Saturday, March 16, 2013

Baseball players rarely get fired

Jack Welch, the famous CEO of General Electric from 1981 to 2001 was famous for his ten percent rule. He believed that every year, the bottom ten percent of performing employees should be fired. We adopted some of his philosophies at the software company I worked for from 1993 to 2008. I ran Customer Service and I kept a spreadsheet on everything related to performance such as call times, customer reviews, and a dozen of other skill sets. Every year, the bottom performers would be let go no matter how long they had been with us. It sounds brutal but those numbers were posted daily and everyone knew what was expected. Such a rule does not exist in baseball. Many of the bottom ten percent of players get to keep playing.

I did a search on both Baseball-reference.com and on Fangraphs.com to identify who the lowest valued players were in the past three seasons. I searched for those batters who had played at least 250 games over the past three seasons with a WAR of zero or less. In Baseball-reference.com, there were twenty-five such players. Of those twenty-five, only four did not have a job this spring.

I did the same thing and looked for pitchers with at least 210 innings pitched over the past three seasons. There were strangely twenty-five of those too. Of those twenty-five, only six did not have a job this spring, at least in a major league camp.

Why is the mediocrity built into the Major League system? Juan Rivera is in the Yankees camp this spring on a minor league deal. But you know he's going to head north with the team, right? He's on our list. Despite being on that list, he's made $19 million dollars over those three seasons. The Yankees also brought Ben Francisco into camp because their injuries keep mounting. He is on my list.

I have never understood how these players continue to get jobs year after year. One of these conversations always leads to the "good clubhouse guys." Craig Calcaterra of NBC's HardBallTalk recently had a post where he talked to Brandon McCarthy, one of the most stat-savvy players in baseball, and McCarthy reiterated the "good clubhouse guy" mythology. Calcaterra, to his credit, was not really buying it.

Each team has a large amount of players playing in the Triple-A and Double-A level. It would seem really difficult for me to believe that a good portion of these could be as incompetent as these bottom twenty-five players in baseball are. And they could be that incompetent for a lot less money. Is there more to this than meets the eye?

Would it be jaded for me to wonder if baseball would rather keep these awful players around instead of turning them over because that would mean more players qualifying for an MLB pension? On the face of it, those financial concerns wouldn't seem to cost MLB that much money. Can it really be more cost effective to play a lousy player a million dollars instead of a replacement player half that much to save money on pensions? I don't know the answer. But purging the sport of the worst ten percent of the players on a regular basis seems so logical that it has to be another logical reason why it doesn't happen.

The aforementioned Yankees have several outfielders in their minor league system that could very easily replicate what Juan Rivera and Ben Francisco bring to the table. So why don't they go that way? They won't. You just know they won't. You just know the Red Sox will go with Jonny Gomes and Mike Carp over Jackie Bradley, Jr.

For me, back in my software company days, a guy that made other employees feel good and smile a lot just did not make a difference if the customers suffered as a result. In the end, there were too many competitors which would be glad to take our customers if we did not take care of ours. It was an easy decision to make. It was tough. But it was necessary. But for some reason, baseball is different. Old catchers last forever despite production when minor league catchers could do just as well and for a lot less money.

Whether this baseball truism comes from something innocent like a manager's comfort level or something more sinister like the pension angle, the bottom line here is that baseball perpetuates a cycle where the bottom ten percent of performers continue to work in their profession. The government might be the only other employer that does that.

Thursday, March 14, 2013

Allen Craig and his projections

I have been writing a series of posts that seem to take issues with how projection systems predict certain players over the course of the 2013 season. It may seem like I have an issue with projection systems. The fact is that I do not. When you put all of their projections together, the margin for error is quite small. All of them pretty much work the same way. You plug in the numbers that have been compiled to date and run the algorithms a few thousand times and see what spits out. Of course, I have no idea what I'm talking about, but that is my weak and general understanding of them. But there are cases where the projections do not seem to make sense. Either they are too pessimistic or too optimistic. One of those for me is Allen Craig of the St. Louis Cardinals.

