"It looks like an infield popup in the boxscore." --Comment about a home run ball snagged by a leaping outfielder.
Besides its towering significance in the standings, still in limbo with a day left in the season, Saturday night's barn burner between the Rangers and Angels was a lesson in how unhelpful even ERA can be, particularly over short stretches like a single season.
In case you hadn't heard, Texas, poised to clinch the AL West before an adoring crowd, broke open a back-and-forth game to establish a 10-6 lead entering the ninth inning. Interim closer Shawn Tolleson, pitching for the fifth straight day, entered the final frame and proceeded to cough up consecutive home runs to Erick Aybar and Kole Calhoun before manager Jeff Bannister correctly cut his losses and handed the ball to veteran reliever Ross Ohlendorf.
After inducing Mike Trout to ground out, Ohlendorf fooled Albert Pujols, who popped up behind first. Poor communication between Mike Napoli and second baseman Rougned Odor led to a collision of gloves and two bags for a hustling Pujols who was inexplicably awarded a double.
Ohlendorf fanned Daniel Murphy, bringing the Rangers to the brink of clinching and reducing the Of Anaheims to playoff life support. He then fooled C.J. Cron into bouncing a 15-hopper to the left of second base that shortstop Elvis Andrus recognized late and failed to retrieve. The Pujols run, which should have been charged to Odor, accrued to Ohlendorf's record.
Next, following a sharp David Freese single, Ohlendorf splattered Carlos Perez's bat on a two-strike pitch. But the ball leaked into shallow center and Cron, who was aboard entirely because of luck, tallied the tying run before a slack-jawed crowd. Charge that run to Ohlendorf.
Up came Johnny Giavotella, who swung awkwardly at a slider and popped a single off the end of his bat, scoring the go-ahead run. Which the record will show was Ohlendorf's fault.
Pitching coach Mike Maddux went to the bullpen after his reliever had failed him, but let's review the record. Ohlendorf allowed one hard hit ball -- Freese's single. His fielders failed to retire Pujols and Cron, and serendipity sang paeans to Perez and Giavotella. Ohlendorf got ahead on every hitter and made the pitches he wanted. Of his 27 offerings, 21 caught the strike zone.
And yet, Ohlendorf is charged with three runs on five hits, a blown save and a loss in two-thirds of an inning. Those searching for explanations will trot out the old moralizing shibboleths: "he isn't clutch," "he isn't a big game pitcher," "he doesn't have heart," "he chokes," and "he wilts on the big stage," when in fact he faced seven batters and should have recorded six outs. The only missing element of the story is a reliever following him and allowing his baserunners to score.
This outing stands as testament to the need for fielding independent pitching statistics and for measures like Deserved Run Average that separate the results of a pitcher's work from what he can actually control. DRA and its cousins are still relatively primitive and may never achieve their goal of teasing out real pitching acumen, but that is a quest worth the effort.
As a coda to this story, the Rangers did nothing to rescue Ohlendorf in their final at bat. With two outs and Elvis Andrus on first, the runner easily swiped second to place himself in scoring position for a potential tying hit. But Andrus slid past the bag and was tagged out to end the game, and consign Ross Ohlendorf, who did everything his team could ask to salvage the victory, to goat status.
(Double coda: the "winning pitcher" was Jo-Jo Reyes, who threw one seemingly meaningless pitch to record an out in the bottom of the eighth, before Anaheim's Big Adventure began.)
Showing posts with label ERA. Show all posts
Showing posts with label ERA. Show all posts
15 June 2014
Case Study: Where ERA Falls Short
During discussions about the irrelevance of won-loss records for pitchers, it's been suggested to me that ERA helps balance the equation. W-L speaks of specific clutchness while ERA describes overall performance.
The small grain of truth there is that ERA certainly is vastly superior in describing pitching performance than pitching wins and losses. Indeed, despite its inability to parse defense, relief aid and luck, ERA is still a good approximation of pitching prowess.
A game today between Cleveland and Boston at The Fens showed how ERA can go wrong.
Former Milwaukee closer John Axford entered a 2-2 game in the bottom of the ninth inning for the Indians. Axford sported a decent ERA but an awful 27/18 K/BB tally in 28 frames. That means he's offering up two free passes every three innings to accompany 24 hits, an unsustainable number of baserunners, particularly because his HR rate is unremarkable.
