Monday, October 29, 2007

Defensive Projections

Here they are for 2008. I projected every outfielder with two ratings, one as a right/left fielder and one as a center fielder. Infielders get a projection for every position they actually played.

Projections are based on 4 years of data, age adjusted, and regressed to the mean. I've changed a few things about regression to the mean - outfielders are regressed to a mean that is a function of their speed score. A thanks to Mitchel Lichtman for opening my eyes to something that should have been obvious through this article.

For infielders, all first and second basemen are regressed to a mean of zero. Shortstops and third basemen are regressed to a lower mean, with the regression target rising depending on playing time. Once a player has the equivalent of a full season of playing time at a position, the mean he is regressed to rises to zero. A shortstop who played only 5 innings or so would have a projection around -7.

The reason I do this is this: Let's say Miguel Cabrera, a bad defender at third, was forced to play a few innings of shortstop and missed a play or two. His projection, had I done nothing, would be about -1. Orlando Cabrera's projection is -1, and it should be obvious that they are not equals. If you can play every day at short and be only -1 runs compared to the best fielders around, you are pretty valuable. So I came up with this trick to force a below average rating to people who really don't belong at a position.

For outfielders, both ratings are determined by outfield games played at all three positions. For infielders, ratings at each position are calculated independently - Maicer Izturis's shortstop rating is based only on his time there and has nothing to do with what he does at third base. Infield ratings are just based on a player's 4 year weighted record, playing time, and age.

The Ratings can be found here.

Wednesday, October 24, 2007

Best infield defensive seasons, 1957-1969

Last Spring I used Retrosheet data to create a play-by-play defensive system, Totalzone. I used the years 2003 to 2006, and while I don't have detailed hit location for those years, I was able to come pretty close to the ratings of a more advanced system like UZR, especially in the infield. For the 2003-2006 the batted ball type is reasonably complete, and it tells you who fielded every out or hit. Its enough to make a pretty good, though certainly not perfect system.

For the older seasons its a lot tougher. Batted ball type is generally available only for outs, and a lot of the time the position that fielded hits is incomplete. I was able to find some work arounds for these, which I may go into later. What I did was:

1. Look at how many plays an infielder fielded a ground ball and turned it into outs. By using retrosheet I can get this number exactly, as opposed to estimating based on assists and putouts. This is especially true for 1st basemen.

2. Look at how many errors and infield hits are allowed at each position.

3. Estimate how many hits went to the OF for each spot. I apply the team groundball ratio (which is based on outs, usually missing for hits) and a position percentage based on the batter's handedness. For example, third basemen make only 7% of infield plays against lefties, but 34% against righties.

4. Calculate a zone percentage on plays/opportunities, compare to league average, and convert to runs. I looked at separate averages for each year and for right and lefthanded batters.

The results passed the smell test. Here are the best and worst seasons, by position, from 1957 to 1969:

Don Mincher, 1966 +16
Ernie Banks, 1964 +15
Vic Power, 1959 +14

Dick Stuart, 1964 -17
Nate Colbert, 1969 -15
Jim Gentile, 1964 -15

All of this was before I was born. I don't know if Mincher showing up at the top speaks good or ill of my system. I know Power was considered a great gloveman, and Banks moved from shortstop due to injuries, but he must have been much more athletic than a typical 1B. Dr. Strangeglove shows up last, and I don't think anyone will find fault with that.

Bobby Knoop, 1964 +25
Nellie Fox, 1959 +24
Bill Mazeroski, 1958 +20

Nothing wrong with this list. Nice to see one of the early Angel stars at the top of the list, and Fox and Maz are in the Hall of Fame mostly due to their glovework. Maz also gets the #4 season.

Glenn Beckert, 1966 -18
Tony Taylor, 1964 -16
Felix Milan, 1969 -16

I really don't know much about any of these guys.

