|2007 Seattle Mariners: The Dangers of Traditional Sabermetrics and Assumptions|
Written by Matt Souders (Contact & Archive) on July 20, 2007
I don't know how many AHP readers also peruse the pages of ESPN Insider.¬† Normally I don't since I don't have the money to buy every subscription site I can get my hands on, but following a terrible start to the 2007 ESPN Fantasy Baseball season, users were given free access to Fantasy Insider so I've been reading the occasional article and getting a sense for the quality of the analysis they present.¬† Naturally I jumped all over their free trial weekend, since most of their material is good and worth reading if you can afford to do so.¬† I was, however, highly disappointed by the half-hearted, lackluster performance of Keith Law in his review of the 2007 Mariners at the midway point in the season.¬† For those of you who want to read that article, you'll need to be a subscriber.¬† I invite you to follow along here if you have access:
I'm going to take this opportunity to provide an alternative viewpoint for your consideration.¬† "ESPN.com's consensus pick for last place in the AL West" (which is in itself a statement to me about how seriously the folks at ESPN take their analysis of those western teams way out yonder...anyone picking Seattle to finish behind Texas isn't using their head even in March) has gotten off to a great start to their 2007 campaign.¬† Is it for real?¬† Let's begin with his discussion of the Mariner bullpen, which by any measure of performance has been one of the two or three best in baseball to date.
"The Mariners' bullpen has been outstanding this year, posting the second-best ERA in the American League, thanks to an incredible record of keeping the ball in the park. Despite throwing 270 innings, fourth-most of any AL bullpen, Seattle relievers have allowed a league-low 14 homers."
The unwritten implication here is that the Mariners' bullpen is nothing special if they allow a few more home runs.¬† Of course he forgets that the entire pen is leading the AL in reliever K/9 and outside of Brandon Morrow and to a lesser extent Sean Green, every one of the regular relievers has above average control as well.¬† These are pitchers having success because they command the strike zone.¬† The fact that they have thus far prevented the long ball as well helps but it does not entirely define their success.
What about his contention that the Mariners have been lucky to avoid taters?¬† I draw your attention to www.thehardballtimes.com and its' valuable batted ball trajectory data.¬† If you select THT Pitching Stats, then filter for the Seattle Mariners' pitchers and specify that the pitchers don't need to qualify for the ERA title to be displayed, you can view the entire Mariner Mariners' staff on one convenient, searchable page.¬† You will see that the Mariners' relievers have indeed been preventing the longball, but in order to claim that this is not a repeatable skill, you first have to demonstrate that their HR/Outfield Fly ratio is significantly lower than their established norms.¬† If you look at the minor league histories of each of these pitchers (go to www.thebaseballcube.com for detailed minor league data) you will see that they are very much in line with their established level of performance.¬† Despite his claims to the contrary this is a bullpen jam packed with pitchers who make their living by keeping the ball down and inducing the groundball.
George Sherrill (the second big arm he conveniently forgot to mention), J.J. Putz, Chris Reitsma, Sean White, Jason Davis, Sean Green, and Jon Huber all live and die to some extent by their sinking pitches.¬† Reitsma has a big overhand curve, sinker, and straight-drop change-up to go with his relatively straight fastball.¬† Davis and White both throw on a high downward plain, creating good sink on their fastball when they're right.¬† Sherrill's slider induces more grounders than anything Glavine or Maddux could muster.¬† Jon Huber throws almost nothing but his sinker.¬† Granted, Sherrill's GB/FB is not particularly high right now (nor is O'Flaherty's ratio), but it's hard to argue with either of their histories of keeping the ball in the park.¬† In fact, none of the Mariner relievers who are scheduled to pitch in the second half have ever shown a tendency toward gopheritis with the possible exception of Reitsma (who is among the team leaders in HR/Fly at 14.1% compared to a league average of 11%).
A case could be made that the Mariners' relievers (who have combined for a HR/Fly of 7.6%) have been a little lucky on their HR rates, but there are two factors working their favor.¬† First, they pitch half of their innings at Safeco Field, which suppresses home runs by as much as 10% (meaning you would expect a HR/Fly closer to 10 or 10.5% than 11% for an average group, perhaps less for this particular bullpen).¬† Second, everyone but O'Flaherty is running enormous K rates, so the number of flyballs they allow is minimized, reducing the importance of HR/Fly.