Axford opened the bottom of the ninth by fanning .200-hitter Daniel Nava on a 3-2 pitch before putting AJ Pierzynski aboard on balls. He moved the winning run to second by walking Grady Sizemore before whiffing Stephen Drew on a 3-2 count. On the brink of escaping, Axford uncorked a wild pitch to advance the winning run within 30 yards of fruition. And then he walked the number nine hitter, Jackie Bradley, Jr.
Ordinarily. the odds of the batting team scoring a run with two outs and the bases loaded are roughly one in three. However, without any margin for error -- most notably a walk or HBP -- manager Terry Francona had no choice but to yank his inconsistent reliever.
Scott Atchison replaced Axford and induced a groundout by Brady Holt, allowing the Tribe to escape unscathed. (They later won on a Nick Swisher home run in the 11th.)
And so, the box score says that John Axford successfully hurled 2/3rds of an inning of scoreless ball, lowering his ERA no matter where it stood before. Of course, there's is no such thing as 2/3rds of an inning of scoreless ball. That's like two-thirds of a crime-free life. A nine-run inning could have been 2/3rds scoreless.
John Axford walked three more batters than did Cleveland's other four relievers during their six frames of work. He drove his team to the brink of defeat, saved only by his successor. The defense played no role. There were no unearned runs to complicate the matter. But his ERA declined despite a thoroughly awful performance.
Axford, sporting a jaunty 3.21 ERA, has now allowed 21 walks and 24 hits in 29 innings. That, not his ERA, is why he is a former closer.
The small grain of truth there is that ERA certainly is vastly superior in describing pitching performance than pitching wins and losses. Indeed, despite its inability to parse defense, relief aid and luck, ERA is still a good approximation of pitching prowess.
A game today between Cleveland and Boston at The Fens showed how ERA can go wrong.
Former Milwaukee closer John Axford entered a 2-2 game in the bottom of the ninth inning for the Indians. Axford sported a decent ERA but an awful 27/18 K/BB tally in 28 frames. That means he's offering up two free passes every three innings to accompany 24 hits, an unsustainable number of baserunners, particularly because his HR rate is unremarkable.
Axford opened the bottom of the ninth by fanning .200-hitter Daniel Nava on a 3-2 pitch before putting AJ Pierzynski aboard on balls. He moved the winning run to second by walking Grady Sizemore before whiffing Stephen Drew on a 3-2 count. On the brink of escaping, Axford uncorked a wild pitch to advance the winning run within 30 yards of fruition. And then he walked the number nine hitter, Jackie Bradley, Jr.
Ordinarily. the odds of the batting team scoring a run with two outs and the bases loaded are roughly one in three. However, without any margin for error -- most notably a walk or HBP -- manager Terry Francona had no choice but to yank his inconsistent reliever.
Scott Atchison replaced Axford and induced a groundout by Brady Holt, allowing the Tribe to escape unscathed. (They later won on a Nick Swisher home run in the 11th.)
And so, the box score says that John Axford successfully hurled 2/3rds of an inning of scoreless ball, lowering his ERA no matter where it stood before. Of course, there's is no such thing as 2/3rds of an inning of scoreless ball. That's like two-thirds of a crime-free life. A nine-run inning could have been 2/3rds scoreless.
John Axford walked three more batters than did Cleveland's other four relievers during their six frames of work. He drove his team to the brink of defeat, saved only by his successor. The defense played no role. There were no unearned runs to complicate the matter. But his ERA declined despite a thoroughly awful performance.
Axford, sporting a jaunty 3.21 ERA, has now allowed 21 walks and 24 hits in 29 innings. That, not his ERA, is why he is a former closer.
30 May 2009
Off the Cliff
When Cliff Lee won the Cy Young award with a dominant 22-3 record last season, anyone paying attention had to know that he'd regress this year. So far, he's 2-6. So last year was a fluke.
Not.
In fact, surprisingly, he's pitched almost as well as he did last year. Really.
Lee has a 3.14 RA. That's ERA without regard for whether runs are "earned" or "unearned." Careful observers of the game have noticed that errors are but a small subset of defensive plays that could go either way depending on the fielder. Parsing out only those plays tilts the record in a fairly arbitrary way.
Moreover, the whole system of errors and earned runs is a swirl of capriciousness, from biased official scorers to transparently silly rules. Here's just one example: An error on a two-out play renders all succeeding runs "unearned," even those that are clearly the pitcher's responsibility. Even worse: pitchers get an ERA pass for runs created by errors they have committed.
