PECOTA Loves the Dodgers

Yesterday, Baseball Prospectus released their initial PECOTA projections for 2014. Nate Silver, noted political projectionist/analyst, introduced PECOTA in 2003. After Silver departed, Colin Wyers took over, and the Astros recently hired him. These projections are famous for being extremely detailed and are well-regarded around the baseball analysis community. The projections aren’t without their issues or inaccuracies, but the level of detail means that they’re fun to look at.

First, let’s talk about some projections for individual players. The PECOTA projections are behind a paywall, so I’m not going to show all of them. If you like good baseball writing, Baseball Prospectus is well worth your money. However, I will talk about a few of the interesting projections below.

—

Alex Guerrero:

PA	BA	OBP	SLG	HR	FRAA	WARP	WARP/600PA
512	.240	.298	.405	17	0	1.6	1.88

Despite PECOTA projecting Guerrero to be the worst starting position player on the Dodgers in 2014, it’s comforting to see another projection system say that he’ll have positive value. Before PECOTA was released, the only other projection that gave Guerrero much value over a replacement player was ZiPS. The high amount of power in this projection (17 HR) surprises me a bit. PECOTA also forecasts Guerrero’s defense to be precisely neutral.

Yasiel Puig:

PA	BA	OBP	SLG	HR	FRAA	WARP	WARP/600PA
595	.288	.352	.489	25	-6.7	3.3	3.36

PECOTA’s projection for Puig is a bit conservative. For comparison, Puig was worth 4.2 WARP last season. PECOTA projects Puig’s walk rate to be 7.6% (a bit less than the 8.3% he managed last season), but the biggest difference between last year’s line and the forecast is caused by a drop in BABIP from .383 to .336, which seems pretty reasonable. Despite the drop in offensive production, PECOTA still projects Puig to be the Dodgers’ best hitter by True Average, Baseball Prospectus’ measurement of total hitting value.

Matt Kemp:

PA	BA	OBP	SLG	HR	FRAA	WARP	WARP/600PA
491	.274	.343	.486	22	-6.5	2.9	3.54

While PECOTA doesn’t see a return to MVP form for Kemp, it does predict a nice bounce-back from last season. He has the lowest forecast playing time among the four outfielders, which is fair given his recent injury history. His WARP/600PA is the highest among all Dodgers position players, which is nice to see.

Clayton Kershaw:

IP	ERA	K/9	BB/9	HR/9	WARP	WARP/180IP
207.7	2.42	9.1	2.4	0.65	5.2	4.51

Interestingly, PECOTA projects Kershaw to be the second-best starting pitcher in baseball next year by WARP. Justin Verlander has him beat out by a hair for best overall pitching value, but only because he’s projected to throw 11 more innings. When ranking starters by WARP/180IP, Kershaw is the best in baseball, as you’d expect. 5.2 WARP is identical to what he put up last season. Baseball Prospectus takes peripherals more heavily into account than some other value ranking systems, so the increase in ERA doesn’t hurt Kershaw’s projected WARP very much.

Dan Haren:

IP	ERA	K/9	BB/9	HR/9	WARP	WARP/180IP
156.0	3.47	7.7	1.5	1.2	1.6	1.85

PECOTA projects Haren to have the third-highest ERA reduction in 2014 (behind Ian Kennedy and CC Sabathia). Despite this, PECOTA doesn’t see much of an improvement in his overall value. Part of that is the reduced innings count (which doesn’t match his previous seasons), but it also doesn’t forecast much of a rebound in his peripherals. Even if this is all Haren can contribute, he’ll still be worth his salary.

Kenley Jansen:

IP	ERA	K/9	BB/9	HR/9	WARP	WARP/50IP
64.0	1.77	13.4	3.1	0.56	2.0	1.56

Kenley Jansen is really really good. That shouldn’t shock anybody at this point. The fun part about this projection is that it ranks him as the 45th best pitcher (by value) in all of baseball. He’s the second-highest reliever on this list (Craig Kimbrel ranks 31st). He’s only two slots behind Hyun-Jin Ryu by value.

Chris Withrow:

IP	ERA	K/9	BB/9	HR/9	WARP	WARP/50IP
45.7	4.45	8.2	4.4	0.98	-0.3	-0.3

I included this projection because it’s one I disagree with the most. PECOTA is really negative on Withrow. It’s forecasting a big increase in BABIP (and decrease in strand rate, most likely), but I think it’s much too aggressive in regressing his walk rate and strikeout rate. My guess is that it is still looking at his multiple years stuck in AA. This projection doesn’t make me think less of Withrow, but it is interesting to see.

—

Nothing from these projections looks particularly noteworthy or unusual. The projections look a bit conservative, as baseball projections tend to be. Yet, the 2014 package made news because it really likes the Dodgers. PECOTA is projecting the Dodgers to win 98 games.

The PECOTA projection for the Dodgers is nine games higher than the second-highest teams (the Red Sox and Rays are both projected to win 89). The Dodgers are projected to win the division by 11 games, by far the highest margin in baseball. PECOTA projects the Dodgers to score 732 runs, second in the national league behind the park-aided Rockies. It forecasts the Dodgers to give up 579 runs, 44 runs less than any other team in baseball.

Projection systems are usually pretty conservative with win-loss totals for individual teams. For example, the maximum projected win total in the recently released Clay Davenport projections is 91 (that system projects the Dodgers to win the west with 88 games). This isn’t because a projection system doesn’t think any team won’t win more than 91, it’s because the system can’t project a specific team to win more than 91 games in more than half of its simulation runs. Deviation over or under the projected total is an expected part of these projections. A 98 win projection is crazy.

The reason why these projections rank the Dodgers so high despite the conservative individual projections above is that no player is projected to have a significantly “down” season. The Dodgers don’t have a breakout position player performance (Adrian Gonzalez has the highest forecast WARP among position players, 3.6, which is only 26th in baseball), but they are projected to be ridiculously deep. Their worst regular position player is forecast to produce 1.6 WARP, and their worst position is forecast to produce 1.9 WARP. Only four other teams have all positions over 1.5 WARP. Even the current staff of utility players that we’re hoping the Dodgers replace have good projections.

The Dodgers’ competition also runs up the win total a bit. The Giants are projected to be pretty good (87-75), but the rest of the division is forecast to be .500 or below. Combine these factors with the great rotation, and it results in a gaudy win total.

If some of this sounds familiar, it’s because PECOTA really liked the Dodgers before the 2013 season as well. Before last season, PECOTA projected the Dodgers to win 93 games. The Dodgers nearly matched that (92 wins), but they took a much different route than what was projected before the season. To me, this win total seems way too high for a projection, and it’s unlikely that nobody will have a down year, as forecast. PECOTA projected the Yankees to have the second-highest win total last season, and we saw how that turned out. There are plenty of ways for projections to go wrong. But, in the end, being projected to have the highest win total in baseball is better than being projected to not have the best win total in baseball.

—

This post uses the following statistics:

• WARP – Wins Above Replacement Player. Baseball Prospectus’ method to measure player value. Explanation here.
• WARP/600PA – WARP, neutralized to 600 plate appearances in order to get a rate of offensive value rather than an overall accumulation.
• WARP/180IP, WARP/50IP – WARP, neutralized to a particular number of innings pitched in order to get a rate of pitching value rather than an overall accumulation. 180IP is used for starters, 50IP for relievers.