Mark Belanger, 1968 +29
Ernie Banks, 1959 +28
Belanger, 1969 +25

Belanger was the Adam Everett of his day, though he was even worse as a hitter. Ernie Banks was a truly great player back then, he also hit 45 homers and 143 RBI that year. Luis Aparicio didn't make the top single season list, but was above average 12 of the 13 seasons, with a high of +14, and about 100 runs above average over the period.

Freddie Patek, 1969 -35
Dick Howser, 1961 -31
Roberto Pena, 1968 -30

I'll have to see what Patek did when I get to the 70's. He was +5 in 1968, his first year.


Brooks Robinson, 1967 +46
Brooks, 1969 +40
Brooks, 1968 +38

Over the 13 years, Brooks was +273. There were only 3 seasons in the infield over +30, and Brooks had them all. These stats back up everything you've ever heard about the defense of Brooks Robinson. Thanks to him and Belanger mostly, the Orioles of this period had DER's around .740 - about .710-.720 was average during this period.

Dick Allen, 1967 -24
Richie Hebner, 1969 -22
Doug Rader, 1969 -22

Allen was 10 runs or worse every year from 1964-1969. I don't know enough about the others to say this list is good or not.

Tuesday, October 16, 2007

Top Free Agent Starters

Its a good thing Bill Stoneman left a strong starting staff to our new GM, Tony Reagins. According to my CHONE projections, which aren't quite done yet but preceeding nicely, there really isn't anyone on this list that should start ahead of any of our 5 guys. Yes, that includes Ervin Santana. In his favor, he has shown some durability and has one of the best arms of any starter on the planet. I have no doubt that were he a free agent now, he'd get paid on potential alone, and command about what Gil Meche got last year. Even in a terrible year he struck out more than twice as many as he walked. If only he can avoid the long ball...

Now, for those unfortunate teams that need pitching, here's what the market will offer, starting from #10:

10. Rodrigo Lopez

Not that good, but kept his ERA under 5 in Colorado, before that he was durable and threw strikes in Baltimore. Sure, a lot of those strikes got hit, but its a weak market.

9. Brett Tomko

He throws strikes. He's able to provide slightly above replacement level pitching as a starter or from the bullpen.

8. Jeff Weaver

Ugh. Weaver the elder can pitch effectively sometimes, like down the stretch and into the playoffs for the 2006 Cardinals. He's also pitches a lot of batting practice. He got 8 million last year after a bad season, so somebody will pay him after this most recent bad season. Expect a bit of a paycut though, like the two guys right behind him, he'll get something in the 6 million range.

7. Kyle Lohse

Last year he might have been the most average pitcher in the National League. If you can do that and pitch 190 innings, you are going to get paid. CHONE doesn't like the quality of his 2006 season, and values him around 7 million, but he'll probably get closer to Jeff Suppan money, 40 million over 4 years.

6. Freddy Garcia

This is one of Bill Stoneman's best non-moves. Remember the stupid rumors of Santana + Figgins for Garcia and Joe Crede? Good thing we passed on that one. Garcia is obviously a health risk as he lost most of 2007 to injury, but was a fine pitcher before that. If your medical reports on him are promising, he'd be a nice guy to take a risk on, maybe in the 7 million range + option + incentives.

5. Bartolo Colon

Pretty much the same story as Garcia, except he's been mostly hurt for two years. I wish Bartolo the best as he attempts to restart his career in another uniform, and I'll always remember his shutout of the A's in his final start in 2004. When healthy last year he was still throwing upper 90's heat with good control, I hope he can stay healthy for more than 2-3 starts at a time.

4. Paul Byrd

Another former Angel. Byrd throws strikes, gives you a decent number of league average innings. He's 37 now, and will be somebody's #4 starter for about 17 million over 2 years.

3. Carlos Silva

Throws strikes, gets grounders, stays healthy. Its enough for a big payday.

2. Greg Maddux

He pitched in the most pitcher friendly place in baseball last year. He'll be 42, but he's a decent bet to keep throwing 200 innings of league average ball until father time finally catches him.

