With last night’s win, the Royals’ total of GWWNHWITP (Games We Would Not Have Won In The Past) moves to three in the last four games, and four for the season as a whole (I would add the April 12th game against the Yankees, when the Royals started a three-run rally with two outs in the eighth.) I don’t know about the rest of you, but I’m having more and more trouble discounting the way the Royals are playing as simply a hot start, or simply a one-man show.
As Sam Mellinger points out, the Royals are only 10-11 when Zack Greinke doesn’t start – but they’re 16-6 when Sidney Ponson doesn’t start. The best way to look at is this: when neither Greinke nor Ponson start, the Royals are 10-6. If the Royals can play over .500 when their
Last night, the Royals should have taken the lead in the bottom of the seventh, when home plate umpire Jerry Crawford blew a call at the plate when A.J. Pierzynski used his Jedi mind trick on him. (Seriously, how is that a player universally considered one of the dirtiest players in the game so consistently gets umpires to make calls in his favor? If I were an umpire, I’d call every close play against Pierzynski on pure principle. Then again, if I were an umpire I’d probably sucker-punch Pierzynski just for the fun of it.) Living in
Steve Stone: “Pierzynski is in perfect position to make the catch, applies the perfect tag…” replay shows Maier’s leg gets in first “…and we caught a break.”
Hawk Harrelson: “I don’t know – Jerry Crawford is in perfect position to make that call.”
And that, my friends, is why Steve Stone is one of the best in the business. And why Hawk Harrelson is not.
- The Royals drew 11 more walks in the game (granted, two intentionally) and now rank 5th in the league in that category. Last night they scored the tying run in the sixth on a bases-loaded walk, three nights after they scored the go-ahead run in the 11th on a bases-loaded walk.
The Royals have three bases-loaded walks this season. They had five all of last year.
- Coco Crisp walked four more times. I’m running out of ways to describe how surreal this is to watch a leadoff hitter who spits on pitches two inches out of the strike zone. Crisp is just the fourth Royal this decade to walk four times in a game. (Alex Gordon set the team record with five last year – two intentionally – on August 30th. Like Crisp, he didn’t score in the game either.) Crisp also quietly tied a team record earlier this year by walking at least twice in four straight games, a feat last done by Jeff King in 1997.
He’s never done anything like this before, but at this point I’d be shocked if he returns to his previous career walk rates. Maybe there’s something in the water this year; one of the two players in the league with more walks than Crisp is Marcos Scutaro, who has 26 of them. Like Crisp, Scutaro has never drawn even 60 walks in a season before.
- You know whose plate discipline is even more startling than Crisp’s? Jose Guillen. Guillen has drawn eight walks in just 14 games. Guillen’s career high in walks is 41; he had 23 all of last season. He has a .414 OBP, people. How’s that portrait coming along?
- I’ll save more Greinke talk for another column - although Craig Brown as a nice column here about how Greinke is maintaining his stuff deep into games - but I did want to address the hysteria over Greinke’s ERA+, which is currently at 1170.
ERA+ is a wonderful stat, in that it adjusts ERA for the context of era and league and ballpark so that you can directly compare the performance of two pitchers from any point in baseball history. But ERA+ has a flaw which only manifests itself at the margins. This is not a flaw that is unique to ERA+. ANY statistic that involves dividing one number by another – in other words, any rate statistic – has this same potential flaw.
The problem with rate statistics is this: as the denominator in the formula decreases, the statistic increases at an ever faster rate – and when that denominator approaches zero, the statistic increases so fast that it quickly becomes unreliable.
An example may make this more clear. Consider three different pitchers, all of whom have pitched 240 innings in a season:
Pitcher A averages 16 innings per home run.
Pitcher B averages 80 innings per home run.
Pitcher C averages 240 innings per home run.
Looked at this way, it appears that the difference between Pitcher C and Pitcher B is far greater than the difference between Pitcher B and Pitcher A. But look at it the other, more traditional way:
Pitcher A surrendered 15 homers in 240 innings.
Pitcher B surrendered 3 homers in 240 innings.
Pitcher C surrendered 1 homer in 240 innings.
