The Young Sabermetrician

While leafing through some old papers at my parents house last summer I came across a blue binder that contained roughly 20 pages of green bar paper filled with dot-matrix printing. Upon further inspection the binder contained my first sabermetric "studies" from the winter of 1982-1983 when I was a wee-lad of 14. While a detour into the mind of a 14-year old might not sound very instructive, looking at these studies reminded me of several concepts I later learned about baseball statistics and so since you obviously have nothing better to do, I thought I'd share some of those with you today.

The studies were performed with the help of an Osborne 1 personal computer that my Dad had purchased the winter before. The Osborne was released in April of 1981 and included 64K of RAM, a 5" screen (we subsequently purchased a 14" orange screen monitor), two 91K disk drives, and the CP/M operating system. It was the first "portable" computer and the keyboard could be attached to the CPU so that the it would pack up like a very heavy and bulky suitcase.

I remember vividly that while in the "showroom" in the winter of 1981 my Dad asked the salesman about the differences between the Osborne and the new IBM PC and its DOS operating system. The slick salesman assured my Dad that CP/M was the way to go and that Osborne would be around for a long time. Well, that didn't really work out but that Osborne did come with a spreadsheet (SuperCalc), Wordstar, Microsoft MBASIC, and Dbase II. It was SuperCalc and Dbase II that I used extensively in my baseball research.



My other tools I used in the studies were my first Baseball Encyclopedia that I purchase around this time and which sits proudly on my shelf next to the desk at which I'm typing, and the book Daguerreotypes which was published by the Sporting News and was a register of I believe Hall of Fame players.

That winter I joined SABR and promptly let my membership lapse as girls and other pursuits caught my attention. I came to my senses and rejoined in 2003.

I didn't read my first Baseball Abstract until 1984 although I must have run across some of Bill James' writing since I see Runs Created mentioned in one of the studies. I'm sure you're all dying to find out just what a 14 year-old in the bad old days before the internet (we did have a 300 baud modem that fit into the floppy drive slot and from which we could dial long distance into bulletin boards with lots of crappy software) could have possibly written about.


So in looking through the binder here were the studies, my sometimes strange conclusions, and what I was missing:

The Effect of the DH. In this study I looked at AL and NL pitchers both before and after the DH in order to see what effect the DH had on the career length of AL pitchers. It was my contention, being an NL kid who looked on the DH as an abomination, that the DH would shorten pitcher's careers since they would throw more innings.

What I reported was that NL pitchers who had pitched 200 innings or more in 1971 went on to pitch 6.18 more years while AL pitchers who pitched 200 innings or more in 1972 had careers 5.24 years longer. I was no doubt pleased to report that the DH shortened the careers of pitchers by almost a year. Of course I wasn't really aware of the dangers of small sample size and so my study included only 25 AL pitchers and 22 NL pitchers and I made no allowance for pitchers changing leagues. In the end, I have no idea whether the DH did shorten careers but I'm pretty sure this study wouldn't have found it if it did.

What Makes a Winner? In this study I recorded the World Champions from 1960 through 1980 and recorded where they placed in their league in homeruns, AVG, fielding AVG, ERA, and SB and computed the average. I assume I was trying to figure out how teams win.

My results were HR - 3.52, BA - 3.43, FA - 3.19, ERA - 2.62, SB - 4.43. I'm not sure what if anything I concluded from this as the rest of the page is blank, but of course this study was flawed since park effects weren't taken into account and because the sample size was small including only World Champions - not to mention the fact that the winner of the World Series is not the best team in baseball a majority of the time.

Ballpark Comparison. On this page I recorded the 12 NL ballparks and made a calculation as to how the park affected AVG, SLUG, Runs, Hits, Total Bases, 2B, 3B, and HR. I'm not sure where I got the data but earlier that year my family visited Cooperstown and my brother and I received permission to go into the library and what we were looking for were all kinds of split information (left/right, home/road etc.). We may have photocopied something that had home/road splits in it that I used for this study.

In any case the data shows that Veterans Stadium increased runs by 20.6% while Dodger Stadium depressed runs by 13%. Other hitters parks were Riverfront, 3 Rivers, and Fulton Country Stadium. Once again I probably looked only at a single year and I now know that park factors, while useful, fluctuate wildly from year to year. They should be calculated over a three or five year time period and then be regressed to the mean.

East vs. West. In March of 1983 a friend who was a Royals fan had apparently claimed that the AL west was the better division in the AL. At that time I was having a brief affair with the Yankees and so I sought out data to refute his claim. To do I looked at all the "main starters" on each team in the AL for 1982 and found how they pitched against each division (I have no idea where I found the data). I then listed each starter, his ERA and record against the AL East and the same against the AL West and computed the difference in ERAs. After totaling it all up the western division starters went 164-191 with a 4.34 ERA against the East and 179-171 with a 4.21 ERA against the West. Eastern division starters went 171-175/4.37 against the East and 218-165/4.01 against the West. Both sets of data show that the East was indeed stronger than the West.

I also noted that the Western division as a whole went 265-323 against the East. For good measure I included the Royals' starters totals and showed they performed better against the West than the East as well. My conclusion: "...all I can say is that my friend who asked me this question should know what he's talking about before he starts talking." Not well written but I can't argue with the conclusion (at least for 1982).

Left vs. Right. As I mentioned previously I was very interested in split data and before the Abstracts and the Internet it was extremely hard to come by. Because I was lacking that data I drafted my then 11 year old sister into recording some data while watching games. Over a period of several months we recorded about 750 plate appearances. I put the data into SuperCalc and my splits showed the following:

Right handed hitter vs. left handed pitcher: .209
Right handed hitter vs. right handed pitcher: .333
Left handed hitter vs. left handed pitcher: .289
Right handed hitter vs. left handed pitcher: .303

My conclusion was that announcers are wrong when they talk about a general platoon advantage and any differences probably have more to do with the individual and how much "practice" they get against each side. Well, my sample size here was really small with left vs. left at only 38 at bats and right vs. left with only 115 at bats and so the results are obviously meaningless. In fact, the batting average over the entire sample was .300. The other problem, of course, is that the sample was taken probably almost exclusively from Cubs games, those being the only games that we were regularly able to watch in eastern Iowa at the time.

Using retrosheet data here are the actual splits for the 1982 season:

Bats   Throws      PA     AVG   OBP    SLUG
R      R        59110   0.255  0.305  0.381
L      R        38048   0.270  0.329  0.405
B      R        15403   0.263  0.317  0.367
L      L        11078   0.253  0.312  0.362
B      L         6799   0.250  0.294  0.350
R      L        30666   0.268  0.321  0.413

This study brings to mind the article "Underestimating the Fog" that Bill James wrote for the Baseball Research Journal last spring and that stirred the waters of the sabermetric community a bit. In that article James criticizes the common technique employed in various sabermetric studies that are typically cited to "prove" conclusions - for example James' own look at platoon differentials in the 1988 Baseball Abstract. That technique involves the search for recurrence or persistence of the phenomena being studied using year to year correlations. In other words, James did a study that attempted to determine if platoon differentials persisted across seasons. He found that they did not for individual players and so concluded that hitters have no individual ability to hit well or poorly against left-handed pitching but that there is a strong group tendency to do so.

