Tuesday, October 21, 2014

The World Series: Something's Gotta--Well, Probably's Gonna--Break

Well, it's the first all-Wild Card World Series since 2002, when the Giants fell to the Angels in seven games, in a series that features four one-run games (Games One, Two, Four, and Six), two blowouts (Angels 10, Giants 4 in Game Three; Giants 16, Angels 4 in Game Five), and the emergence of a bullpen star, Anaheim's Francisco Rodriguez, who racked up 13 strikeouts in 8.2 innings, which is three more innings than he pitched in the regular season. 

One of the reasons I've been wrong in almost all of my postseason predictions (I did guess that the Orioles would beat the Tigers) is that neither of this team's finalists, the Kansas City Royals and San Francisco Giants, have played true to form:
  • The Royals scored 4.02 runs per game in the regular season, below the AL average of 4.18. The league leader was the Angels, with 4.77. In the postseason, when the quality of pitching is above-average, the Royals have scored 5.75 runs per game, about a run per game more than the best-scoring team in the league during the regular season. And, as I've pointed out, they've done it not with speed, but with the homers, walks, and strikeouts that they avoided during the regular season.
  • San Francisco's best starter, Madison Bumgarner, is unimpeachable. But the No. 2 starter, Jake Peavy, had a 4.72 ERA with Boston before joining the Giants via a trade in late July. He had a 2.17 regular season ERA with the Giants and a 1.86 postseason ERA. The No. 3 starter, Tim Hudson, had an 8.72 ERA over five starts in September and a 3.29 ERA in the postseason. The No. 4 starter, Ryan Vogelsong, was almost as bad as Hudson in September (5.53 ERA) and had a strong start against the Nationals (5.2 innings, four baserunners, one run) before a weak one (four runs allowed in three innings) in a game the Giants won anyway. That four-man rotation has put Yusmeiro Petit in the bullpen, where he's allowed two hits and no runs over nine innings. He had a 4.40 ERA in September.
One of these trends will break down. Either the Royals will stop hitting so well, or the Giants starters after Bumgarner will stop pitching so well. They can't both continue like this.

Both teams have strong bullpens. The Royals, as I noted in my Championship Series preview, had the best record in the American League in games they led after six innings, 65-4. That, of course, is attributable to Kelvin Herrera (usually their seventh inning reliever, 1.41 ERA), Wade Davis (eighth inning, 1.00 ERA), and Greg Holland (ninth inning, 1.44 ERA). But the Giants weren't slouches either, going 62-6 in games they led after six, fourth-best in the National League and sixth-best in the majors.* But these teams were also among the most successful at coming back from deficits after six innings: the Giants were fifth-best in the majors (13-60) and the Royals seventh-best (11-58) in games they trailed going into the seventh inning.** Still, the Royals' trio has combined for a 1.05 ERA in the postseason, giving the Royals, in my mind, an edge in games that they lead.

The question is whether they'll get there. Given my initial premise--that the clock's going to strike midnight for either the Royals' offense or the Giants' rotation--I'm more skeptical of Kansas City, which ranked last in the majors in walks and homers in the regular season. I've made it clear that I'm biased in favor of small-market low-payroll teams, and the Royals' $89.3 million payroll (17th in the majors) is almost 40% lower than the Giants' $145.1 million (sixth). But if I had to make a guess, I'd expect the Royals to wind up on the short end of a low-scoring World Series. 

*Top six won-lost records in games leading after six innings: Padres (!) 60-1 (!!), Royals 65-4, Nationals 72-6, Dodgers 81-7, Twins 52-5, Giants 62-6.

**The Nationals were 14-54 in games they trailed after six innings, the only team to win over 20% of such games. But the bottom of the list is more interesting. The Braves were trailing 63 games after six innings, and came back to win only three--a winning percentage of less than 5%. And the Dodgers were even worse, going 2-54 in games they trailed after six. Dodgers fans have a reputation for leaving games early to beat the traffic, and in 2014, it seems, they were justified, as the team went on the win 92% of the games they led after six while losing 96% of games they trailed after six. 