The raw difference between Pitchers B and C is just two homers, whereas the difference between Pitchers A and B is 12 homers. But the first method gives a false reading of Pitcher C’s ability because the method breaks down as homers approach zero. If Pitcher C had not surrendered any homers, then – like Greinke’s ERA+ before he gave up a run – his “innings per home run” would have been infinite.
Like I said, you can have this problem with any rate stat – which is why we have things like innings pitched and at-bat limits, so that someone who goes 2-for-2 doesn’t win the batting title and someone with a scoreless inning doesn’t lead the league in ERA. The problem with ERA+ and the example I presented is that the denominator isn’t expressed in terms of opportunities, but in terms of outcomes. If we reverse the formula – if we expressed their home run rates in terms of homers per inning instead of innings per homer – then the numbers look like this:
Pitcher A: 0.063 HR/IP
Pitcher B: 0.013 HR/IP
Pitcher C: 0.004 HR/IP
Which looks much more reasonable, and in fact, this is the way it’s normally presented (albeit in homers per nine innings, for obvious reasons).
Any stat which involves dividing by the outcome is prone to breaking down as the outcome approaches zero. When you extract the park and league adjustments from ERA+, the formula is essentially (league ERA/pitcher ERA). In this case, the closer a pitcher’s ERA is to zero, the closer his ERA+ gets to infinity. We can solve the problem by reversing the formula, making Greinke's ERA+ roughly 8, but custom has already dictated that with ERA+, like with OPS+, the higher the better. In this case, though, Greinke’s number is so high that it doesn’t mean much other than he’s been really, really good.
Sorry for the math digression.
- Brayan Pena cleared waivers, and I think it’s time I give up and admit that I’m out of my depth when it comes to figuring out which players are going to be claimed and which ones aren’t. You’d think that out of 29 other teams, one of them would want a 27-year-old switch-hitting catcher with a lifetime .303 average in the minors (and who has hit over .300 in all four of his seasons in Triple-A) making the league minimum. You would be wrong, as I was. I’m happy to have him. If he does nothing else for the Royals this year, he’s already earned all the ink I’ve spent writing about him with his game-altering pinch double against the Yankees on April 12th. I suspect we’ll see him again in September, if not sooner.
- I have a confession to make. I do a mean Guy Fieri impression.
- Finally, I’ve been meaning to address Luke Hochevar’s contract situation for awhile, and with Sidney Ponson making (hopefully) his last start tonight, now’s as good a time as any.
Hochevar came into the season with 1 year, 17 days of service time. The rules state that 172 service days = 1 full service year, so in order to keep Hochevar from being a free agent until after the 2014 season, the Royals can not afford to give him more than 154 days of service time this year.
Complicating things is that there are, I believe, 182 days (exactly 26 weeks) from Opening Day to the last Sunday of the season. This is why a player can get a full year of service time even though he’s not called up until a week into the season (see, for instance, Kerry Wood’s rookie year.)
Anyway, if you just count from the last day of the season and work backwards, the tipping point for Hochevar was May 4th. If the Royals had called him up before May 4th, then he would have qualified for free agency a year early. (This became a source of concern in the Jazayerli household when Gil Meche’s back acted up, and it looked like the Royals might need Hochevar to make Meche's next start – on May 3rd.) In other words, the Royals have no reason to keep Hochevar on the farm any longer purely from a service time standpoint.
There is the additional concern of trying to keep him from qualifying as a “super-two” player, a player who’s eligible for arbitration before he has three full years of service time. The math here is a little tougher to figure out, because the line that separates super-two players from the other players with 2+ years of service time moves from year to year. Generally, the dividing line falls between 120 and 140 days of service time. That means the Royals would need to keep Hochevar down on the farm for somewhere in the range of 5-7 more weeks.
This would be dumb, not only because the Royals need him, but because the benefits of avoid the super-two pale to the benefits of keeping a player out of free agency. If the player’s any good, then the couple extra million he’ll make that first year of arbitration is nothing compared to getting to keep a quality player for another season. If he’s not any good, then that first year of arbitration isn’t going to cost you much anyway.
So root for a slugfest tonight – the Mariners are cooperating by sending Carlos Silva to the mound – and hope that the Royals pull out the game despite, not because of, Ponson. And maybe we’ll finally get around to the rotation we should have started the season with in the first place.