James then argues that in this cases as in others the negative conclusion - the phenomena is not real - is flawed because there is too much instability in the data used to make the conclusion. For example, the conclusion that there is no specific ability to hit well or poorly against left-handed pitching is based on platoon differentials where the number of plate appearances against left-handed pitchers is around 120 in a season. The randomness involved in such a small sample size tends to swamp the differential itself, thereby making the results meaningless.

The Count. Using the data I collected for the previous study I showed a table of how hitters hit in the various counts. I then classified each count as a hitters or a pitchers count. My hitters counts were 3-0, 2-0, 0-1, 3-1, 2-1, and 0-0. In each of these counts the hitters hit anywhere from .381 to .500. On the pitcher's counts the hitters hit from .292 on 1-1 counts to .132 on 3-2 counts. I then totaled up hitters and pitchers counts and came to the conclusion that "it was learned that the hitters hit 109 points better on the hitter's pitch." I remember using these results that winter in a speech for speech class on pitching strategies.

While the number of at bats in each count wasn't large, these results were more significant because of the big differences between the various accounts and track pretty nicely with a more recent query I ran on counts for 2004. What I was missing in this analysis is the effect of strikeouts. On several of the pitcher's pitches (0-2,1-2,2-2, 3-2) the batter has the chance of striking out whereas on hitter's counts a swing and miss or a pitch taken for a strike doesn't depress the batting average. By taking the strikeouts out of the picture what you learn is that while batting average is generally lower in pitcher's counts, it is the slugging percentage that suffers most. In other words, as common sense would dictate, with two strikes hitters often sacrifice power at the expense of average as they look to protect the strike zone and the put the ball in play.

Offensive Efficiency Rating. Perhaps my most ambitious project, although not the most well thought out was something I called "Offensive Efficiency Rating". I hailed this as a new way of rating players using their career offensive statistics in order to answer the age old question of who was the greatest offensive player of all time.

To compute this rating I assigned weights to nine different offensive categories as follows:

Single      275
Double      350
Triple      400
Homerun     500
Stolen Base 250
Run         375
Walk        235
Strikeout  -275
RBI         425

I used as my sample all of the players in the book Daguerreotypes and included the recently retired Willie Stargell, Bobby Bonds, and Reggie Smith. To calculate the rating I then divided the career totals in each of the nine categories by the player's number of at bats and then multiplied this rate by the weight. After adding up each of the weighted values I then ranked the players by this rating. For players whose statistics were incomplete I projected their career totals using the rate for that part of their career for which data was present.

This study was carried out using Dbase II and I remember painstakingly typing in the statistics for the approximately 275 players in the book. The exercise of performing the calculations did force me to learn the Dbase programming language, which of course had a pretty long shelf life and which I used in a college job doing FoxBase programming at Iowa State University. I know you're dying to see who really were the greatest offensive players of all time so...

Babe Ruth      346.0
Ted William    342.1
Lou Gehrig     329.5
Billy Hamilton 316.6
John McGraw    306.2
Joe DiMaggio   293.5
Jimmy Foxx     291.8
Ed Delahanty   291.4
Ty Cobb        290.0
Sam Thompson   289.7

As you can see from the list the rating is biased towards players who played before 1901 with four making the list. I also produced a list of post 1900 leaders that went like so:

Babe Ruth      346.0
Ted William    342.1
Lou Gehrig     329.5
Joe DiMaggio   293.5
Jimmy Foxx     291.8
Ty Cobb        290.0
Hank Greenberg 287.7
Rogers Hornsby 280.4
Mel Ott        276.7
Tris Speaker   274.6

Here players who played prior to 1950 are given the nod.

Obviously there are many things wrong with this system not the least of which is that the weightings of the events are out of whack. The weights given to stolen bases, singles, doubles, and triples at the expense of homeruns and walks allows players from the 1800s and those in the deadball era to do well. Players from the 1950s and later are hurt even more by the -275 assigned to strikeouts - a weight that is equivalent to singles and greater than walks and that is wildly inaccurate. Simply put, strikeouts don't matter that much in the big scheme of things. One of the side bars in the book Mind Game makes the point well by showing that team strikeouts have little effect on total runs scored. The same can be said of stolen bases - a weapon that is strategic in nature.

In addition you'll also see that the weighting of runs scored and RBIs is out of whack since other systems don't take these into account at all. The reason is that they are highly dependent on the team context in which the player finds himself.

I've written a bit about the history of creating weights for offensive elements in a series of article on run estimation last year at this time.

• A Brief History of Run Estimation, Part I
• Run Estimation: Runs Created
• Run Estimation: Batting Runs
• Run Estimation: Estimated Runs Produced
• Run Estimation: Base Runs

Outside of the event weights the system suffers from a lack of context as well. You simply can't compare the raw offensive totals of players from different eras because the run environments in which they played differs. Hitting .400 in the 1930s is not the same thing as doing it in the 1960s as illustrated in this graph that shows the runs per game since 1901.



The same can be said of the park in which they played as mentioned previously. As a result the stats would have had to have been adjusted for the context before making comparisons. From an offensive perspective Michael Schell has taken the approach to its highest level in his books Baseball's All-Time Best Hitters and Baseball's All-Time Best Sluggers : Adjusted Batting Performance from Strikeouts to Home Runs.


Reading through these studies it became clear to me how much my thinking about baseball has been influenced by the writing of Bill James, Craig Wright, and Pete Palmer among others. Given the rise of personal computers and the information revolution, however, the kind of analysis they brought to the table was destined to occur. And that kind of analysis is now breeding a new way for teams to draft, shape their roster, and evaluate their talent.

Searching for Significance

My article looking at matchup data for 2003 through 2005 is up on THT today. The article looks at the 30,481 batter/pitcher matchups during that time period in order to take a look at which ones are statistically signficant and make some general observations about matchups.

Although it didn't come out prominately enough in the article Lawrence Weintraub (who is in astrophysics at CalTech and has been up each night waiting for the clouds to part), and John Walsh really helped me work through the issues surrounding the results I was getting and so I'd like to more publicly thank them. John will also have some very interesting articles published on THT in the coming days.

Palocracy

Well, the offseason has certainly gotten off to a roaring start. With seven teams changing GMs and the soap opera like stories around Theo Epstein in Boston, Paul DePodesta in Los Angeles, Pat Gillick in Philadelphia, Josh Byrnes in Arizona, and Gary Hunsaker in Tampa Bay (although not officially a GM), there's lots for the baseball fan to take note of.

The number of changes at the top has resulted in lots of discussion about the increasing notoriety, celebrity, and importance of GMs in the modern game. David Pinto had a nice post on the subject titled GM Stock last week. Pinto argues that free agency and the information revolution have had much to do with innovation in baseball being pushed up the ladder to the GM and away from the managers on the field. That trend has been accelerated by the debate that Moneyball started between the old school and the new.