Monday, October 20, 2014

On The Horror of a Wild Card World Series

Look, I'm not wild about it either. For the second time in baseball history (2002, Angels-Giants), we will have two wild card teams--i.e., teams that couldn't even win their own division--meet in the World Series. That's a far cry from the years though 1968, when a team had to be the best in and eight- or ten-team league in order to make the Fall Classic. Now, you don't have to even win a five-team division.

Except. There is an element of variability in most everything in life, including baseball, that is almost always ignored. Say there's a game, tied 3-3 in the eighth inning. The pitcher throws a ball on the border of the strike zone. The batter doesn't swing. Is it a ball or a strike? 

Well, the average major league hitter had a .251 batting average, .314 on base percentage, and .386 slugging percentage in 2014. After a 1-0 count, that rose to .267/.373/.420. After 0-1, it became .221/.261/.331. 

That difference, in the late innings of a close game, is enough to sway a ballgame. The batter who gets a strike call is far less likely to get on base and score a run than the batter who doesn't. How many times do fluctuations like that occur over the course of a season? Take the Royals. They finished with 89 wins. If you were to replay 2014, over and over again, using the exact same players, same lineups, same injuries--do you think they'd win exactly 89 games every time? I don't. Maybe they'd win 88 or 86 or 92--it all depends on little things like ball/strike calls, a tiny movement in a batter's swing, the precise timing of a fielder's leap. How random are things like that? Maybe, say, 3%? Then over a 162-game season, we'd expect variance of 162 x 3% = 5 games. So the Royals may have won 89 games this year, but it's just as likely that they're an 84-win team that got a little lucky, to the tune of five wins, or a 94-win team that got a little unlucky, to the tune of five losses. Yes, the 89-win Royals knocked off the 98-win Angels and 96-win Orioles, but you can easily build a case that the Royals, if we were to replay the season over and over, might have, in the long run, a better record than the other two. Maybe the Angels and O's are really, in the long run, 93-win teams that got a little lucky this year while the Royals are really, in the long run, a 94-win team that got a little unlucky this year. 

Same in the National League. The Giants won 88 games, and they knocked off the 96-win Nationals and the 90-win Cardinals. But were they an inferior team? Those differences are both within our 3% margin of error. 

I've quoted baseball analyst Joe Sheehan before, but his observation from his newsletter bears repeating: Variance swamps everything. The fact that we're seeing two wild card teams in the Series doesn't mean we're seeing two inferior teams. We're just seeing two teams that didn't compile as good a record over a six-month, 162-game season as the teams they defeated. Maybe they picked the right time to get hot, or maybe they took advantage of their opponents' injuries (the Royals faced an Angels team without its No. 1 starter and an Orioles squad without its starting third baseman, first baseman, or catcher), or maybe some pitchers were more worn down in October than they were in April, or any of many other factors. The point is: If the goal were to insure that the teams with the best record meet in the World Series, then we should return to the pre-1969, pre-divisional play baseball world in which there were two leagues and one champion per league, period. As FanGraphs managing editor Dave Cameron said in his chat last week, "If you crown your champion with a postseason tournament, you’re asking for randomness."

And that's what you get: Randomness. There are two things I hate hearing, incessantly, during the postseason. One is all the myths I addressed last month: that prior postseason experience is important, that having veteran players provides an edge, that momentum matters, that good pitching stops good hitting, that reliance on home runs is bad, that ace starters confer an advantage. All of those, I explained, are dubious. 

