These Numbers are Disconnectedby John Brattain
March 12, 2008
Time for some mischief. Maybe it will get the Fire Joe Morgan treatment (h/t to ‘Ken Tremendous’ who gave me my first scandal … one of the more enjoyable moments from last season).
I have been reading many verbal grenades being tossed back and forth in recent days between the alleged stathead and anti-stathead (although it should be noted that this group should be deemed the “traditional stathead”) camps. A lot of it has to do with the Cincinnati media’s kissing up to new manager Dusty Baker who, quite frankly, mystifies my occasional sabermetric-apostate sensibilities.
It’s not that I’m in the “traditional stathead” camp—far from it. I fully appreciate the importance of on-base percentage and I feel that generally it is a bad idea to give away outs (yet feel there are times when it is necessary), etc. There is, however, a human element to the game that, while it cannot be fully quantified, it has to be taken into account.
The thing is, there is a disconnect between what has been studied and what I have seen. For example, we read about the limited utility of the stolen base and its effect on run scoring. However, to take that as absolute gospel is to say in effect that Rickey Henderson and Tim Raines would have been superior players (or at the very least, no less valuable) had they consistently stayed on first base whenever they got on.
If the stolen base is a weapon of minimal importance, then why pitch from the stretch? Let the other team steal—the numbers indicate that it doesn’t noticeably improve run scoring. Focus on getting the hitter out and don’t worry about the guys on the base paths.
I have never read any sabermetrician advocate that the benefits of pitching from the windup with men on is so great that it is worth letting the opposing team run at will (poor Will). Why is it somewhat insignificant to ignore the running game on offense while putting so much effort to preventing it on defense?
I was discussing Josh Towers with a colleague and friend who is far more sabermetric-savvy than yours truly (and on my must-read list). I suggested that since Towers threw well from the windup in 2007 (opposing hitters were .266/.301/.339) but was mauled from the stretch (opponents batting .350/.380/.632) that it might behoove (behoove?) Towers to simply pitch from the windup. He thought I was stone cold bat[feces] mad as a bloody march hare (I must admit—not only is he sabermetric-savvy, he’s a shrewd judge of character) for suggesting such a thing.
Players could steal at will (again, poor Will) off Towers. However, if stealing isn’t a significant offensive weapon then where’s the problem, right?
Another disconnect I have and one I hope eventually more research can be done is this: as we know, the sacrifice bunt depresses run scoring over the course of a season. What I want to know is which runs? The thing is, the 2007 Blue Jays caused me to re-think many things I used to take for granted. I don’t like giving up outs but I watched too many scoring opportunities missed by replacement level hitters swinging away.
As we’ve discussed before, not all runs are equal—I think runs 2-5 in a given game are generally the most crucial. (Feel free to correct me on what the span is.) The Jays in 2007 lost 41 games in which the other club scored five runs or fewer. Does the run suppression of the sac bunt (although I do prefer in these situations if the player tries to bunt for a hit and the sacrifice is merely the second-best outcome … but I digress) occur more in the crucial runs or the upper level of run scoring? What I mean is this: does it depress scoring where five runs or less come around to score or does it cause a team to score seven runs rather than eight?
Put another way—do the ‘suppressed’ runs show up more often in 9-7 games or 3-2 games? The Jays may have scored fewer runs in 2007 by getting the Sal Fasanos, the John McDonalds and the Hector Lunas to move runners along but would it have improved run distribution where it may have gotten them a few more wins?
I honestly don’t know. As I said—there is a disconnect between what I know and what I saw occur during the 2007 season.
Getting back to the running game (or for that matter—putting on plays) again, I know what sabermetrics says about the utility of such things, but it doesn’t add up when I compare it to the game on the field. In most sports, it is considered advantageous to make sure the opposing team is ignorant of how you plan on defending against them. Can you imagine in the NFL one team’s defensive coordinator come into the opponent’s offensive huddle to inform them of what kind of defense to expect on a given play? Can you imagine a hockey coach text-messaging his counterpart on how they plan on handling the upcoming power play?
Of course not—it’s asinine.