I would also add that the amount of money now at stake in the game has upped the ante for the owners. That has led to owners who treat their teams more as investments and less as play things where efficient and innovative practices are sought out. And that brings us to the culture of baseball.

Last weekend I was re-reading Men At Work: The Craft of Baseball by George Will published in 1990, and ran across a passage that does a great job of painting the picture of baseball's management culture as it existed before the information revolution had made much of an impact and even as it exists in some places still today.

"To the familiar classifications of social systems, now add a new category to cover the peculiar governance of baseball. To aristocracy, plutocracy and democracy add baseball's contribution to government: 'palocracy,' government by old pals. Baseball has traditionally been run by men whose lives have been intersecting and entwined for decades. They have known one another from the rocky playing fields and spartan offices of the low minor leagues all the way up to the manicured playing fields and well-appointed suites of the major leagues. You do not talk long with a baseball person you hear the phrase 'baseball person.' Often it is accompanied by a negative: So-and-so is 'not a baseball person.' No adjective is required, thank you very much. A baseball person is a good baseball person. A palocracy can make for kinder, gentler governance, but it also can make the world safe for mediocrity..."

"Closed systems, such as tenured university faculties or diplomatic corps or military services, are vulnerable to systemic mediocrity. People who have gone to the same schools, climbed the same career ladders, absorbed the same values and assumptions and expectations, become intellectually insular and professionally self-protective. They forgive one another their mistakes, and mediocrity becomes cozy."

One of the effects of the larger role of GMs and the Moneyball debate has been to start to break the palocracy and replace it with a meritocracy that uses all of the tools available to build a winning team. And that is very good for baseball.

Call the Feds

Here's a post that nicely sums up my feeling on the Neifi Perez two year $5M contract.

I've said it before and I'll say it again, you can't give Dusty Baker more than one good field, no hit infielder since he really really really likes to bat them at the top of the order and have them sacrifice willy nilly. I'm sorry but that means that Joe Macias must be transferred into the witness protection program immediately.

Third Base Crossroads

My fellow writer Steve Treder at THT has written a very interestinge piece on third base as a crossroads position of sorts. Although I've written about the changing importance of defensive skills at the position over the years, Steve argues that third base sitting in the middle of the defensive spectrum as it does has consequences:

"...third base is simultaneously easier to play than half of the positions and harder to play than the other half. This makes third base a position that one half of the roster can aspire to as the highest challenge of their skills, as well as a position that the other half of the roster can relax toward as their skills erode due to age or injury. To be played well, third base demands the strong arm and quick reflexes of more difficult positions, but third base can also be played passably by the slow-footed players generally found in the easier positions."

The consequence of this is that third base is viewed as transitional by most teams, which has an impact on the architecture of their rosters.

"...a very high proportion of players are likely to spend at least some time playing third base. This high degree of potential traffic passing through third base—its Crossroads status—has the consequence of causing many teams to plan for third as a transitional position, to leverage its capacity to maximize roster flexibility. It often makes sense for teams to not concern themselves with finding and settling on a long-term answer at third base, thus reducing the opportunity for players to pursue full careers as third base specialists..."

The result of the way third base viewed is worked out in the fact that there are fewer "no doubt" Hall of Famers at third base, just eight, than there are at any other position. It also is a reason that a seeming majority of teams like the Cubs (where there is even a name for the phenomenon - "the curse of Ron Santo"), White Sox, and Yankees have struggled finding a third baseman that could man the position for more than a year or two. Great stuff Steve.

Fun with Matchups

Question: How often do particular hitters face paritcular pitchers in a season?

To answer that question I took a look at matchup data for 2003 through 2005. The following list are those hitters with the most at bats against a particular pitchers over that time period.

I've also included the hitter's average over the period, the pitcher's average against as well as the expected average for the matchup based on the log5 method that Bill James published in the 1983 Baseball Abstract. Dan Levitt has written a nice article on the formula and its accurracy.

So from the table you can see that around 15-20 at bats per season are probably the max.

Batter              Pitcher               AB       H     AVG    BAVG    PAVG   ExAVg
Garret Anderson     Jamie Moyer           45      10   0.222   0.300   0.267   0.300
Juan Pierre         Livan Hernandez       43      13   0.302   0.303   0.262   0.297
Carl Crawford       Tim Wakefield         42      15   0.357   0.293   0.251   0.277
Michael Young       Barry Zito            41       9   0.220   0.317   0.234   0.281
Ray Durham          Brian Lawrence        41      14   0.341   0.286   0.272   0.292
Angel Berroa        Mike Maroth           41      15   0.366   0.273   0.292   0.299
Derek Jeter         Tim Wakefield         40       9   0.225   0.307   0.251   0.291
Garret Anderson     Ryan Franklin         40      13   0.325   0.300   0.272   0.306
Ichiro Suzuki       Barry Zito            40      14   0.350   0.330   0.234   0.293
Derek Jeter         Rodrigo Lopez         40      19   0.475   0.307   0.279   0.321
Miguel Cabrera      Livan Hernandez       39       8   0.205   0.300   0.262   0.295
Jimmy Rollins       Dontrelle Willis      39       9   0.231   0.281   0.254   0.268
Manny Ramirez       Rodrigo Lopez         39      12   0.308   0.309   0.279   0.323
Juan Encarnacion    Livan Hernandez       39      15   0.385   0.265   0.262   0.260
Ray Durham          Jake Peavy            38       6   0.158   0.286   0.230   0.248
Luis Castillo       Livan Hernandez       38       9   0.237   0.302   0.262   0.297
Manny Ramirez       Roy Halladay          38       9   0.237   0.309   0.248   0.288
Juan Pierre         Tom Glavine           38      11   0.289   0.303   0.272   0.309
Luis Gonzalez       Jason Schmidt         38       8   0.211   0.280   0.214   0.226
Torii Hunter        Mark Buehrle          38      11   0.289   0.262   0.271   0.266
Michael Young       John Lackey           38      15   0.395   0.317   0.271   0.322
David Ortiz         Roy Halladay          38      14   0.368   0.297   0.248   0.277
Mark Kotsay         Ryan Franklin         37       9   0.243   0.288   0.272   0.294
Cesar Izturis       Brett Tomko           37       9   0.243   0.267   0.281   0.282
Johnny Damon        Roy Halladay          37      13   0.351   0.298   0.248   0.278
Carl Crawford       Josh Towers           37      15   0.405   0.293   0.290   0.317
Eric Chavez         Kenny Rogers          36       7   0.194   0.275   0.285   0.295
Ichiro Suzuki       John Lackey           36      11   0.306   0.330   0.271   0.335
Johnny Damon        Rodrigo Lopez         36      11   0.306   0.298   0.279   0.312
Michael Young       Joel Pineiro          36      13   0.361   0.317   0.268   0.319
Eric Chavez         Ryan Franklin         36      12   0.333   0.275   0.272   0.281
Derek Jeter         Sidney Ponson         36      16   0.444   0.307   0.294   0.336
Hank Blalock        Joel Pineiro          35       5   0.143   0.279   0.268   0.281
Hideki Matsui       Tim Wakefield         35       6   0.171   0.297   0.251   0.281
Cesar Izturis       Adam Eaton            35       8   0.229   0.267   0.261   0.262
Michael Young       Bartolo Colon         35      11   0.314   0.317   0.255   0.304
Shannon Stewart     Jake Westbrook        35      11   0.314   0.294   0.265   0.293
Juan Pierre         Mike Hampton          35      12   0.343   0.303   0.273   0.310
Alex Cintron        Jason Schmidt         35       9   0.257   0.283   0.214   0.229
Shawn Green         Jake Peavy            35      10   0.286   0.277   0.230   0.240
Jacque Jones        Jon Garland           35      12   0.343   0.268   0.262   0.264
Lew Ford            Mark Buehrle          35      13   0.371   0.285   0.271   0.289
Brad Wilkerson      Tom Glavine           35      14   0.400   0.257   0.272   0.262

Will and Schwarz

Just wanted to mention that my review of BP's new book Mind Game is up on the THT site today.

