Search THT:

Or you can search by:

October 31, 2012

WPS recap post-mortem:  two plays in Game Two

The World Series is over, a lot sooner than most fans were hoping. We're now stuck in that twilight zone where we can either look forward to the 2013 season, or backward to the season that was. I don't have my glasses with me right now, so five months ahead is just a blur. I'm looking backward, thank you.

Firstly, a couple pieces of unfinished business from WPS Recap. Prime among those is to congratulate the San Francisco Giants on their victory. May that flag fly forever.

Next, Game Four was exciting enough to rescue the World Series from being the least interesting Series of all time, as measured by the WPS Index. I failed to mention where it did end up: fifth least-exciting all-time, better than the 2007 Red Sox sweep, but not quite as thrilling as the 1928 Yankees sweep. I could nudge 2012 ahead on the basis of Pablo Sandoval's three-homer game to begin the Series ... except that Babe Ruth had a three-homer game to end the 1928 Series. Fifth place it is.

Now for the post-mortem that the title promised. The crucial game of the Series, if a sweep can be said to have one, was Game Two. Scoreless through the seventh-inning stretch, it was there for either team to take. Had Detroit done so, the complexion and psychology of the series changes markedly, and who knows what happens. That means it's time for, yep, second-guessing!

(Technically, I did touch on one of these matters in the original WPS Recap. I nearly made a between-game post of the second one, but I thought I had been writing quite enough THT Live articles. There, now you can second-guess me on something: that's The Circle of Life.)

The first turning point came as the Tigers rallied in the top of the second. With a hit-by-pitch Prince Fielder on first, Delmon Young doubled into the left field corner. When the ball ricocheted away from Gregor Blanco, third base coach Gene Lamont waved Fielder home. Blanco's long throw sailed over Brandon Crawford, but Marco Scutaro had trailed the play. He reined it in and threw on to Buster Posey, whose quick sweep tag got Fielder a foot short of the plate.

I reported on Thursday night that the break-even mark for sending Fielder in that situation is 87.2 percent (given the 2012 Run Expectancy numbers). Here's another figure to give you some perspective: The break-even mark this season for stealing third base with two outs was 87.8 percent. One of the archetypal bonehead plays in baseball; something players are coached to avoid and lambasted for forgetting; one of those old saws from the proverbial book that is actually dead-on correct. That play is just a tiny bit worse than sending the runner home on a no-out double.

Now, it is possible to be too hard on Gene Lamont for this snap decision. The factor that made sending Prince Fielder so obviously dubious—that it's Prince Fielder, lugging Prince Fielder's weight around the bases—would have worked against him had he halted at third. He would have had a similarly diminished chance of reaching home on a two-hop grounder or a medium fly ball. We could conceivably be lamenting how Lamont clogged up a rally by holding Prince at third, and how three straight teammates failed to bring him home. (Delmon Young got stranded on second, after all.)

It was still a mistake, but it was a mistake of aggressively going for an early run. Given how Detroit's bats were limp noodles for most of the series, it's easier to forgive, or at least understand, in retrospect. Of course, most of that power outage was in the future when Lamont windmilled Fielder home. It was a bit early to be acting desperate. On the whole, it was a blunder, but not something so stupid that it should haunt Gene Lamont forever. (Given the sweep, it probably won't. Had this happened in a Game Seven, though ...)

The second pivot point came in the bottom of the seventh inning, the game still scoreless. San Francisco loaded the bases on a single, a walk and a sacrifice bunt that turned into a hit when it rolled to an unmolested halt a few inches inside the third-base line. Bases loaded, no outs, tie game: That's a jam.

Detroit manager Jim Leyland had two options: He could play the infield in, hoping to cut off the run while risking a greater chance at a big rally, or set the infielders at double-play depth, trading one run for two outs. He did the latter, and got what he was playing for, a 4-6-3 twin killing that still made it 1-0 Giants. He strongly defended his tactical choice after the game. "We were absolutely thrilled to come out of that inning with one run," he told reporters.

