research Lindsay wrote about which baseball strategies made sense and which didn’t. He concluded that the sacrifice bunt was overrated as a strategy and only made sense late in the game when the team needed only one run. Lindsay determined how often one needed to steal bases successfully to make it worthwhile and also wrote about strategies like playing the infield in and the intentional walk.

Later Lindsay would write another article that explained the concept of the relative value of leads, although this article wasn’t published. Work in this field, often dubbed win expectancy, would later be done by the Mills brothers and has been written about at The Hardball Times by Dave Studeman. Lindsay was able to determine the frequency with which teams would win when they were ahead by a certain number of runs at a certain juncture in the game (such as ahead by 2 after five innings; behind by 3 after seven). Lindsay’s work never made him famous or a household name in baseball circles, but it was the first time someone had used play-by-play data to weigh baseball strategies. Schwarz actually had several of Lindsay’s original scoresheets and passed them around the audience. It was quite interesting to see the scoring system Lindsay and his father developed, which allowed them to score over a thousand games and track the data for all the different situations.

Vernon Wells was originally scheduled to make an appearance in the lecture, to give some insight into how baseball players deal with and relate to statistics. However, this was the weekend of the birth of his second child and Wells returned to Arlington to be with his wife. Luckily, J.P. Ricciardi agreed to spare about twenty minutes from his busy day and answer a few questions. Schwarz asked J.P. some questions to begin with, such as how J.P. used statistics to evaluate players. The conversation evolved to talking about pitchers, as J.P. revealed that strikeouts are a very important tool to be used when evaluating pitchers and can take precedence over ERA in many cases.

Some seemed surprised by that revelation, but J.P. pointed to the example of Matt Clement, who struck out a ton of players last year and who the Jays pursued aggressively in the offseason. JP also mentioned that the Jays target groundball pitchers because of their home park, but that they don’t follow that policy stringently, as he mentioned Ted Lilly as an exception who the Jays felt could succeed in Toronto.

An audience member (who may or may not write for Batter’s Box) asked JP how confident he feels in using statistics to evaluate defence. J.P. replied that he still feels like the best way to evaluate defence is with your naked eyes. “Keith’s always trying to sell me on the defensive statistics,” J.P. said, but added that he didn’t use them often, if ever. He mentioned Derek Jeter as the whipping boy for those who use stats for defensive evaluations, but said there’s “no one” he’d rather have fielding a ground ball for his team in the bottom of the ninth.

Schwarz then returned to the probabilities of scoring chart and asked J.P. about some of the in-game strategies the Blue Jays practice. J.P. admitted that he couldn’t quote you the numbers on the sheet, but that he was familiar with the conclusions it allows one to draw. J.P. explained that in the American League it make sense to play for the big inning, where often the winning team can outscore the losing team’s run total for the entire game in one inning. Contrary to what some have implied and reported, J.P. doesn’t use the spreadsheet exclusively and will amend the strategies in certain games or late-inning situations. He mentioned that afternoon’s matchup of Halladay vs. Sheets as one where the Jays would try to put anything they could on the board early.

The conversation soon turned to minor leaguers and the impact of coaches. Schwarz made the point that he would spend the extra $50,000 or whatever is necessary to hire the best possible scouts and coaches, as their impact on players could be extremely valuable. J.P. went on to express immense confidence in Dickie Scott and their entire minor league scouting staff. While he uses statistics to project college players and minor league players statistics, J.P. says he also relies heavily on the advice of his scouts and minor league staff. He also mentioned that he sees the first-round draft pick at least once a year in person, in addition to reviewing video, as the first-rounder is the guy “I have to be right on.” J.P. said that he saw Adams four times, and saw the other first-rounders once each, although he knew within “five minutes” of watching Hill that “he was our guy.” So far J.P. looks like he made quite a find with that first-round selection.

The case of David Purcey was raised, who had good college stats but who also was seen as very projectable by some. J.P. explained that they really liked his power arm from the left side, which they thought was especially important for the AL East. He admitted that Purcey had some wildness, but they felt his strikeouts could help offset that. The point was also made in the audience that statistics could not explain what problems could potentially be corrected and that cutting down Purcey’s control shortcomings even partially could become huge. J.P. was also asked about the wealth of pitching prospects in the system, from Rosario to Marcum to Jackson to Vermilyea, and he said that a couple might be up this September, but that most would appear in 2006 or 2007.

J.P. had to leave shortly afterwards and the lecture became very informal afterwards. Schwarz signed some books and talked with some audience members who had additional questions, while others mingled and talked. It was a very enjoyable time and Schwarz was a very good lecturer. He was engaging and funny and was able to accommodate an audience with a large range of baseball knowledge.

His book contains many interesting stories that were not covered in the lecture or this article, from a segment on Bill James to a part that details the fight over whether or not to preserve old baseball records, when in many cases it was later determined the totals were inaccurate. The book goes into greater detail on many of the people mentioned in the talk, including George Lindsay and Allan Roth. The Numbers Game is very accessible and flows very well, with many characters reappearing several times in different capacities. It’s highly recommended by both myself and Named For Hank.

Batter’s Box would like to thank the Learning Annex and Mr. Schwarz for their generosity in relation to this event and would also like to apologise for the delay in publishing this story.

Alan Schwarz Book Talk | 3 comments | Create New Account
The following comments are owned by whomever posted them. This site is not responsible for what they say.
Mike Green - Tuesday, July 05 2005 @ 11:52 AM EDT (#121561) #
In the sixties/early seventies, Allan Roth provided in-depth statistical support to TV's Game of the Week. At the same time, the print media statistical reports were highly focused on batting average. Many, if not most, major league teams had high average/low OBP hitters at the top of the lineup. Proving that the pen is mightier than the screen!
TangoTiger - Tuesday, July 05 2005 @ 11:54 AM EDT (#121562) #
Contrary to what some have implied and reported, J.P. doesn’t use the spreadsheet exclusively and will amend the strategies in certain games or late-inning situations. He mentioned that afternoon’s matchup of Halladay vs. Sheets as one where the Jays would try to put anything they could on the board early.

J.P. does something here that everyone should: accept that the main charts are based on average conditions, but that we should be using custom charts based on the given conditions of the day.

The small-ball strategies, like the sac bunt and stolen base, are much more effective with the run environment is very low.

I think we all realize this (which is why the SB numbers are so much lower than the 80s, and why you'd be crazy to steal at Coors), but I'm not sure the "average" discussion really tackles that issue head-on.

Great summary...

StephenT - Tuesday, July 05 2005 @ 01:23 PM EDT (#121587) #
Thanks for the event summary.

I just read the book a couple weeks ago. It's awesome. I shouldn't have waited so long to read it. Anyone who writes about baseball is at risk of looking pretty dumb if they don't know this information.
Alan Schwarz Book Talk | 3 comments | Create New Account
The following comments are owned by whomever posted them. This site is not responsible for what they say.