I was also pleased to find a piece on Baseball America (subscription side) from two of my favorite writers, an interview Alan Schwarz did with George Will. In the intro to the interview Schwarz notes that Will's Men At Work (perhaps my favorite baseball book) is the best selling baseball book of all time.

While the interview is worth the read (and an online subscription to BA is also worth it) my favorite exchange, and one I empathize with, was this one about Will's past as a long suffering Cubs fan.

AS: What has been the worst moment for you as a Cubs fan?

GW: Losing Game Six of the Marlins playoff in 2003--the Bartman Game. Game Seven was an afterthought. The man who should love Bartman and put him in his will is the Cubs' shortstop--Alex Gonzalez. No one remembers that Gonzalez dropped a double-play groundball the same inning. I remember walking out of Wrigley Field and someone saying, "Mr. Will, we'll get 'em tomorrow." I said, "not a chance."

I remember the 1984 playoffs with the Padres. The Cubs had won the first two games, and Don Drysdale was broadcasting for ABC at the time. I'm walking out of Wrigley field next to him and he says, "All right, Will, now do you believe?" and I said, "Every Cub fan knows it's the Padres in five." Which of course it was.

After game six in 2003 I had exactly the same feeling throughout the day of game 7. I allowed my hopes to rekindle when Moises Alou hit his two-run homer in the 3rd inning to give the Cubs a 5-3 lead but I should have known better.

Will's mention of Alex Gonzales is another example of the "but for this..." canonical story in action that I wrote about last week on THT. The Bartman play becomes the fulcrum on which the series tipped while the Gonzales play is lost because of our propsensity to simplify the story in order to fit our model. I wouldn't be surprised if 20 years from now folks will forget there was a game 7 at all.

I also found interesting Will's analysis of the 2005 post season (the interview was done just as the playoffs were beginning).

GW: I may annoy people, but my reasonably longshot bet--which is the only kind worth making--is to look out for Houston in the postseason, because they've got Clemens, Pettitte and Oswalt....Jerry Reinsdorf has been a very good baseball owner and cares passionately about it. Some people think that all he cares about are the Bulls. He really wants to win in the American League and I'd like to see him get a chance. That's an exciting team.

Esptein Walks

Another blow for the performance analysis crowd as Theo Epstein does not re-sign with the Red Sox as their GM. However, rather than being thrown out it appears Epstein left because he felt his relationship with CEO Larry Lucchino was compromised and so he walked away. You can read his statement here.

Earlier reports I'd seen said money was at issue as well although at the end he appeared to be able to get his $1-1.5M per year. Of course that makes him the most marketable general manager out there and who knows, he may have gotten some information on another job opening up or be interested in the Dodgers, Phillies, or Tampa Bay jobs. It's surprising to me that the Sox would mess around with this and didn't get the deal done sooner. And of course this will be another interesting situation to watch as the new GM will have decisions to make on Johnny Damon, David Wells, and the perennial off season attempted trade of Manny Ramirez.

Fallout from this could very well include the cadre of analysis gurus the Red Sox employ including Bill James and Voros McCraken. There could be two more free agents in the making.

More But for This...

First, just wanted to say thanks for all the feedback on my article on umpiring in the post season over at THT.

One reader noted that:

"...your listing of blown calls this post-season omitted a crucial call from this year's World Series: Jermaine Dye's walk in Game 2, which turned out to be a foul ball (which even Dye, to his credit, admitted after the game). Can't say what would have happened if Dye's swing had been correctly called foul. But we do know that the ball 4 led to reliever Dan Wheeler being taken out of the game, and the new reliever, Chad Qualls, gave up a grand slam HR on the next pitch (to Konerko)."

And of course the reader is absolutely correct, however, I have a small excuse. The article was written prior to game 2 but wasn't run until today due to scheduling. I should have updated it to reflect this play since it was clearly another blown call and one that had a potentially huge impact on the game and the series.

Another reader makes a case that the 1969 series had an even more egrarious call than the one I mentioned - the game 5 reversal of the ball call on Cleon Jones to a hit by pitch after the Mets disputed it. Jones went on to score on a Don Clendenon home run and the Mets won the game 5-3 and the series. The difference on this call is that to me it wasn't clearly blown since the Mets had some evidence (the shoe polish) to back up their claim. The reader would argue, however, that the practice of huddling and reversing a call was not standard for the time and so the call should not have been reversed.

This is also an example of the main point of the article since two Baltimore errors in the 8th along with a Ron Swoboda double are what plated the final two runs for the Mets. Everybody remembers the hit by pitch but not the errors.

Relying on the Homerun

Here's a quick question. Which teams relied most on the homerun in 2005?

One way to answer that is to rank the teams by the percentage of their runs that scored on homeruns. Here they are:

Team          RS   RS-HR     Pct
TEX          865     413   47.7%
CIN          820     364   44.4%
NYA          886     390   44.0%
CHA          741     314   42.4%
CHN          703     296   42.1%
ARI          696     288   41.4%
CLE          790     321   40.6%
BAL          729     293   40.2%
ATL          769     300   39.0%
BOS          910     348   38.2%
MIL          726     275   37.9%
NYN          722     269   37.3%
HOU          693     257   37.1%
DET          723     262   36.2%
PHI          807     290   35.9%
TBA          750     262   34.9%
LAN          685     238   34.7%
SLN          805     275   34.2%
COL          739     246   33.3%
SFN          649     213   32.8%
OAK          772     252   32.6%
MIN          688     219   31.8%
ANA          761     240   31.5%
PIT          680     211   31.0%
TOR          775     238   30.7%
SEA          699     212   30.3%
SDN          684     206   30.1%
FLO          717     215   30.0%
KCA          701     200   28.5%
WAS          639     166   26.0%

Texas leads both in percentage of runs scored on homeruns and total runs scored on homeruns while the playoff teams come in 3rd (Yankees), 4th (White Sox), 9th (Atlanta), 10th (Boston), 13th (Astros), 18th (Cardinals), 23rd (Angels), and 27th (Padres).