One post-game analysis, by Ben Lindbergh at Baseball Prospectus, supported Leyland's call. He used Run Expectancy to compare the presumed optimum results of the defensive orientations: two out, man on third, one run scored for double-play depth; one out, bases loaded, no runs scored for playing in. The double play dropped the Giants' RE from 2.260 to 1.363, while a force at home would have lowered it from 2.260 only to 1.537. Lindbergh thus advised everyone to put away their pitchforks.

I'm going to keep something pointy at hand, myself.

Run Expectancy is an excellent tool for measuring general situations, especially those early in a ballgame. When the game is late and close, however, its general applicability gets overwhelmed by the specific situation. When that happens, analysis is served better by looking at the more complex but more precise Win Expectancy numbers.

Fortunately, there's an app for that, and it's right here at The Hardball Times. I fed the situations into Dave Studeman's Win Probability Inquirer. I assumed a run environment of 4.0 runs per team per game—the average in the majors this year was 4.32, and AT&T Park is a pitcher's park. These were the results.

Base/Out/Score Situation        SF's Win Probability
Bases loaded, no outs, 0-0             0.830
Man on third, two outs, 1-0            0.796
Bases loaded, one out, 0-0             0.738

Both optimum results improved Detroit's chances, but the comparison is clear. Getting the force at home gives Detroit a 9.2 percent boost; the run-scoring DP was just 3.4 percent. (I also checked the plays with a 3.5 run environment, and the spread was even wider.) It's difficult to balance the probabilities of getting these optimum results with the risks that both defenses, especially the infield in, offer. If you are hoping for the best, though, the infield in gives you a much superior best to hope for.

There's one other factor in play. Leyland was content to play a run down to San Francisco, a team whose bullpen was one of its greatest strengths. Maybe his team's two-run homer while down seven in the ninth the previous night affected his estimation of the Giants' pen, but one can argue he should have been working to avoid having any deficit to make up against those relievers. As it unfolded, Detroit did not score against the bullpen Giants, in that game or for the rest of the World Series.

That is how Game Two got away from the Detroit Tigers: a little too aggressive in the second, a little too conservative in the seventh. It's tough and kind of unfair to argue perfect causality in a contingent game like baseball, but a run here, a run there, and pretty soon that's the ballgame. And just maybe the whole season.
Posted by: Shane Tourtellotte

October 29, 2012

WPS Recap: World Series, Game 4

The 2012 World Series avoids the fate of being judged by WPS as the least exciting World Series ever. As for avoiding the sweep, well, we can't have everything.

Game       1   2   3   4   5   6   7   8   9  10    F
Giants     0   1   0   0   0   2   0   0   0   1    4
Tigers     0   0   2   0   0   1   0   0   0   0    3
(Giants win World Series, 4-0)
WPS        1   2   3   4   5   6   7   8   9  10 
Giants     5  29  17  13   7  44  21  24  14  59
Tigers     7  14  31   5  11  32   9  24  20  20
WPS Base: 402.7  Best Plays: 87.5  Last Play: 4.7  Grand Total: 494.9

After a drought of excitement that turned the tail end of this postseason into a debatable and desert land, we finally got a really good game at the last possible moment. Not quite a great game—my cutoff there is at 500—but well worth watching.

The WPS line score shows how this contest built up its numbers. Few 1-2-3 innings to depress scores, meaning a lot of innings that had at least one baserunner and, thus, some scoring threat. The top of the ninth saw the Giants go down in order, but it was late and close enough that this still produced excitement, moreso than a couple earlier innings with someone getting aboard. Despite some interesting rallies late, notably in the seventh through ninth, it was still the scoring innings that produced all the highest WPS numbers.

I won't get much into covering the highlights here—you likely know them by now—but I do still have observations on this final game of the season. One is how Bruce Bochy declined to bring closer Sergio Romo into the game in the ninth inning, needing to keep the Tigers off the board to extend the game. Sabermetricians consider this the fallacy of the save situation: in waiting for the closer's standard situation and sending out inferior pitchers in the last half of the ninth (or later), you make it all the less likely that you'll get to that save situation.