But the other thing that I hear a lot, and which is probably more offensive, is that the postseason is a test of character. It's a short series of baseball games, not a morality play, for crying out loud. The Royals and Giants didn't advance because they're better people than the players on the Angels, Nationals, Orioles, and Cardinals. They didn't win because they were able to "put it together," "draw on their inner strength," "function better as a team," "will themselves to win," execute under pressure," "perform when they had to," and whatever other crap you read and hear, while their vanquished opponents didn't. They're each here because they won eight baseball games, spread over a one-game play-in and two short series. And, to the delight of baseball fans, most of those games were tight and exciting. Isn't that enough?

Friday, October 17, 2014

The Royals: This Changes...Well, Nothing

With eight straight wins in the postseason--a new record--and their first trip to the World Series since 1985, the Kansas City Royals have become the "it" team of this October. But it's for more than their long playoff drought and low payroll ($89.3 million, 17th of 30 teams). It's because of the way they play, and how they've advanced in the playoffs. Pitted against two slugging teams, the Royals swept the Angels (top scoring team in the AL) in three and the Orioles (top home run team in the majors) in four. The Royals are throwbacks: Last in the majors in homers, last in strikeouts, last in walks. All the things that modern analysis embraces, the Royals reject, and are now four wins away from being world champions.

Here, look at this table:
Tm HR% SO% BB%
Kansas City 1.6% 16.3% 6.3%
League Average 2.3% 19.8% 7.7%
KC's Opponents 2.9% 19.6% 9.8%
Generated 10/17/2014.

The percentages are the percentage of plate appearances. So, for example, while the average AL team hit a homer in 2.3% of plate appearances, the Royals did in only 1.6%, while its playoff opponents--the Angels and the Orioles--averaged 2.9%. (I excluded the one-game wild card game against Oakland because the roster rules allow teams to pack extra hitters for the one game, making the conditions unlike other games). As you can see, the Royals eschewed the strategy of standing at the plate, taking pitches, and waiting for a ball to drive over the fence. And it's worked. Can the Royals' success be a blueprint for other teams in the future?

My answer is clearly no, for two reasons. First, let's not forget the regular season. The Royals won 89 games, playing in arguably the weakest division in baseball. They didn't win their division. As a wild card team, they finished only one game ahead of Oakland, two ahead of Seattle. A couple balls going bouncing differently than they did, and the Royals are traveling to Oakland for that wild card game, or maybe watching it on TV. And let's not forget the wild card game: If the A's don't blow a 7-3 lead with one out and one on in the eighth, or an 8-7 lead with one out and nobody on in the twelfth, and we're not even talking about the Royals.

Second reason: I lied. That table above? The third line of figures aren't combined totals for the Angels and Orioles. They're the numbers compiled by the Royals in their seven wins. Yes, for all the talk about small ball, the Royals have been hitting a bunch of homers, drawing a bunch of walks, and striking out in nearly one of five plate appearances--just like the other teams. It hasn't been a case of the Royals winning because they're playing a new brand of baseball; they're winning because they're playing like their opponents. It's not a case of showing us a new way; it's a case of imitation being the sincerest form of advancing in October. The Royals have won by slugging.

Three aspects of the Royals have endured, though. First, their bullpen, as I explained earlier, has been historically great, and has remained so in the playoffs. But a strong bullpen isn't a new idea. Second, their defense has been airtight, with multiple highlight-reel plays in every game, seemingly. But that's not unique to Kansas City. Every team aspires to that. The Royals' success is one of execution, not strategy. Third, the Royals steal a lot of bases. During the season, they led majors with 0.94 steals per game. No other team had more than 0.85. During the playoffs, the Royals have continued to steal a lot, albeit at a slightly slower (0.86 per game) pace. But those steals haven't been a major factor:

  • Divisional Series Game 1, tenth inning: Terrance Gore steals second, doesn't score.
  • Divisional Series Game 2, second inning: Alex Gordon steals second, doesn't score.
  • Divisional Series Game 2, ninth inning: Gore steals second, doesn't score.
  • Divisional Series Game 2, eleventh inning: Gordon steals second, later scores, due to a throwing error that allowed him to advance to third and score on an infield single.
  • Divisional Series Game 3, third inning: Billy Butler steals second, doesn't score.
  • Championship Series Game 1, seventh inning: Jarrod Dyson caught stealing.
  • Championship Series Game 2, fifth inning: Lorenzo Cain steals second, doesn't score.
  • Championship Series Game 2, seventh inning: Dyson caught stealing.
That's it. For all the noise about the Royals' stolen bases, they led to only one run, and that one was lucky (helped along by an error) after the Royals already had a two-run lead. It really hasn't been a factor.

