Saturday, January 20, 2007

PECOTA on the Yankees

Nate Silver has finally made the 2007 PECOTA cards available for Premium subscribers. Combined with his PECOTA Does Prospects series from last year, I decided to see how some of the Yankees' young players developed the past year.

We're going to be looking at three stats here. WARP stands for a player's projected wins above replacement over a five-year peak. Upside is a player's potential runs above average for the same five years. Combined is those two stats added together.

In short, WARP measures a reasonable expectation while Upside measures potential. According to Nate Silver, a Combined score of 300 or better equates to an "elite" prospect. Obviously, major leaguers should score much higher than prospects. Here, then, are all the Yankees covered in Silver's series last year who also have a 2007 projection.

Robinson Cano
2006: 19.0 WARP, 129.5 Upside, 319.6 Combined
2007: 31.2 WARP, 262.4 Upside, 574.4 Combined

PECOTA adores Cano. He has the tenth highest Upside in all of baseball (at least among position players). Sabermetricians supposedly only like players who draw beau coup walks. This is why that mentality is silly.

Melky Cabrera
2006: 12.5 WARP, 78.9 Upside, 203.4 Combined
2007: 23.9 WARP, 110.8 Upside, 349.8 Combined

This one is a no-brainer. Melky established himself as a legitimate major leaguer at the age of 21. That's reflected in the doubling of his WARP score. However, his potential now seems much better, too. Considering his comparables list includes Carlos Beltran, Roberto Alomar, Pete Rose, Tony Gwynn, Derek Jeter and Carl Yastrzemski, Yankee fans have every reason to be optimistic.

Philip Hughes
2006: 17.9 WARP, 131.3 Upside, 310.8 Combined
2007: 21.1 WARP, 197.2 Upside, 408.2 Combined

Last year, PECOTA saw Hughes as the #3 pitching prospect (barely behind Francisco Liriano and Yusmeiro Petit) with the highest Upside. From that already lofty position, Hughes managed to increase his Upside by 50%. He's third among pitching prospects again, but the two ahead of him are Daisuke Matsuzaka--not a prospect--and Tim Lincecum--who has all of 31.2 IP to his name. No one else is anywhere near Hughes.

Hurricane Tyler Clippard
2006: 8.8 WARP, 29.9 Upside, 118 Combined
2007: 9.9 WARP, 49.6 Upside, 148.6 Combined

Clippard made a successful transition to the high minors last year. PECOTA certainly acknowledges that with a significantly higher Upside. But it also obviously agrees with the scouting belief that Clippard is a good-but-not-great pitcher. His stats really aren't anything special.

Kevin Thompson
2006: 13.8 WARP, 57.7 Upside, 195.9 Combined
2007: 13.4 WARP, 30.9 Upside, 164.9 Combined

Thompson is now 27, so he's basically reached his peak already. That his Upside declined is really just a reflection of his age. He's still every bit as likely to be a quality major leaguer. In fact, PECOTA thinks he has nearly an 80% chance of being worthy of an everyday job this year, and a 33% shot at being a star. He deserves to be in the bigs.

Kevin Whelan
2006: 7.3 WARP, 48.3 Upside, 121.7 Combined
2007: 4.4 WARP, 17.2 Upside, 61.2 Combined

23-year-olds who have never pitched above A-ball probably have some major deficiency holding them back. He simply isn't progressing the way kids his age should.

Eric Duncan
2006: 10.4 WARP, 72 Upside, 176 Combined
2007: 12.6 WARP, 33.4 Upside, 159.4 Combined

Is Duncan finished as a prospect? His WARP climbed merely because he reached Triple-A; he's more likely to get a cup of coffee at some point. But his Upside evaporated. It's unlikely Duncan will ever live up to his first round draft status.

Marcos Vechionacci
2006: 8.8 WARP, 13.0 Upside, 101.4 Combined
2007: 8.8 WARP, 2.6 Upside, 90.6 Combined

Vechionacci had a poor 2006, but he's still just 20-years-old. He didn't progress at all, but it's still too soon to write him off.

To recap, Cano, Melky and Hughes took huge steps forward; Clippard progressed nicely; Thompson is the same player but a year older; Whelan and Duncan look more and more like fringe players; and Vechionacci is nothing special.

More PECOTA stuff to follow, I'm sure.

Sunday, January 07, 2007

The Lefty Dilemma

With the signing of Eye Chart, the Yankees' regular lineup will probably look like this:

L - Damon
R - Jeter
L - Abreu
L- Giambi
R - Rodriguez
L - Matsui
L - Cano
S - Posada
L - Mientkiewicz

That's really the only way you can put those nine guys together without having three lefties in a row. The spacing is nice, actually: it follows a constant LLR pattern.

The downside is that there's no avoiding LOOGies. No matter what, teams will be able to use their crappy lefthanded middle relievers for at least two batters. Is it possible, however, that the Yankees' extreme lefthandedness gives them the advantage in the matchup game?