It worked perfectly for Bochy, though. Jeremy Affeldt and Santiago Casilla handled the ninth, and when Marco Scutaro drove Ryan Theriot home in the visitors' 10th, Romo came in to administer death by slider (and sinker to one lefty). The doctrinaire sabermetricians can perhaps comfort themselves that it was Miguel Cabrera caught looking by the final strike. Mike Trout gets a modicum of preemptive revenge.

Bochy also splashed some cold water on the suggestion I made in yesterday's recap, stating categorically after Game Three that Tim Lincecum will be back in the starting rotation for 2013. I guess my suggestion didn't pass his horse-laugh test.

My Tim-ism of the night came as the top of the ninth led off. I cannot swear to the exact quotation, but Tim McCarver said something along the lines of "If [Hunter] Pence can get on base, it would go a long way toward helping the Giants score in this inning." Correct. Tautologies for $400, Alex. This wasn't as egregious as saying the chance of scoring goes up "dramatically," but it's in the same spirit.

I could almost have forgiven McCarver if he had been a little better with a good line earlier. As a replay of one of the myriad broken bats of this series played, McCarver observed, "Kindling for the winter." That line was a clean single, but "Kindling for the hot stove" would have cleared the fence.

One casualty of the San Francisco sweep is that we won't get to see how far Joe Buck was going to carry his torch for The Who, after getting their music into broadcast intros and between-inning outros. I was actually looking forward to the inevitable dugout interview with Pete Townshend in Game Six.

And I am glad to report that our favorite Marlins jersey-wearing fan was at the game! Laurence Leavy sure does get around, and the World Series was that much more interesting because of him. He even broke out his white panda hood in the eighth inning when Pablo Sandoval came up, but the charm did no good, as Panda grounded into a pickle-licious double play. I have to wonder, though, who he apparently was texting during the game. The one time this series served up a hot dish, and his attention was split.

And that is it for WPS Recap this year. Whether I do this again next season depends on how fast the trauma of actually replicating a sliver of the life of a sports reporter fades. One last time, I would like to acknowledge FanGraphs for its steady flow of real-time game data that made these reports possible.

There may be one sequel to these recaps: I might be doing a slightly deeper analysis of two of Detroit's tactical moves in Game Two that could have cost them greatly. (I nearly wrote this up on Saturday, but I figured two THT Live articles in 24 hours would be trying our readers' patience.) The Series is over, after all: the time is perfect now for all those "what ifs."
Posted by: Shane Tourtellotte

October 28, 2012

WPS Recap: World Series, Game 3

Game Three of the World Series is over ... and suddenly, the Series itself almost is.

Game       1   2   3   4   5   6   7   8   9    F
Giants     0   2   0   0   0   0   0   0   0    2
Tigers     0   0   0   0   0   0   0   0   0    0
(Giants lead series 3-0)
WPS        1   2   3   4   5   6   7   8   9 
Giants     7  40   4   9   4   6   7   8   2
Tigers    17   5  21   9  42  11  13  14   9
WPS Base: 227.4  Best Plays: 31.6  Last Play: 1.5  Grand Total: 260.5

The most exciting game of this World Series still could not reach the median WPS Index of 300. The culprit is the iron grip that the first team to score (meaning the Giants) puts on the game. No lead in this Series has yet been lost. In fact, in the last seven Giants games, and the last six Tigers games, the first team to score never has relinquished its lead. This goes very far in explaining how their League Championship Series and this World Series have put up such poor WPS numbers.

Anibal Sanchez arguably was the better of the two starting pitchers on the night. He went seven innings, throwing 117 pitches, and allowed only seven baserunners. Ryan Vogelsong didn't get out of the sixth and put nine men on base, including first and second with one out in three of his five complete frames. (The fifth went to bases loaded, one out.) Vogelsong, though, got two key double plays to snuff early rallies and bore down in the fifth to strike out Quintin Berry and induce a pop-up from Miguel Cabrera.

Sanchez, on the other hand, committed the cardinal sin: he bunched his hits, at least two of them. After a solid first, his control in the second deteriorated, not to Nuke LaLoosh levels by any means, but definitely into Ricky Vaughn territory. Several balls went well away from the zone, including a leadoff walk that seriously mattered and a wild pitch that didn't quite. Gregor Blanco's triple mooted the wild pitch, and Brandon Crawford drove Blanco home. This opened, and closed, the scoring for the game.