So when you hear, as you undoubtedly will, that the Royals have won because of a brand of baseball featuring speed and defense with a de-emphasis on homers, walks, and strikeouts, you'll be listening to someone who hasn't been paying attention to the Royals this postseason. They've won because of defense and their bullpen, yes, but also because of an above-average rate of homers and walks, with little upside from stolen bases and little downside from a lot of strikeouts. That's the same brand of baseball as the teams they've swept.

Monday, October 13, 2014

Free Advice for Bruce Bochy

Right handed pitcher Hunter Strickland was drafted by the Red Sox in 2007, traded to Pittsburgh in 2009, and selected off waivers by the Giants in 2013. At 25, he made his major league debut in September, allowing five hits and no walks while striking out nine over seven innings spanning nine appearances for San Francisco. 

Here are his lines against right-handed and left-handed hitters so far in the postseason (four games pitched):

  • RIGHT-HANDED BATTERS: Eleven batters faced, one single, one walk, three strikeouts. That's a .100/.182/.100  slash line allowed, .282 OPS allowed, 2.31 fielding independent pitching (a proxy for ERA based on home runs, walks, strikeouts, and hit by pitches). 
  • LEFT-HANDED BATTERS: Seven batters faced, four home runs, two strikeouts. That's a .571/.571/2.286 slash line allowed, 2.857 OPS allowed, 23.70 fielding independent pitching.
It may be just me, but I think he's been more effective against right-handed hitters.

Sunday, October 12, 2014

On The Mound, Jake Peavy

Tonight the San Francisco Giants will seek to match the Kansas City Royals: Wild card team, winning on the road to take a 2-0 lead in their Championship Series. Here is a list of teams that have lost the first two home games of a seven-game Championship Series (it was a best-of-five series from 1969 to 1984) that have gone on to win:

That's right, none. If the Orioles come back against the Royals, they'll be the first team to have accomplished it. The Giants are trying to put the Cardinals into the same situation.

Pitching for the Giants will be Jake Peavy, whom they acquired from the Red Sox for two minor leaguers on July 26. His performance before and after the trade has been an interesting contrast:
BOS AL 1 9 4.72 20 20 124.0 131 67 65 20 46 1 100 1.427
SFG NL 6 4 2.17 12 12 78.2 65 24 19 3 17 1 58 1.042
Provided by Baseball-Reference.com: View Original Table
Generated 10/12/2014.

Adjusted for league and ballpark, his ERA was 18% worse than the AL average with Boston, 61% better than the NL average with San Francisco.

Here's his breakdown against the Cardinals:

St. Louis Cardinals342.979957.25024199180611.179
Everyone Else1361073.543293262090.018518668242206294019661.187
St. Louis - Postseason0210.133313.125151535292.250
Generated 10/12/2014.

I imagine people are going to make a big deal about that postseason record against the Cardinals. I'd ignore it for three reasons. First, three games are not enough to draw a conclusion about anything. Second, his regular season record against the Cardinals--which is still not large enough to be meaningful--indicates no particular problem, and it was all compiled before he was traded from the National League to the American in 2009. At that point, the Cardinals had, if I counted correctly, only two batters who are still with the team today (catcher Yadier Molina and left fielder Matt Holliday), so the prior record's not really all that relevant. Third, two of those postseason games occurred in 2005 and 2006 when Peavy was a young pitcher with the Padres. He's not the same pitcher today at 33, when he ranked 51 of 88 ERA qualifiers with 7.0 strikeouts per nine innings pitched, that he was back then at 24 and 25, when he led the league with 9.6. 