So Lee's 3.14 RA, which ranks seventh in the AL this year, would have placed him third at the end of last season behind Daiske Matsusaka and...Cliff Lee. But his low 2009 RA is deceiving: he stumbled out of the gate, giving up 11 runs and 17 hits in his first two starts. Since then, he's channeled the '08 (not to mention the '05) Cliff Lee, relinquishing a stingy two runs a game.
But even RA has much to apologize for. It makes no provision for defense and is heavily influenced by the quality of relief pitching. It doesn't account for the parks a pitcher pitches in or the quality of opponents he faces.
Fortuntely, Progressive Field (nee': The Jake) is a fairly neutral park for run scoring and Lee's opponent roster, while considerably easier than Roy Halladay's -- who pitches in the brutal AL East --is nonetheless unremarkable. The Indian defense this year, according to the advanced metrics devised by the eggheads at Baseball Prospectus, has been just a tad below average. So let's dig deeper and see how he's doing.
Lee is eating 6 2/3rds innings per start and striking out 3 1/2 times as many batters as he walks. His 6.5 K per nine innings is another positive. Lee is putting way too many runners on base -- his 1.42 WHIP borders on lousy -- but he's offsetting some of that by keeping the ball in the park. Four home runs in 74 innings is excellent, and reminiscent of last year.
All this indicates that Lee is taking care of business on his end -- walks and home runs are entirely the province of the pitcher -- but being victimized by seeing-eye singles and bloop doubles. Ironically, this is the resume of a ground ball pitcher, but Lee tends towards the flyball. I don't really know what to make of that, but it appears all in all that his strong showing is the real thing.
Still, 2-6 stinks, right? Sure, but it's a team game, and the team around Lee in Cleveland gives him less support than his jock strap. In his last five outings, the Indians have plated a grand total of seven runs, en route to the worst record in the AL so far. Their power core -- Hafner, Sizemore, Peralta and Garko -- have socked 18 home runs combined, two more than the Rays' Carlos Pena.
In baseball, luck tends to even out over time. Lee withdrew a lifetime's share from the karma bank last year, so maybe he's paying it back in '09. But pitching deep into games, striking guys out and keeping the ball in the park are a great foundation for success. If he keeps it up, he'll have a great year, whether or not his won-loss record reflects it.
b
Not.
In fact, surprisingly, he's pitched almost as well as he did last year. Really.
Lee has a 3.14 RA. That's ERA without regard for whether runs are "earned" or "unearned." Careful observers of the game have noticed that errors are but a small subset of defensive plays that could go either way depending on the fielder. Parsing out only those plays tilts the record in a fairly arbitrary way.
Moreover, the whole system of errors and earned runs is a swirl of capriciousness, from biased official scorers to transparently silly rules. Here's just one example: An error on a two-out play renders all succeeding runs "unearned," even those that are clearly the pitcher's responsibility. Even worse: pitchers get an ERA pass for runs created by errors they have committed.
So Lee's 3.14 RA, which ranks seventh in the AL this year, would have placed him third at the end of last season behind Daiske Matsusaka and...Cliff Lee. But his low 2009 RA is deceiving: he stumbled out of the gate, giving up 11 runs and 17 hits in his first two starts. Since then, he's channeled the '08 (not to mention the '05) Cliff Lee, relinquishing a stingy two runs a game.
But even RA has much to apologize for. It makes no provision for defense and is heavily influenced by the quality of relief pitching. It doesn't account for the parks a pitcher pitches in or the quality of opponents he faces.
Fortuntely, Progressive Field (nee': The Jake) is a fairly neutral park for run scoring and Lee's opponent roster, while considerably easier than Roy Halladay's -- who pitches in the brutal AL East --is nonetheless unremarkable. The Indian defense this year, according to the advanced metrics devised by the eggheads at Baseball Prospectus, has been just a tad below average. So let's dig deeper and see how he's doing.
Lee is eating 6 2/3rds innings per start and striking out 3 1/2 times as many batters as he walks. His 6.5 K per nine innings is another positive. Lee is putting way too many runners on base -- his 1.42 WHIP borders on lousy -- but he's offsetting some of that by keeping the ball in the park. Four home runs in 74 innings is excellent, and reminiscent of last year.