However sabermetrics seem to advocate a risk-averse philosophy to the game: don’t steal, don’t hit and run, don’t sac bunt with men on first and second with nobody out. Let the opposing pitcher know that you will not try to steal second or try to stay out of a double play and be sure to inform the opposing infield that when two men are on with nobody out that they can feel free to set up at normal double-play depth and not worry about who covers which base in the event of a bunt.
This is considered good baseball strategy. That is the bottom line of this philosophy.
The thing is, sabermetrics tells us the most efficient way to score runs. What has to be borne in mind is, while run scoring and winning tend to correlate (duh) it is not an absolute correlation. A team can score nine runs and lose. They can score one run and win. This is where the human element of the game has to be taken into account.
No one would suggest that a NASCAR team’s success is predicated strictly on the technology in the car. To the best of my knowledge there are no NASCAR fantasy leagues where fans measure the horsepower, torque or RPM of an engine and tally up wins and losses of a given season. Nobody would suggest that the driver is of minimal importance in NASCAR since MPH wins races and not people.
It is the same in baseball. If the game were so easily predicted based on the cumulative totals of each individual player on the roster with extra weight being given to recent performance there would be no need to play the games. There would be no need for advance scouts since certain approaches will guarantee certain results. Back when Yogi Berra was a young catcher, opposing scouts noticed that with a fast runner on first, Berra would call for high fastballs to better enable him to throw out a potential base thief. There was another example (it may have been Rube Walker) where in a tight spot he’d signal for the curveball from his pitcher. His reason? He couldn’t hit a curveball and assumed nobody else could either.
Human foibles like that affect game strategy.
Pitchers get distracted, infielders get fooled, and the ability of the defense factors into what a pitcher will throw. If the catcher is a sieve he may be reluctant to break off his nastiest breaking ball with a man on third with two out. A pitcher may nibble with men on trying to get a strikeout because he’s worried that the middle infield might boot the ball should it be put into play. An otherwise capable relief pitcher struggles when asked to don the mantle of “closer” since he muscles up trying to get more on the ball due to his new role. A batter presses when he’s sitting on 19/29/39/49 home runs in a season or 199/299/399/499 for a career. A pitcher wishes to pitch another inning even though he knows he’s gassed but it’s a tie game, he’s sitting on 19 wins and his final start of the season is against the league’s offensive juggernaut with their ace getting the ball.
It is knowing these things that enables a team to steal a run or a game that may otherwise have been unwinnable. Rickey Henderson causes a pitcher to lose focus on the man hitting causing him to throw a mistake pitch (heck, Rickey drove the Blue Jays' pitchers to distraction in the 1989 ALCS). A hit-and-run creates a hole that turns a double play ground ball into a seeing-eye single. Pat Borders wins the 1992 World Series MVP, not because he batted .450, but because guys like Jack Morris, Juan Guzman, David Cone, Duane Ward and Tom Henke could throw their nastiest forkballs, split finger fastballs and sliders with Braves on third base knowing that the ball wouldn’t go to the backstop. Dave Stewart absolutely owned Roger Clemens—so much so that the Rocket blew a gasket in the 1990 ALCS. Kirk Gibson’s memorable home run in the 1988 World Series was because he knew that with a full count Dennis Eckersley’s security blanket in that situation was his back-door slider.
This is why I stand so firmly on the fence when it comes to the great statistics debate. Baseball is an amazingly complex game with almost infinite possible outcomes in a given situation. Men who have been in the game for decades state repeatedly that several times in a season they’ll witness something that they had never witnesses before. Mathematics is the ultimate truth, the slickest lawyer or the most ingenious accountant cannot nullify that 2+2=4. Baseball is the ultimate sporting expression of humanity where more than raw talent is required to succeed. The more information we have, the better. Despite some protests, neither side possesses the ultimate truth of the game. I hope that one day the factions will embrace each other and work together to explore the game to new heights of understanding.
Our good friend, and THT stalwart, John Brattain passed away on March 24, 2009. John was a prolific writer, whose work can also be read at Sympatico/MSN Sports and Baseball Digest Daily. John's work was also featured at USA Today, MLBtalk, ESPN Insider, Baseball Prospectus, The Baseball Analysts and The Baseball Journals. Never afraid to express himself in any medium, he was also a frequent radio speaker.
<< Return to Article