Thursday, February 20, 2014

Attempting to project Alex Guerrero's statisitics for his rookie season

It is notoriously difficult to translate performance in the Cuban league to the major leagues. The competition level is similar to A-ball in the minors. There's only about a dozen people who have any kind of record in both places, so that's not much to go on.

Nevertheless, some have made now their best attempt to guess at what our own Alex Guerrero might offer in 2014.

In the field, people expect him to be maybe average at second baseman once he gets used to it. On the bases, he might be about average as well, or a little below. He was signed for his bat though, so let's look at that.

His last few years in Cuba he was hitting something like .310/.400/.600, with a 14 percent walk- and strikeout rate. That sounds good, but it might only be equivalent to .225/.285/.400 (according to the foremost Cuba stats guy, Clay Davenport)

Some have taken a stab at projecting his 2014 hitting line, and they mostly see him doing better than that in the batting average department.

BB% / K% / ISO
8 / 14 / .128
.250/.309 /.418
7 / 16 / .168
.215 /.277 /.389
8 / 28 / .174
.240 /.298 /.405
? / ??  / .165
My Own Crazy Ideas
8 / 18 / .165

I don't have the Pecota details, but to get to that line, it'd have to be around an 8 percent walk rate and 25-26 percent strikeout rate.

My crazy ideas are based roughly on the idea of treating Cuba stats like A-ball, but also based on translating traditional scouting reports into Major League performance. If a guy has a 45 hit tool and 55 power tool, that's pretty much the line I'm giving you since 50 is average by definition. And that's what they say he has.

These five projections are not saying terribly different things. He looks to be about an average hitter, and if he's capable of playing something close to a major-league average second base, that's really good. His hitting projection is not average for a second baseman, it's average for all hitters. That makes him a solid regular approximately worth his salary ($28 million / 4 years = $7M AAV). This winter, free agents expected to give about that level of performance (call it ~2 Wins Above Replacement) got right around that same dollar figure give or take. That's guys like David Murphy at $6M, James Loney at $7M, Chris Young at $7.25M and Marlon Byrd at $8M.

And critically, it's a comparable or higher level of performance than can be expected from Mark Ellis next year, so the Dodgers are likely to have improved their team overall and paid a fair price to do it.

Wednesday, February 19, 2014

Dodgers do well defending the corners, could use help up the middle

The standard theory of baseball defense says to be strong up the middle. If you have a good shortstop, second baseman, center fielder and maybe a catcher, you have good defense.

The corner positions don't matter as much. Certainly, the middle positions are harder to play and you need to be a better athlete to do well there. An average SS is a better defender than an average first baseman (or even a good 1B). Because fewer people can play an adequate major league SS, you're more willing to put up with a weak bat there. That's the essence of the rationale for the positional adjustment in the WAR framework.

First defense is rated relative to others at your position in terms of runs saved over/under average and then you get the positional adjustment thrown in -- more runs for middle defenders and runs taken away for corner defenders as you move down the spectrum C, SS, 2B, CF, 3B, RF, LF, 1B. It's about five runs per season for each notch down (although 2B/3B are about the same and LF/RF are about the same) -- so catchers get +12.5 and 1B get -12.5. Designated hitter would be another notch below 1B.

What about chances? First basemen get the most chances by far, if you count catching throws from other infielders. But that's such an easy job, the average 1B fielding percentage is around .994 (vs., say, 3B is at .959 -- seven times the error rate). If you try to ask about just batted balls, these also have varying grades of difficulty. Fly balls are easier than opposite field grounders, which are easier than pulled grounders, generally speaking.

On the individual side, the atomic measurement of all the modern fielding metrics is something like plays made - expected plays made, where expected plays made depends on how many balls were nominally the responsibility of a fielder and how fieldable each were (i.e. how likely is an average major league defender at that position to make that play). As a result, the runs saved above/below average defender at your position is a really nice idea for a metric. It is labor-intensive and somewhat subjective, but analysts are doing it fairly well nowadays. Of course Adrian Gonzalez is not a better defensive player than Matt Kemp, but he's better at 1B than Matt Kemp is at CF, and that's what's interesting in terms of the Dodgers defensive situation

The Dodgers have turned the standard theory inside out. They have above average defenders for their position at all four corners, and below average at all four middle spots.

In 2013, counting everyone who played at the position, the Dodgers were rated as follows (in units of runs saved per 150 games) in terms of each position relative to average at that position:

7 8 3 23 -20 -4 4 -6

That's +41 on the corners and -26 up the middle. This difference of 67 runs between the two groups -- the biggest in baseball. Next were the A's at 54 and the Yankees at 45. At the opposite end of the spectrum, the Nationals' middle defense was 46 runs better than their corner defense. The Cardinals, Brewers and Red Sox were also toward the bottom.

So, there are both winning and losing teams, big market and small market teams, at both the top and bottom of the list. How can we judge if this non-traditional strategy is effective? First off, we can just add up all the numbers. A run is a run. That's a total of 14 runs above average from all fielders. That's in the top third of baseball teams.