The bullpens were better than the two very good starters, sealing the decision. Tim Lincecum gave up only a walk and a reached-on-error in his 2.1 innings and is posing a conundrum for the Giants next year. He has struggled badly as a starting pitcher in 2012 but has been dominant in several multi-inning relief appearances in October. My mind goes back to another pitcher who wasn't cutting it as a starter but blew people away in postseason relief appearances and got his team's brass thinking. The year was 1995, and the pitcher was Mariano Rivera.

This is a far more complicated choice. Lincecum had great success starting before this year, something Mariano never had. But "had" might be the operative tense, and it could be that Lincecum could now best be used as an old-fashioned, Rollie Fingers/Bruce Sutter fireman, throwing more than one inning, closing or getting the game to a closer as the situation fits. It would mean a paradigm shift in how to use a $20 million starter and in how to deploy an ace reliever. But that it's possible to propose such radical changes and not be horse-laughed out of the room is one of the most intriguing subplots of the 2012 postseason.

I mentioned an error while Lincecum was pitching. Two batters earlier, the Fox broadcasting team praised Giants shortstop Crawford's fielding, including 19 straight errorless games. This is right where you expect the jinx to kick in, and "kick" was the right word when Delmon Young's grounder eluded Crawford's glove and forced a bad throw to first. Of course, if the hoodoo doesn't happen, you forget they ever talked about it, but where's the fun in commenting on that?

Not that the Fox crew can be too smug in talking about others' errors. The graphics guy made a couple of boners, putting up Berry's playoff numbers with Austin Jackson at bat and hilariously identifying the reliever in San Francisco's bullpen as Gregor Blanco. (It was actually Sergio Romo, who is a loose enough character that you could almost believe he arranged that as a joke.)

There was even a mild Tim-ism tonight. After Buster Posey made a wonderful stab with his glove to corral a way wide pitch by Lincecum, McCarver said that Posey's play took "an awful good set of hands." Yep, a set of hands, all one of them. Buster's right hand might as well have been in another county as far as that play was concerned, but praise, even deserved praise, knows no factual bounds sometimes.

The Detroit fans pulled hard for their club most of the night, but the wear showed late. They poured out a generous bowl of Boo-Berry when Quintin struck out swinging to end the home seventh; Prince Fielder and Andy Dirks got the same treatment for their whiffs in the eighth. I was reminded of Yankee Stadium crowds at the start of the ALCS, showering boos on most of their offensively hamstrung lineup (Raul Ibanez and Ichiro Suzuki were maybe the only ones spared; Derek Jeter likewise would have been, had he still been ambulatory.) Comerica Park wasn't as dead as Yankee Stadium got during the sweep, but what went around for the Tigers may be coming around.

The Sunday Fox football schedule got plenty of play during the broadcast, and I noted a certain irony. Late in the afternoon, (Tony) Romo will be throwing against the Giants. A few hours later, it could be (Sergio) Romo throwing for the Giants, to win the World Series.

There's at least one more game to go, but it feels like there's only one more to go. Cross your fingers for some sparks: a bad enough game—say, Game One bad—and WPS will score this as the least exciting World Series ever. I, for one, don't want to see that kind of history.
Posted by: Shane Tourtellotte

October 26, 2012

WPS Recap: World Series, Game 2

Game Two of the World Series, and the season is one game closer to its end. Enjoy this while you can, folks: it's a hundred umpteen days to pitchers and catchers.

Game       1   2   3   4   5   6   7   8   9    F
Tigers     0   0   0   0   0   0   0   0   0    0
Giants     0   0   0   0   0   0   1   1   X    2
(Giants lead series 2-0)
WPS        1   2   3   4   5   6   7   8   9
Tigers     4  13   5  15   6  10  20  11   6
Giants     4  23   5   5   6  10  31  11   X
WPS Base: 186.0  Best Plays: 31.6  Last Play: 0.9  Grand Total: 218.5

This one was better than San Francisco's last five games, but still not close to the median index of 300. Why did a tight game produce a low score? Two reasons. One, like those five games before it, the first team to score never relinquished the lead, never even fell back to a tie. The first run may as well have been the last. Two, the pitchers' duel between Madison Bumgarner and Doug Fister was a little too good. They rarely gave the opponents even a chance to score in an inning. Inevitability, even the appearance thereof, is the enemy of excitement.