Let me cover my bases here a bit and admit that I am a bit obsessed with Allen Craig. And it is quite complicated an obsession. Part of it is a deep-rooted dislike for Albert Pujols and the deep-seated hope that Craig can replace Pujols' value to the Cardinals. I admit that none of this is particularly sane. But just because I write about baseball an insane amount of time does not mean that I am an impartial metronome of players. There will always be this Fan screaming for a place like the human emotion in Spock or something. These obsessions do not always work out well. I had obsessions with Reid Brignac and Micah Owings. Those didn't exactly pan out. Though, Owings did hit a grand slam today.

But I think this one with Craig has a chance. The guy now has 857 plate appearances under his belt and has a cool triple slash line of .300/.348/.515. He has compiled 5.6 fWAR in about a year and a half of playing and has a wOBA of .370 and a wRC+ of 135. That's pretty impressive. So I am a bit flummoxed that his projections predict his numbers will go down instead of up. How could that be?

In case you want to see what I am talking about, Fangraphs.com lists the projection systems on each player's page. I'd throw in the BaseballProspectus.com projection (PECOTA) too but I need to renew my subscription and haven't yet.

First, the projection systems have him only playing anywhere from 117 to 140 games. This apparently takes into account the injury Craig suffered before the season last year. But that was last year and he is playing first base, one of the safer places on the diamond. I think he'll play 155 games. But the pessimism goes on from there. What follows are Craig's 2012 numbers followed by the average projection number for that stat (add them all together and divide). Here we go:
  • Batting Average: .307 / .291
  • On-base Percentage: .354 / .348
  • Slugging Percentage:  .522 / .501
  • wOBA:  .374 / .364
  • ISO:  .215 / .210
  • fWAR: 3.1 / 2.8
Are you getting the picture here? And remember, those average projections are averages and there are some that are significantly lower. ZiPS is the most pessimistic. Three of the five projections predict his strikeout rate will rise despite the fact that his swing and miss rate has gone down three years running. Two of the five predict his homer total will be less and two of the five predict his doubles total will be less. Allen Craig has a career BABIP of .329. None of the projection systems think it will be that high in 2013. 

Okay, perhaps you are saying, "Okay, smartypants, what are your predictions for Allen Craig since YOU know so much?" Well, first off, I don't know diddly. But my prediction for Allen Craig (barring injury) would be: .315/.364/.540 with 30 homers, 28 doubles, a wOBA of .384 and an fWAR of 5.0. Those are as dumbed-down a set of predictions you are going to find anywhere. There are no spreadsheets or computer readouts. But those are my numbers. And just remember, I can beat the computer at Hearts 81% of the time.

Wednesday, March 13, 2013

Norichika Aoki - regression candidate?

The Milwaukee Brewers caught a nice surprise in 2012 when Norichika Aoki turned in a 2.9 fWAR season and ended up compiling 588 plate appearances in 151 games. There are a lot of nice elements in his numbers as we shall see. With only one year in the Major Leagues (in this country), can he repeat those numbers? Only ZiPS doesn't think so. All the other projection systems think he will improve further. Let's take a closer look and see which system is going to be correct..

First of all, ZiPS does not believe in Aoki's 2012 power numbers. For a fairly small guy, Aoki racked up 37 doubles, four triples and ten homers. His .433 slugging percentage and .144 ISO had to be a surprise. ZiPS isn't the only one who thinks his power numbers will slip. But ZiPS is the most aggressive in his projected fall. The only natural place to look is to see what Aoki did in Japan.

And the numbers in Japan show us that what he did in 2012 with power numbers for Milwaukee look sustainable. His number of doubles was right in line with his production in Japan and if anything, his ten homers were somewhat less than his normal Japan production. My conclusion then is that his ISO should remain in the same ballpark as what he did in 2012 with a possible chance of it being slightly higher.

But power is only a part of the OPS and wOBA categories. Aoki would need to maintain his batting average and on-base percentage. The one thing that helps Aoki is that he nearly always makes contact. He only swung and missed on 4.6% of the pitches he saw in 2012. That is a very low number. And his plate discipline seems stable as he swung at only 27% of pitches out of the strike zone. So he'll make contact and he should walk fifty or so times this season.

The type of contact he makes concerns me a bit. Aoki hits a lot of ground balls. In fact he hit exactly twice as many ground balls as he hit fly balls. Only fifteen players in the majors had a higher ground ball to fly ball ratio than Aoki did. And of those fifteen, none had a lower line drive percentage than Aoki. And yet, Aoki's batting average on ground balls (and BABIP, of course) was .290.