One Step Forward, One Step Back

In the world of sabermetric-minded general managers, this weekend saw one step forward and one step backwards. First, the Arizona Diamondbacks have hired former Red Sox assistant GM Josh Byrnes replacing Joe Gariagiola Jr. who took a job with Major League Baseball. Byrnes got a four year contract with a club option for a 5th. Byrnes was the assistant GM here in Colorado for two years before going to Boston. Speaking of Boston, GM Theo Epstein is still in contract negotiations, and it appears he's being a tough negotiator.

While Byrnes is considered by many to be more of a Moneyball type he also has a reputation for being somewhat of a moderate. That was the step forward. Interestingly, Will Carroll noted on BP Radio on Saturday that the GM position in Arizona may have scared some potential candidates away since it appears that Matt Williams may be a sort of "GM in waiting" and has the ear of owner Jeff Morad. It'll be interesting to see what kind of patience Morad and his partner show with the young Byrnes and how much impact Byrnes will be able to have given that the D'Backs are still saddled with some large contracts, namely $11.5M to Luis Gonzales and $11M to Javier Vazquez and of course Troy Glaus ($9M) and Russ Ortiz ($7.4M). The story in the Arizona Republic noted that the owners recognize they're not baseball men and so Byrnes will be given the autonomy to make the decisions. It'll be an interesting situation to watch.

The step back came on Saturday when Paul DePodesta, the Dodgers GM who was hired before the 2004 season was fired. The way the story's being played in the mainstream media is that DePodesta is a poor communicator and neglected team chemistry, which led to the Dodgers 71-91 record this season after winning the NL West in 2004. The chemistry angle comes from the trade of fan favorite Paul LoDuca and the acquisitions of Jeff Kent, Milton Bradley, and J.D. Drew, all of whom are not exactly Shawon Dunston in the clubhouse.

While John Wesiman over at Dodger Thoughts will provide much better commentary over the next few days, my initial reaction is that essentially one year is too little by which to evaluate a GM unless they absolutely prove inept. After all, DePodesta has never even hired a manager let alone had time for draft picks during his tenure to help the club. It seems as if Frank McCourt is a guy who is quick to change directions when things don't work out immediately and probably hired DePodesta not because he understood the philosophy of "be the house" but rather because it was viewed as the hot new thing. McCourt's management style having now fired two GMs in 22 months doesn't bode well for the Dodgers franchise in the years to come. One could see him swinging between one extreme and the other in the same way as when Martin Luther compared human nature to a drunkard trying to ride a horse. He gets on and falls off on the left side. He resolves not to make that mistake again, so he remounts, careful to avoid falling off on the left, and promptly falls off on the right.

And it looks like McCourt will be leaning the other way in considering a longtime baseball guy like Pat Gillick, a man who has openly criticized the Moneyball approach as I noted in my article on the book Scout's Honor.

Angus on Baserunning

Jeff Angus, author of the Management by Baseball weblog has written a nice piece on baserunning over at Baseball Analysts where he mentions my work on baserunning. Thanks Jeff.

In that article he notes that he's using data that goes through August. I now have run my framework for 2005, complete results from which will be published in the 2006 Hardball Times Annual (order yours today), and here are the top 10 baserunners listed by opportunities. Keep in mind that these opportunities consist of those where the runner was on first with second empty and the batter singles, the runner is no first with second empty and the batter doubles, and when the runner is on second with third empty and the batter singles.

These numbers are also park and situation adjusted and the columns mean:

• Bases - the number of bases gained in these situations. This number differs from what Angus showed since he calculates a "safe %" indicating how often the runner took the extra base


• OA - the number of times the runner was out advancing the extra base


• IB – the difference between the actual and expected bases gained. Negative values indicate that the base runner did not advance as many bases as an average player would have given the same opportunities


• Base Runner Runs or BRR – the total number of runs a runner can be attributed given his base running performance in the opportunities he had (given our three scenarios above)


• Incremental Runs or IR – the difference between the BRR and the expected number of runs. Negative values indicate that the runner performed poorly in his opportunities by not taking as many bases and therefore contributing fewer runs than he should have


• Incremental Run Percentage (IRP) – a rate statistic calculated as the ratio of IR to BRR. IRP is akin to OPS+ in that it shows at what rate players contribute runs with their base running. Values over 1 indicate players who contributed above what would have been expected while values under 1 indicate a poor performance. IRP is needed for player comparisons since IR, like RBIs, is weighted by opportunity.

Drum roll please....

Team    Name             Opp Bases OA    IB   BRR    IR  IRP
NYA     Derek Jeter       69   105  3   4.4  7.99  0.99  114
BOS     Johnny Damon      66   100  2  -1.4 10.38 -0.24   98
BAL     Miguel Tejada     65    99  1   3.1 10.47  0.35  103
SLN     David Eckstein    64    95  1   1.9  7.58  0.42  106
PHI     Jimmy Rollins     63    96  2   3.8  7.04 -0.25   97
SEA     Ichiro Suzuki     63    99  1   5.3  9.26  1.56  120
CHN     Jeromy Burnitz    61    83  2  -2.7  6.97 -0.51   93
ANA     Darin Erstad      59    95  0   9.0 11.14  3.05  138
BAL     Brian Roberts     59    83  1  -1.8  8.09 -0.63   93
CHN     Derrek Lee        59    91  1   3.0  9.79  1.16  113

From this you can see that even though Johnny Damon may have taken the extra base upwards of 60% of the time, when considered in his context complete with David Ortiz and Manny Ramirez as hitters coming behind, he actually contributed negative .24 runs and was 2% below average. On the other hand Darin Erstad of those Angels contributed an additional 3 runs with his baserunning and was 38% above average, which incidentally was just slightly more than Chone Figgins contributed (3.02).

Bernanke on Baseball

Just found out courtesy of the SABR-L list that Bush's nominee to head the Federal Reserve Board, Ben Bernanke, is a baseball fan. There was a short discussion on NPR of Bernake's view on ERA on "All Things Considered".

Bernanke's view is that ERA is not a "fair" measure by which to gauge relief pitchers. In the NPR piece Dwight Jaffee, the Willis Booth Professor of Banking Finance and Real Estate at the Haas School of Business at the University of California at Berkeley discusses how he an Bernanke came to this realization but never really developed anything better.

Of course this realization has occurred to many people over the years and a few of the ways to correct for it I mention here. Our Fielding Independant Pitching (FIP) statistic on THT is another way to correct for it.

Bernanke's interest in baseball was also discussed in the October 26th New York Times.

"Perhaps the best way to establish transparency, Mr. Bernanke has written, is to set up an easily understandable process. It is in many ways the opposite of the image that Mr. Greenspan has helped cultivate, that of an oracle.

He will probably work toward depersonalizing monetary policy,' Mr. Gertler said.

A useful analogy is to Bill James, the baseball writer whose books Mr. Bernanke has discussed with fellow economists. Mr. James's work, which often argues for statistical analysis over intuition, 'was always a topic of conversation,' Mr. Gertler said.

He remembers Mr. Bernanke's being especially interested in finding statistics that forecast future performance. In the same vein, he has argued that central bankers should rely as much as possible on the enormous amounts of economic data now available and less on hunches.