Many managers will be tempted to leave a LOOGY in to face four-of-five lefties. That means Jeter, Rodriguez and Posada will gain at bats against crappy lefties, often in crucial spots. If, on the other hand, managers switch to a righty, that means the next wave of lefthanded batters gain the platoon advantage back.

That the Yankees' lineup now features three matchup opportunites and no team has three quality lefty relievers (some lack even one) should guarantee the Yankees the platoon advantage more often than not.

Friday, January 05, 2007

Mientkiewicz or: Why Steven Goldman is Wrong

The Yankees have signed Doud Mientkiewicz to a 1-year, $1.5 million deal. I like it. Steven Goldman doesn't:

There is very, very little reason for a contending team, or any other, to be interested in Mientkiewicz as a starting first baseman. Or a platoon first baseman. He might make a decent reserve. Mientkiewicz is entering his age-33 season, the point in a player's career when he is in danger of running rapidly downhill. The problem with Mientkiewicz is that he's already gone downhill. His bat is a known quantity: he has little power, some plate judgment, and no ability to hit for average.

Goldman is right when he says Mientkiewicz is getting older and has little power. So why do I like the signing?

Well, first, Goldman is wrong on his other two points. Mientkiewicz has more than "some" plate judgment. He has drawn 377 walks in his career, when an average player would've drawn 281. That's a walk rate 34% above average. It's slipped in recent years, but for 2004-2006, his walk rate is 21% above average.

Second, saying Mientkiewicz has "no ability to hit for average" is absurd. His career average is .270. Last year he hit .283. Not a world beater, but a far cry from "no ability."

So what can we expect from Eyechart going forward? ZiPS has his 2007 projection as .255/.342/.383. Marcel has him at .260/.336/.402. Let's average them out: .258/.339/.393.
Using a simple Runs Created formula, that works out to 77 runs per 650 plate appearances.

Now that's not very good. The average AL 1B hit .280/.352/.467. In the NL, it was .290/.372/.507. Let's say that the average 1B hits .285/.362/.487, or 102 runs/650 PA. Maybe that's unfair to Mientkiewicz, since the NL is the easier league and those stats should be discounted a bit more, but whatever.

Anyway, we can put those numbers on a scale of batting average with this formula: (1.8*OBP + SLG) * .2595. For Mientkiewicz, it's .260. For the average 1B, it's .295. You can look at those numbers the same way you look at batting average, so they're easy to understand.

Baseball Prospectus figures replacement level in a couple different ways. This one works on the batting average scale, this one on the runs scale. These would set the replacement level for first basemen at a .261 average and 75 runs. What was Mientkiewicz again? .260 and 77 runs. In other words, he's the definition of replacement.

So with the bat, Eyechart gives you nothing. What about defense? He's always enjoyed a fantastic defensive reputation. Checking the stats, there are three that I put any real stock into: UZR, The Fielding Bible and The Fans' Scouting Report.

Mitchel Lichtman hasn't made UZR fully public since 2003, as he was working for the St. Louis Cardinals. For 2000-2003, however, Mientkiewicz was fantastic: 17 runs above average, second best in the majors to Todd Helton.

For 2006, the Dial ratings had him as the second best in the AL at +12 runs per 150 games. I don't put much stock in Dial's numbers, but I mention it only because in his midseason update, Lichtman chimed in to note that Mientkiewicz rated a +12 in UZR. So UZR still loves him.

The Fans' Scouting Report is exactly that: fans rate players in various skills based purely on their own observations. Tango Tiger compiles these scouting reports as a sort of "wisdom of the crowd" judgment. In 2003, the fans tabbed Mientkiewicz as the top first baseman in the majors. Tango worked out his ratings to be equal to +23 runs.

For 2004, fans rated him a 73 (out of 100). No idea how that turns into runs, but it's a great number for a first baseman. In 2005, he was a 67. And 2006 a 69, fourth-best among first baseman. So from a scouting perspective, he's still got it.

Lastly, The Fielding Bible. I don't have it with me, but it loves him. He was rated the second-best first baseman (behind Teixeira). What's more, their stats can't take into account saving bad throws, perhaps a first baseman's most important job. Baseball Info Solutions was curious though, so they watched every play from 2005 and recorded every bad throw saved. The result: Mientkiewicz and Albert Pujols rated ridiculously ahead of everyone else.

Put it all together and it is quite reasonable to say Mientkiewicz saves at least ten runs over a full season with his glove. Considering his brilliance with snagging bad throws, quite likely more like fifteen.

What we have, then, is this: his bat is worth 0 runs and his glove is worth 10-15. That makes him worth 1 win, maybe 1.5. That's below average, but not terrible. By TangoTiger's salary research, Mientkiewicz is worth between $4 million and $6 million.

The X factor is his back. How does his defense hold up following the surgery? It's a fair question, and entirely possible one that will hurt the Yankees. But consider this: he'll make $1.5 million in 2007. Even assuming his true value is
on the low-end ($4 million) then the Yankees are paying ~38 cents on the dollar. That seems a more than reasonable gamble to take. And if a better option comes along later, it's an eminently movable contract.

There really is no downside here. Nice move, Cashman.