By the way, I seem to recall comments made about the 2002 Angels' pen along these lines.¬† "They're all fringe talents.¬† We've never heard of these guys, so they're probably not going to stay this good."¬† Five years later, everyone knows who Brendon Donnelly and Scot Shields are.¬† It is a surprisingly easy thing to do (building a bullpen out of spare parts).¬† Sherrill and Putz both kicked around the minors (and in Sherrill's case, the independent leagues) trying to start and then turned into fire breathing monsters when they got a crack at relief.¬† Mark Lowe (last year's Mariner phenom who underwent elbow surgery to repair a structural defect and is nearly ready to rejoin the Mariners) underwent a similar transformation when he jumped from the rotation to the bullpen.¬† Turning marginal talents into effective relievers is relatively easy because pitching for one effective inning at a time is relatively easy for guys who have at least two out pitches.
Now let's move on to the starting rotation.¬† Mr. Law focuses his attention on the Mariners' rotation ERA (third worst in the AL at 5.00 after Tuesday's loss to the Orioles), and makes several proclamations that need further analysis.
"Seattle's ace, Felix Hernandez, has the stuff to be one of the five best starters in baseball, but he has struggled with an elbow problem and a tendency to overuse his fastball, which (despite coming in at 92-96 mph) is probably his worst pitch."
Mariner fans who have been watching Hernandez' starts intently (not to mention THT's great pitch selection charts) will tell you that Hernandez is NOT overusing his fastball with the possible exception of his match-ups against the first few hitters in each start.¬† Hernandez' troubles stem from trying too hard not to walk people.¬† He tends to aim at the center of the strike zone and hope the movement on his pitches is enough to induce bad contact.¬† His 2-seam fastball doesn't move as much as it did before his elbow surgery (though that has been slowly improving) and his tendency to be constantly around the strike zone gives hitters the chance to "see ball hit ball" with a fully loaded swing.¬† If they make contact, they usually hit it VERY hard, leading to the elevated batting average on balls in play (.356) and the high HR/Fly (15.6%).¬† If he allowed himself to walk a few guys every now and then, and pitched more at the edges of the strike zone, the HR/Fly and the BABIP would plummet and he would resume his dominance over baseball.¬† He has made steady progress in this regard over his last several starts and I expect improvement down the stretch.
"Jeff Weaver has been better since coming back off the DL, helped by appearances against three NL teams and the depleted Oakland A's, yet he still has an ERA over 6.00."
Let's not talk as though he hasn't pitched well against good hitting teams too.¬† He kept the bullies of Boston in check and has in general pitched markedly better since his return from the DL.¬† His "ERA above 6.00" is not a statement about how bad he is now.¬† It's a statement about how bad he WAS when his fastball topped out at 86 instead of 92 and he couldn't get any break on his slider.¬† Weaver is no great shakes, but this dismissive comment strikes me as snide sarcasm, not analysis.
In general, the Mariners rotation has been effective when they're not getting destroyed.¬† That sounds like a silly thing to say, but if you take a look at the game logs, you'll find strings of quality performances interrupted by catastrophic blowouts.¬† Weaver got lit up 6 times in a row to start the season.¬† Feierabend was crushed twice spaced between three solid performances.¬† Cha Seung Baek was either really good (games against the Tigers, Indians, Rangers etc) or really bad.¬† We have a tendency to overuse average (mean) based metrics like ERA.¬† Two pitchers with 6 ERAs are not necessarily of the same value (even if you buy that ERA is a good stat on which to base value judgments).¬† If you have the choice between a guy who gives up a steady 5 runs a game and a guy who gives up 3 runs a game 4 times for every 13 run meltdown he hands you, go with the wildly inconsistent pitcher.¬† Inconsistency can be your friend.¬† The Mariners' rotation is about as extreme an example of this as I've ever seen.¬† They are wildly, spectacularly inconsistent, and the result is a team that does far better than you would think given their ERA.