One of those few early moments of danger came in the Tigers' second, when with none away, Prince Fielder tried to score from first on Delmon Young's double down the left-field line. A good relay by Marco Scutaro and a nifty tag by Buster Posey halted the runaway cement mixer a foot shy of glory. Beyond Fielder's widely-known unfamiliarity with the high end of the speedometer, it was a dubious situation to send a runner. Using Run Expectancy numbers for the 2012 season (from Baseball Prospectus), I calculate that sending the runner has to succeed 87.2 percent of the time to be a break-even play. A green light here was probably a bad idea.

It must be said that this play, and two others like it, squeezed through a loophole in the WPS system to appear less exciting than they were. When base advancement and an out combine in one play, the positive and negative elements cancel, and you're left with a much reduced score. The Fielder play was one example. Two others were the plays where the Giants scored their two runs. Hunter Pence scored in the seventh on a 4-6-3 double play by Brandon Crawford, and Angel Pagan came home in the eighth on Pence's sacrifice fly.

The game may have been decided by the varying quality of the shutout work the starting pitchers were doing. Madison Bumgarner threw just 72 pitches through six innings. Doug Fister had to throw 108 to get through six, and he was pulled after yielding a leadoff walk in the seventh. The Tigers bullpen, called on earlier, showed itself weaker. The first two Detroit relievers walked the first man they faced, contributing markedly to the Giants' runs in the seventh and eighth. Bumgarner, pulled only for a pinch hitter, was backed up by perfection from Santiago Casilla and Sergio Romo.

(By the way, Romo was truly impressive in filling air-time during his dugout interview with Buck and McCarver in the fourth inning. The man has a lot to say, and says it fast. The interview also confirmed that he's a bit of a nut, but this is fairly standard for relief pitchers, and he's probably still No. 2 in the pen behind Brian Wilson.)

Maybe I've started tuning out the announcers, because my Tim-ism file is not too thick. One he did produce was his amazement at Miguel Cabrera's pitch-by-pitch recall of an at-bat in the 2003 World Series where he homered off Roger Clemens, waxing rhapsodic over his incredible memory. Cabrera may have a great memory, but really. If I hit a home run off Rocket in the World Series, it'd be something I'd be reliving often, by myself and with others asking about it. That memory would be kept fresh with repeated recall, plus maybe the occasional look at the video. Let's see him remember an anonymous May 2007 game against the Devil Rays, and then I will be amazed.

I've noticed that, before the bottom of the first in each game this postseason, the PA at AT&T Park has been playing "Eye of the Tiger." I would suggest that the public address staff check the opponents' jerseys ... but really, can you say it's been hurting the home team here?

And finally, our Marlins fan was back in his appointed spot, and it does turn out that he is a Marlins fan, not a Giants backer who threw on the first orange shirt in reach. SB Nation actually interviewed the man, and Laurence Leavy sounds like a pretty good guy, so I gladly withdraw any slights on his character I may have made, suggested, or implied yesterday.

But I stand by calling the Marlins' new logo something out of Lovecraft. I mean, look how fast it drove Ozzie Guillen mad.

We reconvene on Saturday in Detroit. The Tigers, and the WPS Index, could use a nice jolt there.
Posted by: Shane Tourtellotte

October 24, 2012

WPS Recap: World Series, Game 1

And here we are in the World Series. A nice, sedately paced seven games in nine days. It was just a couple weeks ago that we could get eight playoff games in 36 hours.

Once again, WPS Recap will be reviewing the game, starting with my method for measuring the excitement of baseball games. (I am compelled to mention once again that it wasn't quite original with me. Our own Dave Studeman wrote up a very similar method in the 2007 Hardball Times Annual, but I came up with mine in independent ignorance of his.) I gratefully acknowledge FanGraphs for providing the data I use to calculate my Win Percentage Sum numbers. I'm not the only one working during these games.