1. Curt Schilling

He's 41, and doesn't have nearly the stuff he once did, but any pitcher who can strike out 4 for every walk still has value, and his stuff is still not bad. He and Maddux are probably both worth 11-12 million, and his age is not a bad thing in that by signing him you should be able to get a 1 year deal, and not risk the chance of an Adam Eaton type albatross.

Others to consider: Jason Jennings, Kenny Rogers, Tom Glavine,

Thursday, October 04, 2007

So how did the pitcher projections do?

I compared my pitcher projections to those of 4 other projection systems - Zips, Pecota, Marcel, and Hardball Times. The creaters of those systems are, respectively, Dan Szymborski of Baseball think factory, Nate Silver of Baseball Prospectus, Marcel the monkey (Tangotiger asks for the monkey to get the credit), and David Gassko and Chris Constancio of THT.

To compare I took a pitchers actual ERA, and compared it to the projected ERA. Sounds simple but there are a few things to take care of first:

1. Take pitchers who threw at least 50 innings
2. Adjust for league average. CHONE projected ERAs were lower across the board than the others. I figured the projected league ERA based on the innings pitched of the pitchers in my sample, and adjusted the projections accordingly. I multiplied CHONE projections by 1.008, PECOTA by .907, all the others somewhere in between.
3. There's a little over 300 pitchers in my sample, but not every system projected every pitcher. The ones not projected were mostly minor leaguers, or coming from Japan. What I did was if I had less than 3 projections for a pitcher, do not use him. If one or two systems do not project a pitcher but the others do, the pitcher gets a 6.00 ERA projection. Of the 5 systems, ZIPs projected the most pitchers and takes the fewest 6's.

Now, the evaluation criteria. I did two- the first is correlation, the second is average run error. Just take the absolute value of the difference between projected and actual ERA, and figure how many runs you are off. That way I adjust for playing time, it hurts more to be off by 0.50 in ERA for a fulltime starter than for a 50 inning reliever.

And the results:

Chone .405
Zips .370
Pecota .366
THT .340
Marcel .339

Not bad for my Chone. Now for the more advanced test, avg run error:

Chone 9.0
Pecota 9.2
THT 9.3
Marcel 9.4
Zips 9.5

Now what did my system do better than the others? For all I know it just got luckier. There isn't a huge deal of difference in the results. Last year Chone did not do so well with pitchers, and for all I know it was just unlucky, but it might be the use of batted ball data. From a pitcher's mix of grounders, pops, line drives, and flies, I attempt to get better estimates of a pitcher's batting average per ball in play and his homerun rate.

Wednesday, October 03, 2007

I'm going to Disneyland

The Angels need to regroups after being shut down by Josh Beckett, but my fantasy team in the Halosphere league is busting out the champagne right now.

Six years ago I named my cat, Brian Downing Kaat, after my favorite Angel ballplayer of all time. This year I named my fantasy team after him. Brian Downing's cats finished second in the regular season to the Bonn Bon-Bons, a team run by Bjoern from Germany, an Angel fan who makes me look like I live on Katella Avenue by comparison.

After securing a first round bye, the Cats clawed their way to a 7-6 win over the Reverend, who finished in third place. Then, in the finals they faced the Bon-Bons in a 2 week head to head playoff. It was close most of the way, tied through the first week, but the Cats pulled ahead late for a 9-5 victory. Both opponents had strong seasons and were worthy adversaries.

Following a strong draft, the Cats did not make a single trade all year, but solidified the roster with pickups of Brian Roberts, Ryan Braun, Joe Saunders, and Lance Berkman as the season went on.

Monday, October 01, 2007

Minor League Translations

2007 MLE's

For players at full season levels. Don't take these too seriously. I have Ryan Braun at a .603 slugging percentage in a little over 100 at bats. Who's going to believe that?