The hope eventually is to come up with a statistical formula that processes this information and gets the best forecast,' said Jean Boivin, a Columbia economist who has done research with Mr. Bernanke. 'Once you do that, the question is how much is left over for judgment.'"

Chone Figgins and the Defensive Spectrum

In my most recent article on THT I wrote about the shift in the defensive spectrum that occurred in the 1940s that resulted in third base moving left on the spectrum while second base moved right.

In my brief analysis I noted that the prevalence of bunting and chop hitting in the first part of the century was likely responsible for third base being considered the more demanding defensive position. Dave Studeman, who knows alot of cool stuff, pointed out to me that Bill James had written on this subject in his book Win Shares in an article titled "Why Did the Defensive Spectrum Jump?". I have the book and have read it but my fast aging mind seems to have forgotten that article. Tonight I took my copy off the shelf and re-read the article.

He notes the same trend I show in the graph as to the offensive production of second base versus third base in the decades from 1900 through 1969 using Runs Created and shows the shift occurring sometime in the 1930s. My graph shows the same trend and its clear by the mid 1930s that third baseman had overtaken second sackers.

However, James does not attribute the shift in the spectrum to bunting and chop hitting but rather to the dual factors of increasing double plays and decreasing errors. Since 1947 there have been more double plays in the majors each year (except 1963 and 1975) than errors. His reasoning is that as double plays became more common managers were more willing to sacrifice offense in order to get double plays. At the same time as errors became less frequent managers were willing to put lesser defensive players at third base since the cost wasn't as high.

That makes perfect sense to me and so I certainly don't disagree. But I still think that bunting and chop hitting probably had an impact, although lesser, as well.

Interestingly, he notes how this trend is detectable in the common language we use to describe the positions. In the early part of the century second baseman were referred to as "keystone men" whereas today they are referred to as "pivot men" emphasizing their ability to turn two.

Air Force Taboo

Well, here in Colorado Springs Air Force coach Fisher DeBerry apologized today for the comments he made after his team's 48-10 loss to TCU over the weekend. In those comments he noted that TCU

"had a lot more Afro-American players than we did and they ran a lot faster than we did. It just seems to me to be that way. Afro-American kids can run very well. That doesn't mean that Caucasian kids and other descents can't run, but it's very obvious to me that they run extremely well."

He also on Monday told The Gazette of Colorado Springs that "you don't see many minority athletes in our program", implying that the academy needed to recruit more minorities.

For those that have read my review of the book Taboo you'll not be surprised that I agree with DeBerry's comments in general. Now I have no specific knowledge of the TCU or Air Force athletes, but as a general principal there is no doubt that at the elite level athletes of west African descent dominate in sports where speed and jumping ability confer an advantage. Football, at some positions, fits the mold and so it would in fact be surprising if blacks weren't overly represented. It also wouldn't surprise me that if a team with 85% African Americans (who are primarily of west African descent) as compared to a team with 15% African Americans wouldn't dominate the top positions when both teams are ranked from fastest to slowest.

So then why was DeBerry forced to apologize? For the simple reason that physical group differences are a taboo subject because of their link to possible mental group differences. That's why, rather than the stated reason that it was a violation of the academy's racial and ethnic anti-discrimination policy, Athletic director Hans Mueh found himself saying:

"What we're talking about is speed. There's speed that cuts across black, white, gray, blue, whatever. It was just an inappropriate comment and you all know it was an inappropriate comment."

The first part of his comment is simply and demonstrably not true. The second part reflects the politically correct orthodoxy.

The simple fact is that DeBerry's comments weren't disciminatory. Discrimination occurs when an individual is treated differently (usually badly) only because of their race and not based on their talents. Acknowledging that there are group differences in no way discriminates against anybody because it has nothing to say about any individual.

Streaky Chat

Just saw that Alan Reifman has posted a chat with Jim Albert and Jay Bennett on his Hot Hand in Sports site. Albert and Bennett of course are the authors of the excellent book Curve Ball: Baseball, Statistics, and the Role of Chance in the Game.

I thought that their answer to the question of why there is more luck in baseball was interesting. Their three main reasons were:

"(1) The major sports allow the best team to be fielded in each game. Because pitchers can only start once every four or five days, the most important player in baseball varies from game to game. This tends to equalize teams in a baseball game. How would the Patriots do if Tom Brady played in only one quarter of the games? It would tend to bring them back in the pack.

(2) Most major sports allow unlimited use of their best players. In football, if you have Jim Brown or Walter Payton, your use of his exceptional playing ability is limited only by defensive adjustments (which also helps to make the rest of your offense look better). In basketball, you can give the ball to Michael Jordan or Allen Iverson as much as you want. Baseball limits the use of your best players in the lineup. You can't allow Barry Bonds to bat any time you want. The use of an advantage in ability is severely limited in baseball. I think this is the primary reason for your example about the 49ers vs. the Patriots at Foxboro and the Devil Rays sweeping the Yankees.

(3) Baseball has a longer season than all other major sports. It's more likely to achieve extremely high or low winning percentages with a shorter season. This is especially true for football, less so for basketball and hockey."

To these I would simply add that in baseball a single event such as the swing of the bat can more often alter the score and therefore make the difference between winning and losing whereas in sports such as basketball and baseball it usually takes a combination of skillfully executed plays to result in a score. When that is the case a less skilled player can have a disproportionate affect on the game.

Grand Slams and Game Enders

I've had some folks ask me who was the last player to hit a grand slam in the World Series after Paul Konerko's game 2 blast. Luckily, the guy who keeps track of everything related to homeruns, SABR's David Vincent published a list on SABR-L this week complete with the inning. And for good measure he threw in the walk-off homeuns as well.

World Series Game-Enders

                                    
10/05/1949 Tommy Henrich       NYA AL  9
09/29/1954 Dusty Rhodes        NY  NL 10
10/06/1957 Eddie Mathews       MIL NL 10
10/13/1960 Bill Mazeroski      PIT NL  9
10/10/1964 Mickey Mantle       NYA AL  9
10/21/1975 Carlton Fisk        BOS AL 12
10/15/1988 Kirk Gibson         LAN NL  9
10/18/1988 Mark McGwire        OAK AL  9
10/26/1991 Kirby Puckett       MIN AL 11
10/23/1993 Joe Carter          TOR AL  9
10/26/1999 Chad Curtis         NYA AL 10
10/31/2001 Derek Jeter         NYA AL 10
10/22/2003 Alex Gonzalez       FLO NL 12
10/23/2005 Scott Podsednik     CHA AL  9




Grand Slams

Elmer Smith         CLE AL 10/10/1920   1 
Tony Lazzeri        NYA AL 10/02/1936   3 
Gil McDougald       NYA AL 10/09/1951   3 
Mickey Mantle       NYA AL 10/04/1953   3 
Yogi Berra          NYA AL 10/05/1956   2 
Bill Skowron        NYA AL 10/10/1956   7 
Bobby Richardson    NYA AL 10/08/1960   1 
Chuck Hiller        SFN NL 10/08/1962   7 
Ken Boyer           SLN NL 10/11/1964   6 
Joe Pepitone        NYA AL 10/14/1964   8 
Jim Northrup        DET AL 10/09/1968   3 
Dave McNally        BAL AL 10/13/1970   6 
Dan Gladden         MIN AL 10/17/1987   4 
Kent Hrbek          MIN AL 10/24/1987   6 
Jose Canseco        OAK AL 10/15/1988   2 
Lonnie Smith        ATL NL 10/22/1992   5 
Tino Martinez       NYA AL 10/17/1998   7 
Paul Konerko        CHA AL 10/23/2005   7

The Birth Pangs?