I'm not saying it is impossible to bat .290 on ground balls. Jeter and Suzuki have been making a living doing that for years. But it is not the norm. As I have stated here before, the league average for ground balls is .238. That is the norm. Jeter and Suzuki have beat the norm for years and years. Is Aoki one of those guys too? Time will tell.

I'm not quite sure what to make of Aoki in the outfield. The scouting reports are that Aoki has a weak arm. And yet, he had seven assists from right field last season. Right field seemed to be his best position in 2012 and the only one that garnered him positive numbers of the three outfield positions he played. But a weak armed right-fielder doesn't make sense unless he keeps throwing guys out, right?  He is best in a corner position and is really not a center fielder. But his batting game is more like a center fielder or a second baseman type. So he really doesn't fit.

Norichika Aoki will be a big part of whether the Brewers compete in 2013 or not. If he has the kind of year he did in 2012 with his 30 steals in 38 attempts and his ability to get on base at a .350 or higher clip, he will aid that team's cause greatly. But if he cannot sustain his BABIP on ground balls, then he could fall down to the .260 and .270 range in batting average and that will drag off most of what he brought to the table last season. After weighing it all together, my conclusion is that Aoki is a regression candidate and is 31 years old. We'll have to see how it plays out for the Brewers.

Monday, March 11, 2013

Adrian Gonzalez's optimistic projections

Adrian Gonzalez has changed as a hitter the last two seasons and that change has not been reflected in his projections from six different sites culled for this post. And unless he reverts back to where he was four years ago, his projections are kind of messed up. Some might say that his time in Boston might have screwed him up a bit and thrown him off his game. But I maintain that some of the signs were there in his last season in San Diego.

Let's start with his walk rate. From 2008 until 2011, 95 of his 2,783 plate appearances were intentional walks. That is a 3.5 percent walk rate all by itself. And partly because of the intentional walks and other walks that were not intentional but were just as intentional (if you know what I mean), his walk rate was inflated. If you look at his entire career, subtract his intentional walks from his walk total and he has walked 8.2% of the time.

And yet his walk rate projected by the projection systems for 2013 is anywhere from a low of 8.9% to a high of 10.6%. This would seem to indicate that the projections expect him to be intentionally walked at least in double digit numbers.

But will that be the case? He was only intentionally walked five times in all of 2012. And he was only walked once intentionally in his 157 plate appearances for the Dodgers. As a result, his walk rate fell to a career low, 6.1%.

And there is more to this equation. Starting in 2010 in his last year with the Padres, Adrian Gonzalez's plate discipline has gotten lost a bit. In 2009, Gonzalez swung at 23.1% of the pitches he saw out of the strike zone and his total swing rate was 44.7%. In 2010, those figures rose to 31.8% (O-swing) and 48.8%. In 2011, they went up again to 35.5% and 49.2%. The progression continued into 2012 when those numbers went up again to 37.3% and 51.4%.  In other words, his patience at the plate has deteriorated to the point where his O-swing rate was the highest of his career and his total swing rate was the highest its been since 2005.

Obviously, Gonzalez can reverse those numbers and go back to being patient at the plate again. But he is trending against that and a four year slide doesn't seem that easy to break.

The other thing that looks optimistic to me is his ISO projections. After five years in a row of having an ISO well over .200, it fell to .164 last season. All of the projections expect that number to bounce back. None of them expect it to bounce back to where it was, but the range is from a low of .189 to a high of .211.

There are two things to look at when looking at ISO. First is the fly ball percentage and the second is the homer to fly ball percentage. Gonzalez has averaged a fly ball percentage of 37.4% for his career. But has been lower than that for the last two seasons.

His homer per fly ball percentage peaked in 2008 and 2009 with both seasons over 20%. The last three seasons have been 16.4%, 16.4% and then 9.6% last season. So he is hitting less fly balls and less of them are going over the fence. Unless that trend stops and reverses itself, how can his ISO go up?

Obviously, I could be all wet here. Adrian Gonzalez could regain his patience at the plate and hit more fly balls and have more of those fly balls go over the fence. But those factors have been trending in the reverse direction since his last season in San Diego. I'm not saying that Adrian Gonzalez is not still a dangerous hitter. He is. But the trend is that he is not as dangerous as he once was and his projections are not yet reflecting the direction he has been going.