Going into 2014, the situation is about the same. Mark Ellis is gone, but hopefully replaced by other average defenders. Otherwise it's the same starters. I'll get into details about my projections later, but based on who I think is going to play how much where, I see the 2014 picture as substantially similar to 2013.

Certainly all four corner spots should be above-average, while all four middle spots have a pretty good chance of being below-average. Is this going to work for them? Well, team defense is perhaps even harder to measure than individual defense. The Dodgers' outfield situation makes a point about team defense -- it doesn't matter if Kemp is not a great center fielder because he's surrounded on both sides by guys with incredible range in Carl Crawford and Yasiel Puig. So, if they race over and catch a ball in his zone, it's still an out.

The simplest way to measure team defense is just how many balls in play became outs. That's called defensive efficiency. Various attempts have been made to spruce it up for double plays, park effects, etc. The Dodgers ranked about in the middle of the pack last year, maybe a little above the median. A lot depends on Alex Guerrero and Kemp. With Guerrero, we'll just have to see he has to offer this spring -- not much to go on yet. Kemp was looking less and less like a viable center fielder before his injuries. Given his serious ankle injury and hamstring issues, a reasonable person might expect him to be terrible out there this year (although, I do think we're going to see him hit like we're used to).

We can hope for the best, but the team's best lineup in the second half might be Kemp in a corner with Puig or Andre Ethier -- or even Joc Pederson -- in CF.

Photo credit: File photo

Friday, February 14, 2014

Looking at the Dodgers' numerous bullpen options for 2014

The Dodgers have a LOT of guys who are viable bullpen options, not even counting starting pitchers who are biding time. It's likely all these guys see action at some point during the year, but what should the pecking order should be?

I looked at the peripheral stats for all the guys and projected how I think they'll do this year. The rigorous approach led to some surprises and insights. Here's my list of candidates, and projections for these players, with Dodger Stadium as their home park, in 2014.

Kenley Jansen
Brian Wilson
Jose Dominguez
Paco Rodriguez (L)
Jamey Wright
Yimi Garcia
Javy Guerra
Chris Perez
J.P. Howell (L)
Scott Elbert (L)
Brandon League
Seth Rosin
Onelki Garcia (L)
Chris Withrow

Kenley Jansen, of course, is a beast, and I feel good about Brian Wilson ($10 million looks like an overpay in light of later "closer" signings this winter, but whatever). Rodriguez ought to bounce back from a poor close to the season. That's three good guys, which goes a long way to having a solid bullpen. Relievers are volatile, but you can be pretty sure about the A-listers here at least.

Chris Perez and Brandon League are what they are -- and it could go a lot worse than what you see here.

Some guys I thought were pretty good actually just got lucky last year, or temporarily elevated themselves (I'll believe it's permanent if they do it again), and I'm not sure I see them carrying their performance level through to 2014. I'm speaking specifically about J.P. Howell and Chris Withrow.

A lot of people are down on Withrow, and I couldn't see why. I remember him getting a lot of outs last year. He did better in the majors than he ever did in the minors in terms of strikeouts and walks. He also had a completely unsustainable .205 batting average on balls in play (BABIP). Howell also had a crazy-low BABIP (.241) and a HR rate 1/3 of his career average.

On the other hand, I think there's a few guys who I weren't thinking of that highly who could surprise this year.

Jose Dominguez , Yimi Garica and Javy Guerra look ready to make the jump to solid big-league contributors. Dominguez is going to get the strikeouts with that 100 MPH fastball, and I think he'll continue to limit the number of home runs he allows. His control isn't great, but it'll play. Yimi Garcia has a great slider that's striking out everyone. Guerra has been up and down, but if you normalize his BABIP & HR/FB, he actually looked good last year in both the majors and Triple-A. Jamey Wright is an ageless marvel. Seth Rosin, I don't see what the Dodgers see.

Rodriguez and Howell have the two lefty spots pretty much locked up. Scott Elbert Onelki Garcia are both recovering from surgery. Elbert is a mid-year return, while Garcia hopefully is ready close to opening day.

As far as minor-league starters spending a little time in the bullpen, I see Zach Lee (4.22 ERA) and Stephen Fife (4.28) as not that great this year. Chris Reed (4.69) and Matt Magill (4.96) just do not look good to me. Ross Stripling (3.56) is the one guy I think could contribute in that role. The numbers in parentheses are how I project them as major league starters, so perhaps they'd do a bit better in the bullpen (on average, bullpen ERAs are about 0.3 lower despite having lower quality pitchers than starting rotations). However, maybe they'd have a hard time adjusting (Fife's season was basically ruined by the disruption last year).

Paul Maholm (3.83), Josh Beckett (3.80) and Chad Billingsley (3.55) all might or might not be asked to get in the bullpen business too, with Dan Haren (3.62) as an extreme outside shot. Most likely, they'd be at least toward the middle to upper-end of the pack if they had to do that.

A lot will go into the decision besides performance -- who has what kind of contract situation in terms of minor league options, Rule 5 status, etc. There's also who the Dodgers might want to showcase for a trade, or how Beckett's recovery is going, and so on. But opening day is just a point in time. People will get injured or be ineffective and others will get their chances.

Photo credit: Dustin Nosler, Feelin' Kinda Blue