Last weekend we visited a church in Colorado Springs where the pastor was in the midst of preaching a series on the book of Revelation. Personally, I feel that there is so much mystery, symbolism, and ambiguity in the book, that the topic is better left alone from the pulpit unless it is to educate on the various eschatological positions out there. Be that as it may I always enjoy listening to a sermon on the topic.

In this case it became immediately clear (although he didn't make it explicitly so) that the pastor was preaching from a dispensational premilennialist and pre-tribulation position - that is - the position that there will be a "rapture" of Christians before a tribulation period in which the antichrist is revealed and conquered followed by a literal 1,000 year reign of Christ on earth and culminating in the final judgment. This view also takes literally many of the scenes painted in the book of Revelation including the "mark of the Beast" and the 144,000 witnesses. This view was popularized in the 1830s and is the dominant view of the evangelical community as evidenced by the popularity of the Left Behind series of books by Tim LaHaye and Jerry Jenkins.

He used as his text Revelation 6 which discusses the opening of the seven seals by the angel. What I found interesting was that he used the Olivet Discourse in Matthew 24, specifically verses 4-14, as general signs of the coming of the rapture and hence the tribulation. In that passage Jesus' disciples ask him what will be the signs of his second coming.

And Jesus answered and said to them, "See to it that no one misleads you. For many will come in My name, saying, 'I am the Christ,' and will mislead many. You will be hearing of wars and rumors of wars. See that you are not frightened, for those things must take place, but that is not yet the end. For nation will rise against nation, and kingdom against kingdom, and in various places there will be famines and earthquakes. But all these things are merely the beginning of birth pangs." (Matt. 24:4-8)

The pastor then specifically focused on the "birth pangs" phrase and went on to provide evidence that we're fast approaching the beginning of that time in history with the expected and welcome call to repentance.

His "evidence" was in the form of increased earthquake activity and the increase in natural disasters in general. He put on the overhead a chart that looked as follows:

Magnitude 6+ Quakes
1900-1909   3
1910-1919   2
1920-1929   2
1930-1939   5
1940-1949   4
1950-1959   9
1960-1969  13
1970-1979  51
1980-1990  86
1991-1996 100

There was no source noted nor did he explain where these numbers came from.

Unfortunately, for the pastor this view that seismic activity is on the increase isn't really supported by good data. Back in 1999 the Christian young-earth creationist group ICR published an article on what they called this Christian "urban legend" and showed using data from 1900-1999 that large magnitude earthquakes are actually on the decline overall and that earthquakes tend to follow a periodic trend as shown in their graph below:



Since 1999 the trend hasn't changed and following table was published by the US Geological Survey in response to a question as to why we're having so many earthquakes.

 Magnitude      2000    2001    2002    2003    2004    2005

 8.0 to 9.9        1       1       0       1       2       1
 7.0 to 7.9       14      15      13      14      14       9
 6.0 to 6.9      158     126     130     140     140     121
 5.0 to 5.9     1345    1243    1218    1203    1509    1429
 4.0 to 4.9     8045    8084    8584    8462   10894   11085
 3.0 to 3.9     4784    6151    7005    7624    7937    6282
 2.0 to 2.9     3758    4162    6419    7727    6317    3483
 1.0 to 1.9     1026     944    1137    2506    1344      21
 0.1 to 0.9        5       1      10     134     103       0


No Magnitude    3120    2938    2937    3608    2939     679


      Total    22256   23534   27454   31419* 31199 * 23110

In answer to the question the USGS also noted that...

"Although it may seem that we are having more earthquakes, earthquakes of magnitude 7.0 or greater have remained fairly constant throughout this century and, according to our records, have actually seemed to decrease in recent years. A partial explanation may lie in the fact that in the last twenty years, we have definitely had an increase in the number of earthquakes we have been able to locate each year. This is because of the tremendous increase in the number of seismograph stations in the world and the many improvements in global communications. In 1931, there were about 350 stations operating in the world; today, there are more that 4,000 stations and the data now comes in rapidly from these stations by telex, computer and satellite. This increase in the number of stations and the more timely receipt of data has allowed us and other seismological centers to locate many small earthquakes which were undetected in earlier years, and we are able to locate earthquakes more rapidly."


So the perception that there are more earthquakes in these "last days" is just that - a perception - based on the fact that increasing technology allows for better detection and quicker reporting of the quakes that do occur. It is likely that the pastor got his data from either an antiquated or biased source that didn't take these factors into account.

Although I don't have a cool graph to support it, I did hear a report on NPR the other day on the seeming increase in hurricanes in this and recent years. The gist of the answer by the expert they had on was that while this is a big year there have been other years like this in the 20th century such as 1933. The point is that this also is cyclical phenomena as has been reported in the media recently driven by warmer currents in the Atlantic. The most recent cycle started around 1995 and if history is a guide, will last another 10 to 20 years.

In the final analysis I think it was not only a bit presumptuous on the part of the pastor to appear so sure he knows exactly what God has in mind, but also intellectually lazy to present so-called evidence that is easily refuted by 10 minutes of Googling.

Game 2 Musings

More musings while watching the World Series game two...

Lane the Speedster - Strategic stolen base by Jason Lane in the second. He got a great jump on a left-handed pitcher and without good speed still made it easily. That put a runner in scoring position for a hot hitting Brad Ausmus. As it was Ausmus hit the dribbler to third and Adam Everett got to "hit" if you can call it that. He's so messed up that he looked at his third base coach apparently hoping that he was to sacrifice with two outs. He then showed bunt on a 1-1 pitch for a called strike to put him in a big hole (although one could argue that he walks to the plate in a big hole). He then waves through a high fastball. Well, at least he's cleared for the third.

Major Leaguers? - Terrible outfield play by Chris Burke-tober in the 3rd misjudging A.J. Pierzynski's fly to left. He did the same thing in the Cardinals series and strangely looks like a converted infielder. The desire to get Burke's bat in the lineup may end up costing Phil Garner, especially since that desire is predicated on a few good at bats for a player who is not really that good (small sample size). Aaron Rowand also inexplicably wasn't on third but stood on first assuming the ball would be caught as he told his manager between innings. I agree with Tim McCarver that it was the wrong play. Then Craig Biggio drops a popup to let Pierzynski score after the single. Petitte's getting no help. Both runs should be unearned but both are earned.

Where's Lidge? - Well, the umpires continue to have one of the worst post seaons in memory and Jeff Neslon can now be added to the list. Be that as it may, my 10-year old daughter who was lazily watching the game off and on from the couch while reading a book had asked me a few minutes before Jermaine Dye was "hit" with the pitch who the Astros best pitcher was. I responded that Brad Lidge was. After the homerun, as she was heading off to bed asked why the Astros didn't bring in their best pitcher in that situation. I had no response. As it turned out Lidge may have given up the long ball as well seeing as how he's now given up homeruns to two of the last three batters he's faced and this one to a guy who hit zero in 568 plate appearances in the regular season.