Game       1   2   3   4   5   6   7   8   9    F
Tigers     0   0   0   0   0   1   0   0   2    3
Giants     1   0   3   1   1   0   2   0   X    8
(Giants lead series 1-0)
WPS        1   2   3   4   5   6   7   8   9
Tigers    17   5  11   8   2   6   1   0   0
Giants    16   4  29   5   2   0   0   0   X
WPS Base: 108.3  Best Plays: 35.4  Last Play: 0.1  Grand Total: 143.8

One of the limitations of the WPS method is that it's built to measure the competitiveness of a baseball game. A game can grow lopsided very quickly, and WPS will see that, but if there is some performance transcending the struggle to win, WPS is not equipped to register it. In the matter of competitiveness, this was a deadly dull baseball game, the fifth straight snoozer for the Giants.

But Pablo Sandoval saved it for the rest of us who are not simple algorithms.

We've been getting a little spoiled lately by all the no-hitters and perfect games being thrown in baseball. Now a similar thing is happening with players hitting three home runs in a World Series game: Albert Pujols last year, and Sandoval in Game One this year. And Sandoval is not particularly a power hitter, with a mere 12 homers this regular season. I won't say he is to World Series triple-bombs what Philip Humber is to perfectos, but maybe he's around Mark Buehrle-level.

Of the five times a player has performed this feat, this was only the second time that he came up later in the game with a chance to hit a fourth. Babe Ruth did it in Game Four of the 1926 Series, drew a walk, and was on deck when the Yankees made their last out of the game. Sandoval had his chance in the seventh, but only managed a single off Jose Valverde, who was in the course of flunking his audition to pitch meaningful innings in this Series. If Pablo couldn't homer off Valverde in his current state, it just wasn't meant to be.

I kept an ear open for a Tim-ism, and caught something I had heard him say before. He made a point of stating that when the first batter of an inning gets on base, the odds of his team scoring that inning "go up dramatically." Well, depends on what you consider dramatic. Assuming the batter reaches first, in 2005 (the most recent year I have data for that I could reach across the desk and grab) the chances for a scoring inning went from 28 percent to 41.7 percent. In 1959-1960 (I think this may have come from The Hidden Game of Baseball), it was 25.3 percent to 39.6 percent. In neither case do the chances go up by as much as half. Maybe it's me, but I expect something more dramatic of out a dramatic rise.

McCarver did a little better later, when the Tigers had a man on third with one out in the sixth and Miguel Cabrera coming up. Tim said that, facing Miggy, you would gladly yield a sacrifice fly, as outs in this situation are more important than runs. True, but not because a Triple Crown winner is batting. I suggest going to THT's own Win Probability Inquirer and running the numbers. It's a win for the pitching team facing an average batter, probably a well below-average batter.

Everyone expected the game to be a pitching mismatch, the awesome Justin Verlander against the cover-your-eyes Barry Zito. They forget that this is baseball, and things get turned upside-down because baseball. I was a lot less surprised than everyone else seemed to be, though you'll have to take my word for it.

Now, there was something at Game Seven of the NLCS that I did not mention, other matters crowding it out, but it happened again and I have to speak my mind. In the stands behind home plate, right above the green-screen ad board, second row, was a fan wearing, in that sea of Giants regalia, a Miami Marlins jersey. One could overlook this once as coming from a brief act of thoughtlessness, but no. He came back with the very same jersey. This was premeditated.

It's obvious what he was thinking: the jersey was orange, a Giants color, so close enough. But remember, it's the Marlins. Their logo possesses every conceivable color, several inconceivable colors, plus for all I know The Colour Out of Space. (Admit it: that new logo has a touch of Lovecraftian horror to it.) It matches with everything, and clashes with everything. I would offer to buy him a Giants jersey, but if he can afford such good seats for these post-season games, he can afford to buy his own!

So it's on to Game Two. We'll see if it manages to confound the WPS system again.
Posted by: Shane Tourtellotte


Click here for more THT Notes.