Now let's discuss the offense.¬† Law boldly states that the Mariners' aggressive "Ichiro-like" approach makes them a poro hitting team outside of Ichiro himself.¬† He cites information like Pitches/PA (the Mariners are dead last in baseball in this measure of discipline) and BB rate to make the case that the team is so impatient that they don't stand a chance of consistently scoring runs.¬† Pitches/PA doesn't tell the whole story about a team's patience at the plate any more than BB rate does.¬† They're both excellent indicators normally, but the Mariners break the mould as far as how teams normally drive up starters' pitch counts.¬† It is true that the Ms don't take a lot of pitches in each plate appearance.¬† However, opposing starters, especially the ones who are around the strike zone a lot, don't tend to survive past the sixth inning against the Mariners because their self-defense pepper swings produce one or two baserunners every inning.¬† There are many ways to get extra pitches, one of which is to get extra plate appearances.¬† This is something the Mariners are quite good at (they're above average in terms of PA/Game despite the lack of walks).¬† Analysts used to say that the 1995-1997 Mariners' offense was like a wrecking ball.¬† They would swing back and slam into that wall over and over.¬† Even when they didn't score, they weakened you, preparing for the fatal blow, and when you were tired from the stress of facing them, you would topple and they would put up crooked numbers.¬† The 2007 Mariners are similarly a wrecking ball offense, but in a different way.¬† Those 90s teams walked and hit the longball.¬† This team just gets 10 relentless hits every night.¬† It's only a matter of time before those hits come in sequence and they get the runs they need to hand the ball to Putz.¬† Their offense actually stands a better chance of hitting great pitching in the post-season than (say) the Yankees or As.¬† Great pitchers don't throw lots of walks and don't get behind in the count so the patient BB+HR approach doesn't work.¬† The Mariners get 10 hits off of the weak pitchers and the strong ones alike. They can beat you with 10 singles or a few well placed HRs.
Now, Law claims that it's unlikely that Seattle will continue to hit .280.¬† I would simply ask "why is that unlikely?"¬† Their BABIP is a perfectly normal .303 (the league is averaging .298, but if you take out Ichiro's at bats, the Mariners are too, and Ichiro is always going to have a high BABIP due to his speed).¬† It flies in the face of traditional sabermetric wisdom, but the Mariners are playing a moneyball of their own.¬† They're getting cheap (by market standards) consistent production from all nine line-up slots by playing the odds and counting on DIPS theory.¬† There is absolutely no reason at all to expect Seattle to struggle to hit in exactly the same way they've hit all year.¬† There is this ugly rumor going around that because individual singles hitters tend to be streaky, a team of singles hitters should be streakier.¬† It's not true.¬† The larger the sample of singles hitters you amass, the more their individual streaks and slumps will cancel each other out (or so says the law of large numbers).¬† The Mariners are hitting .280 as a team despite getting surprisingly bad results from Richie Sexson and Yuniesky Betancourt and despite month long slumps from Jose Lopez and Kenji Johjima.¬† There's no reason that shouldn't continue.
Law further claims that the Mariners have been lucky with runners in scoring position, citing their BA w/RISP (second in the Al, currently at .286).¬† Of course, if he'd bothered to check, he'd know that a .286 BA w/RISP is not much higher than their batting average in any other situation.¬† He might also be surprised to know that a positive split with RISP is NORMAL for a contact hitter.¬† The defense changes when there are runners at second or third, increasing the odds that a grounder or line drive will find a hole for a single or double.¬† Contact hitters like Ichiro and Yuni Betancourt (1st and 2nd on the team in BA w/RISP) will always do well with men on base for that reason.¬† The Mariners haven't been lucky at all with men on.¬† They're proving that a contact hitting club can be nails tough when the game is on the line and runners are on base.
Law's summary comments focus on Davenport's PythagenPort W% estimates which claim that the Mariners are 6 games lucky right now due to their lukewarm arm run differential (which is currently only +12).¬† I will now refer you to my own Pythagorean Win Estimator, which I have shown through correlative studies and observation to be a better measure of the real value of past performance and which is thus far proving to be a better predictor of future performance than standard PythagenPat and PythagenPort.¬† Follow this link and check out the Mariners' place on my leaderboard: http://detectovision.com/?p=1054
The Mariners are, in my opinion, a true 92 win team as it stands right now.¬† They've benefited from very little of this luck of which Law speaks.¬† The difference between PythagenMatt and PythagenPort is almost entirely explained by the inconsistency of their starting rotation and the dominance of their bullpen.¬† They're 41-2 when leading after 6 innings and this didn't happen by accident.¬† This is a good club that stands an excellent shot at competing all season despite a seemingly unimpressive RS/RA differential.¬† Far from being "hard to argue" that the Mariners haven't been lucky...I find it quite unlikely that they've gained from any significant good luck.
I encourage you to read both pieces and form your own conclusions.¬† As for me, I'm enjoying an exciting season for my club and the luxury of being right for once (it's a rare thing, I know).¬† I predicted the Mariners would win 93 games back in March.¬† They're amazingly on pace for (wait for it) 93 wins.¬† And they're doing it in much the same way I predicted.¬† With a deep, combative offense, a dominant bullpen, and an inconsistent rotation that fits well with the team's defensive strengths and the idiosyncrasies of Safeco Field.¬† Here's to a good second half under a new manager and here's to 5 more years of Ichiro in center field!
Thanks for reading!