Jenks Jinxed? - Bobby Jenks wasn't throwing as hard tonight and his location was obviously not good. The book on Jeff Bagwell is to go up the elevator with him and being ahead in the count he should have. Instead he got one thigh high on the outside corner and singled to start the rally. Burke did a great job in being patient to get the walk and executed a perfect slide to tie the game at 6 in the ninth.

Long Ball Again - The Small Ball/Smart Ball Sox again win on the strength of the long ball - 2 in this case. 16 in 10 post season games and four in the two World Series games accounting for seven of their twelve runs. Homeruns are the smartest way to score - and win - after all.

Senor Tumbalina

While watching game one I tried a new technique for charting pitches. I was going to use it to compare Roger Clemens and Jose Contreras but since Clemens only went two that didn’t work out so well. Anyway, Contreras went 7 innings and one batter so here’s my summary given the data I collected.

Inning One
Contreras starts off each of the three hitters with fastballs (2 called strikes and a ball) and hits 96 on the gun on his sixth pitch to Craig Biggio on which he grounds out to short. He gets Willy Taveras to pop out to short on his first forkball of the night on pitch 10 and then uses the forkball on the final four pitches to Lance Berkman, the last of which comes in at 78mph and catches Berkman looking.

Inning Summary: 15 pitches, 9 fastballs (93.4mph), 5 forkballs (79.6mph), 1 slider (88mph)
Game Summary: 15 pitches, 9 fastballs (75%), 5 forkballs (33%), 1 slider (6.7%)

Inning Two
Starts each of the four hitters with fastballs and throws just one forkball in the inning, the third pitch to Jeff Bagwell which hit him. Mike Lamb hits a homerun to center on an 0-1 93mph fastball down and in. Fastest pitch is the 95mph fastball to start Bagwell. Seems to be making a conscious effort to pitch Baggy up and in with hard stuff.

Summary: 9 pitches, 8 fastballs (91.9mph), 1 forkball (78mph), 0 sliders (-mph)
Game Summary: 24 pitches, 17 fastballs (75%), 6 forkballs (25%), 1 slider (4.1%)

Inning Three
Longest inning of the game for Conteras as he throws 19 pitches to six hitters. He starts the leadoff hitter Brad Ausmus with a 92mph fastball and then Ausmus singles to right on the next pitch – a 93mph fastball up and out over the plate. He starts Adam Everett with a 92mph fastball and gets him to roll into a force out on an 88mph slider on the fourth pitch. He seems to fall in love with his slider after that and throws it on four of his five pitches to Biggio. He mixes in a forkball at 78mph that Biggio fouls off before singling to center on an 88 mph slider. He also throws it on the first two pitches to Taveras who helps him out by sacrificing the runners to second and third. Berkman follows with a double to right on a handing 78mph forkball on the first pitch and then finally rights the ship and throws three straight fastballs to Morgan Ensberg who grounds to short after starting him with a forkball followed by a slider.

Summary: 19 pitches, 8 fastballs (92.3mph), 3 forkballs (79.7mph), 8 sliders (87mph)
Game Summary: 43 pitches, 25 fastballs (58.1%), 9 forkballs (20.9%), 9 sliders (20.9%)

Inning Four
He seems to have learned his lesson and throws only fastballs and forkballs getting a strikeout of Lamb and flyouts from Bagwell and Lane. The strikeout of Lamb he gets on his third straight forkball (79mph) down and in. To Bagwell he muscles up on the fastball once again throwing fastballs on five of six pitches all at 94 or 95mph. He gets him on a fastball up and in that Bagwell just misses and flies to left. Lane flies to right on a 93mph fastball on the outside corner. Started all three hitters with fastballs.

Summary: 14 pitches, 10 fastballs (93.7mph), 4 forkballs (80mph), 0 sliders (-mph)
Game Summary: 57 pitches, 35 fastballs (61.4%), 13 forkballs (22.8%), 9 sliders (15.9%)

Inning Five
In the 5th Conteras got Ausmus to fly to right on the first pitch, a 92mph fastball. Everett grounded to third on the second pitch which may have been a slider and Biggio grounded to third on the fourth straight fastball he’d seen. The Astros are exhibiting very little patience. Conteras didn’t need to break out the forkball this inning and all hitters are started with fastballs.

Summary: 7 pitches, 6 fastballs (91.7mph), 0 forkballs (-mph), 1 slider (89mph)
Game Summary: 64 pitches, 41 fastballs (64%), 13 forkballs (20.3%), 10 sliders (15.6%)

Inning Six
Taveras leads off the inning with a line drive double to left on a 92mph fastball up and in on the first pitch. Berkman hits an 0-1 pitch to first and advances the runner to third, both pitches fastballs. Ensberg swings at a 1-0 pitch and grounds to third, both pitches fastballs. To Lamb he goes with a first pitch fastball and follows with three forkballs at 80mph each, the last of which Lamb grounds to second. Once again the Astros don’t work the count. All hitters started with fastballs.

Summary: 9 pitches, 6 fastballs (92.8mph), 3 forkballs (80mph), 0 sliders (-mph)
Game Summary: 73 pitches, 47 fastballs (64.4%), 16 forkballs (21.9%), 10 sliders (13.7%)

Inning Seven
Conteras hits the leadoff hitter Bagwell on another forkball up and in after starting him with an 88mph fastball up he took for a called strike. Lane pops to first on a 92mph first pitch fastball. Ausmus is then hit with a first pitch 92mph fastball. Everett continues not to be patient and swings at an 0-1 fastball and grounds into a double play. Biggio then also swings at the first pitch, an 87mph fastball and grounds to third. Astros getting less patient as the game wears on. Four of the five hitters started with fastballs. The fastball has noticeably slowed.

Summary: 7 pitches, 5 fastballs (90.2mph), 1 forkball (79mph), 1 slider (86mph)
Game Summary: 80 pitches, 52 fastballs (65%), 17 forkballs (21.2%), 10 sliders (12.5%)

Inning Eight
Taveras hits another double to left center on a second pitch 85mph slider. He’s relieved in favor of Neal Cotts.

Summary: 2 pitches, 1 fastball (88mph), 0 forkballs (-mph), 1 slider (85mph)
Game Summary: 82 pitches, 53 fastballs (64.7%), 17 forkballs (20.7%), 11 sliders (13.4%)


For the game Conteras started 24 of the 29 hitters with fastballs. Astros hitters offered at the first pitch 11 times but didn’t work the count after the first pitch seeing just 2.83 pitches per plate appearance (Carlos Silva was tops in the majors last season at just over 3). I was surprised he didn’t use the forkball a bit more throwing it about one in five pitches. It’s possible that a few of those 11 sliders weren’t actually sliders but hard forkballs but his forkball was very consistently from 78-80mph. He got 12 groundball outs of his 21.