All this indicates that Lee is taking care of business on his end -- walks and home runs are entirely the province of the pitcher -- but being victimized by seeing-eye singles and bloop doubles. Ironically, this is the resume of a ground ball pitcher, but Lee tends towards the flyball. I don't really know what to make of that, but it appears all in all that his strong showing is the real thing.
Still, 2-6 stinks, right? Sure, but it's a team game, and the team around Lee in Cleveland gives him less support than his jock strap. In his last five outings, the Indians have plated a grand total of seven runs, en route to the worst record in the AL so far. Their power core -- Hafner, Sizemore, Peralta and Garko -- have socked 18 home runs combined, two more than the Rays' Carlos Pena.
In baseball, luck tends to even out over time. Lee withdrew a lifetime's share from the karma bank last year, so maybe he's paying it back in '09. But pitching deep into games, striking guys out and keeping the ball in the park are a great foundation for success. If he keeps it up, he'll have a great year, whether or not his won-loss record reflects it.
b
11 August 2008
It's Smarter to Be Lucky Than It's Lucky to Be Smart
Humans like to ascribe causes to effects. We like to connect meaning to events. We're loath to admit how much of life is simply dumb luck, and how often it is not simply "the residue of design."
That's why mis-employing statistics in baseball can really distort our perspective on actual performance.
I'm watching the Braves-Mets game in Atlanta. (Take I-26 west to Columbia for 100 miles and then west on I-20 for about 150 miles.) One out in inning two and Tim Hudson's pitch entices Carlos Delgado's left hip and right elbow to snap in opposite directions. The result is a puff ball that bubbles just to the right of second base. Shortstop Yunel Escobar gets a late jump and the ball scoots under his glove.
"Renteria would have had that," I noted to Tater, who doesn't seem to remember who played short for the '07 Braves.
Rightfielder Ryan Church follows that by grounding one under the mis-timed dive of second-baseman Martin Prado. Delgado rumbles safely to third when Chipper Jones goes to the wrong side of the runner on Jeff Francouer's throw.
"That kind of defense," I mutter to Tater, "is going to inflate Brave ERAs." She seems more focused on the ongoing petting.
Angel Pagan bats with first and third and smacks a double-play ball right at Escobar. He flips to Prado who fumbles the transfer, allowing Delgado to score and the inning to continue.
So Tim Hudson, a groundball pitcher, lures three consecutive batters into hitting groundballs, every one of which should have been an out. Instead, he's given up two hits and an earned run. Pagan, who did the only thing he had to avoid gets an RBI. Delgado, who was fooled worse than a Mugabe supporter, is credited with a hit and a run scored despite a semi-sorta-slide into third that would embarrass a Little Leaguer.
Multiply that by 162 games, and it's easy to see how much of a role luck can play in team results and personal statistics alike.
That's why mis-employing statistics in baseball can really distort our perspective on actual performance.
I'm watching the Braves-Mets game in Atlanta. (Take I-26 west to Columbia for 100 miles and then west on I-20 for about 150 miles.) One out in inning two and Tim Hudson's pitch entices Carlos Delgado's left hip and right elbow to snap in opposite directions. The result is a puff ball that bubbles just to the right of second base. Shortstop Yunel Escobar gets a late jump and the ball scoots under his glove.
"Renteria would have had that," I noted to Tater, who doesn't seem to remember who played short for the '07 Braves.
Rightfielder Ryan Church follows that by grounding one under the mis-timed dive of second-baseman Martin Prado. Delgado rumbles safely to third when Chipper Jones goes to the wrong side of the runner on Jeff Francouer's throw.
"That kind of defense," I mutter to Tater, "is going to inflate Brave ERAs." She seems more focused on the ongoing petting.
Angel Pagan bats with first and third and smacks a double-play ball right at Escobar. He flips to Prado who fumbles the transfer, allowing Delgado to score and the inning to continue.
So Tim Hudson, a groundball pitcher, lures three consecutive batters into hitting groundballs, every one of which should have been an out. Instead, he's given up two hits and an earned run. Pagan, who did the only thing he had to avoid gets an RBI. Delgado, who was fooled worse than a Mugabe supporter, is credited with a hit and a run scored despite a semi-sorta-slide into third that would embarrass a Little Leaguer.
Multiply that by 162 games, and it's easy to see how much of a role luck can play in team results and personal statistics alike.
Subscribe to:
Posts (Atom)