He did look decent against Washington in the Divisional Series opener, giving up no runs in 5.2 innings, though he struck out only three and walked the same number, running his pitch count up to 104 and getting strikes on only 60% of his pitches (the league average was 64%). So I'd disregard his line from those three games against the Cardinals--two from long ago--but at the same time assume a busier evening for the Giants bullpen than last night, when Madison Bumgarner lasted until there were two outs in the eighth.

Friday, October 10, 2014

Time for Awards

Yeah, I know, they won't be announced until after the Series. But the voters have to submit their ballots before postseason play. More significantly, the deadline for Baseball Prospectus's Internet Voting Awards is next Friday, and I voted (my ballot is here), so here we go.

Annual disclaimer: Note that this blog is called On The Field of Play. Awards aren't on the field of play. So this is going to be short. I'm going to list who I think should get each award and who I think will receive it. Then let's move on, OK? It's the postseason.

Disclaimer: I don't feel a team's won-lost record is relevant. I don't think awards should be limited to players whose teams play in the postseason. I also don't buy that pitchers shouldn't be eligible for the MVP award. That's my opinion. You might not agree. Many voters don't. That's cool. Just want you to know where I'm coming from.

Who should get it: Last year, I agonized over this one. This year, it's one of three absurdly easy awards to call. The Angels' Mike Trout led the league in runs, RBI, and total bases; was third in slugging percentage, home runs, and OPS; and played a key defensive position (center field), and played it well.

Who will get it: The only question is whether it'll be unanimous.

Who should get it:
There are several players who, in my view, had good years. I'll list them by OPS:

Andrew McCutchen .952 PIT 548 89 172 38 6 25 83 84 8 115 9 18 3 .314 .410 .542
Giancarlo Stanton .950 MIA 539 89 155 31 1 37 105 94 24 170 16 13 1 .288 .395 .555
Anthony Rizzo .913 CHC 524 89 150 28 1 32 78 73 7 116 8 5 4 .286 .386 .527
Yasiel Puig .863 LAD 558 92 165 37 9 16 69 67 3 124 7 11 7 .296 .382 .480
Buster Posey .854 SFG 547 72 170 28 2 22 89 47 5 69 16 0 1 .311 .364 .490
Jayson Werth .849 WSN 534 85 156 37 1 16 82 83 3 113 9 9 1 .292 .394 .455
Jonathan Lucroy .837 MIL 585 73 176 53 2 13 69 66 3 71 13 4 4 .301 .373 .465
Carlos Gomez .833 MIL 574 95 163 34 4 23 73 47 0 141 10 34 12 .284 .356 .477
Anthony Rendon .824 WSN 613 111 176 39 6 21 83 58 2 104 11 17 3 .287 .351 .473
Provided by Baseball-Reference.com: View Play Index Tool Used
Generated 10/10/2014.

However, the question isn't which hitter had the best year; it's whether the award will go to a hitter or a pitcher. Dodgers starting pitcher Clayton Kershaw went 21-3, led the league in ERA for the fourth straight year--his 1.77 ERA was less than half the league average of 3.66--and he was first in WHIP (0.86), strikeouts/walks (7.7), and complete games (6) and third in strikeouts (first in strikeouts per nine innings) despite starting only 27 games due to an early-season stint on the disabled list. I voted for Pittsburgh's McCutchen (followed by Miami's Stanton) because of that DL stay; I weighted McCutchen's 648 plate appearances (tied for 16th) over Kershaw's 749 batters faced (37th). But I'd be fine with Kershaw winning...

Who will get it: ...which is a good thing, because he will.

AL Cy Young
Who should get it: I thought Cleveland's Corey Kluber and Seattle's Felix Hernandez stood out this year. Let's break them down:
Corey Kluber 34 34 3 18 9 .667 0 235.2 207 72 64 51 269 2.44 14 .233 .279 .345
Felix Hernandez 34 34 0 15 6 .714 0 236.0 170 68 56 46 248 2.14 16 .200 .243 .303
Provided by Baseball-Reference.com: View Play Index Tool Used
Generated 10/10/2014.

Kluber generated his record despite playing with a weak defense behind him. Hernandez pitches in a pitcher-friendly ballpark but got limited run support (3.5 runs per 27 outs when he was in the game, 12th lowest in the league). Both pitched for teams that made a run for, and fell short of, the post-season. I was all set to vote for Hernandez given that he played in the same division as two of the best teams in baseball (Angels and A's) while Kluber pitched in the weak American League Central, but Behind the Box Score calculated that Kluber's opponent-adjusted run prevention was superior. So I voted for him.

Who will get it: With no obvious choice among the three division winners, I suspect King Felix will win his second award of the decade.

NL Cy Young
Who should get it: Kershaw, obviously. Adam Wainwright of St. Louis and Johnny Cueto of Cincinnati had good years, but there is not a race.

Who will get it: As I said, there are three absurdly easy races to call this year. Trout for AL MVP is the first. This is the second.

AL Rookie of the Year
Who should get it: And this is the third. Chicago slugger Jose Abreu was fifth in batting average and on base percentage, first in slugging, third in homers, and second in OPS. Yankees starter Masahiro Tanaka was an early favorite but was sidelined from early July to late September with elbow woes, limiting him to 20 starts.

Who will get it: Abreu. It won't be close.

NL Rookie of the Year
Who should get it: The favorite much of the season was Cincinnati's Billy Hamilton, who had a surprisingly good glove in center field settled into the Reds' leadoff position, stealing 56 bases. I give him two major demerits, though. First, his .292 on base percentage really doesn't cut it for a leadoff hitter. The league average was .326, and the Reds were second to last in leadoff hitter on base percentage. That's why Hamilton, despite his blazing speed and starting 136 games in the leadoff position, was tied for only 34th in runs scored. Second, Hamilton somehow managed to get caught stealing on  29% of his steal attempts, worse than the league average of 28%. So my pick is Mets starter Jacob deGrom, who compiled a 9-6 record with a 2.69 ERA and more than a strikeout per inning after being called up in mid-May.

Who will get it: It'll be close between Hamilton and deGrom. I think the voters will probably be swayed by Hamilton's raw steals and strong defense without considering his poor on-base skills and pedestrian stolen base success percentage.

AL Manager of the Year
Who should get it: I voted for Baltimore's Buck Showalter, New York's Joe Girardi, and Cleveland's Terry Francona, in that order. Showalter guided a team picked by many to finish last in its division to the second-best record in the league. Girardi had to deal with injuries to his entire starting rotation, two free agent busts (Brian McCann, Carlos Beltran), and (ahem) the worst OPS in the league from the shortstop position, and still got 84 wins. Francona led a team expected to decline after a postseason appearance last year to 85 wins and postseason contention well into September. 

Who will get it: I imagine Showalter will walk away with it. Angels manager Mike Sciocia will probably get votes too.

NL Manager of the Year
Who should get it: My pick's Pittsburgh's Clint Hurdle who, like Francona, guided a team to a surprising playoff bid in 2013 but was expected to decline in 2014. Instead, the Pirates made it to October again (barely, getting bounced in the wild card game). I suspect San Francisco's Bruce Bochy will get support too, as will Washington's rookie skipper Matt Williams. Again, these votes take place before the postseason, so the voters didn't have the benefit of seeing Bochy kind of manage circles around Williams in their Divisional Series.

Who will get it: I imagine this will be horse race between Bochy, who was dealt a tough injury hand this year, and Williams. I hope for the sake of the Baseball Writers Association of America electorate that it's Bochy.