Evaluating the Tigers through a Sabermetric lens

Thursday, December 18, 2008

2009 CHONE Projections

Sean Smith, who blogs on baseball and the Angels, has released his 2009 CHONE projections for hitters. He's in the process of getting this all set up on a new hosting site as his old one just stopped working (why, I'm not sure and from what I know, neither is he). The CHONE Defensive projections are what I had been using in my previous attempts at valuing a player, and those are currently gone. However, he's assured everyone they will be back when he gets them on his site.

So, go peruse the hitters tabs for your favorite team. The R150 column is Runs Above Average per 150 games based on his own custom Linear Weights system (linear weights is the best method for offensive evaluation). They are also park-adjusted, league-adjusted and there is an "expanded" tab you can click. These are only up for a few (Angels) players. However, what they give is a percentile projection, projections through the 2014 season, and incorporates a speed number which he uses to evaluate baserunning. These are fantastic projections. To see an expanded tab, here's one for Brandon Wood and Vlad Guerrero.

When the defensive projections come back, those are what I would use for defensive prowess until at leas the mid-way point of the 2009 season. Fielding data takes roughly twice as long to stabilize in comparison to offense. So 1 year of fielding data is on par with roughly a half of season of offensive data.

These are, however, the best projections on the internet that I know of (yes, I think they're better than B-Pro's PECOTA projections). And they are 100% free. Please check them out. The Pitchers projections should be up in a few weeks.

Thursday, December 11, 2008

Edwin Jackson Deal; James Skelton

There have been many things written already on the Edwin Jackson-Matt Joyce swap that occurred last night. I am not a fan. Not in the least. He's similar to getting the remote control car that you've wanted for the last three Christmas' and then you finally get it. Open it up, play around with it and realize that the shiny outer-coat is just a facade and that the gift is a cheap knock-off of the name brand toy you really wanted.

In this case, the shiny outer-coat is Edwin Jackson's ERA. His command improved to a mediocre level from downright terrible, but his strikeout rate really dropped -- and essentially -- offset everything good he did in moderately improving his control.

The anguish over such a deal is less about the talents of Matt Joyce, rather on the apparent lack-of-talents of Edwin Jackson. Jackson is a trade-able commodity -- I'd pause and think about a deal involving Jeff Larish rather than Matt Joyce -- it's all just cost dependent. And I don't like the cost of this deal.

In other news, the neglect of adding James Skelton to the 40 man roster, and thusly leaving him unprotected for the Rule 5 draft, results in the Tigers losing Skelton to the Arizona Diamondbacks. Not a fan of this, either. Skelton's not a big-league ball player in 2009 based on his talents alone. However, he can project to being a competent catcher at some point and the Tigers needs at catchers span beyond the Major League Level. So long as Dane Sardinha has procured a back-up job, James Skelton should still be within the Tigers organization.

It feels like there have been a lot of swings and misses by this front office in the most recent years. Couple that with numerous questionable decisions from the medical staff, head-scratching decisions from the manager on both in-game decisions and player personnel and I'm not sure I'm liking the way the team is currently heading. 2006 feels like decades ago already.

Monday, December 8, 2008

2009 Marcels Projections

Tom Tango released his 2009 Marcels Projections a few weeks ago. I just got around to downloading them. The projections (along with projections from Bill James) can be found on each players page on Fangraphs, the best baseball site online in my estimation.

I have quickly turned the 2009 projections in Wins Above Replacement (WAR) for both hitters and pitchers, and then turned that into a runs value for next season. The Hitters spreadsheet can be found here and the pitchers sheet is here. I used the wOBA turned into Runs Above Average method to get to my WAR numbers, which I should disclose, isn't 100% ideal; wOBA is not park-adjusted. So, Matt Holliday's wOBA calculations aren't taking into account that he's playing 81 times in Coors Field, while Brian Giles is playing half of his games in Petco.

Some caveats: There are no defensive numbers in here. This is straight-up offense. To stick with the 2009 projection theme, I would suggest using the 2009 CHONE Projections.

Also, there are no positional adjustments. This is due to my laziness, as I didn't want to add the position for each of the hitters in the spreadsheets. However, this is very, very easy once you know a person's position. The positional adjustment spectrum is just this:

CA: +12.5 runs

SS: +7.5 runs

2B/3B/CF: +2.5 runs

LF/RF: -7.5 runs

1B: -12.5 runs

DH: -17.5 runs

To add in the defense and the positional adjustment in, you simply add (or subtract) those to the number in the RAA column on my spreadsheet. You divide that new number by 10.5 (the relationship between runs and wins among players) and then add 2 (replacement level is 2 wins below average).


Lets run through an example. We'll use Curtis Granderson. His RAA in my table comes out at +15.7. His CHONE defensive projection has him at +13 runs. That makes him now at +28.7 runs above average. He gets another 2.5 run boost for being a Centerfielder, which puts his RAA at 31.2. Divide that by 10.5 and you get 2.97 Wins Above Average. Add in another 2, and he becomes 4.97 Wins Above Replacement in 2009. That equates out to being $24.85 million. His 2009 salary is currently just $3.5 million, a difference of $21.35 million -- quite the bargain.

More Shortstop Evaluations

With Fangraphs adding the very lauded UZR fielding metric invented by Mitchell Lichtman, I decided to now use that into my evaluations on Jack Wilson, Cesar Izturis, Adam Everett, and with talks in the comments section on Bill's site, I'll throw in Nick Punto, as well.

To keep this short, I'll just go with the nitty-gritty method.

Jack Wilson

Marcel's 2009 wOBA: .311
CHONE 2009 defensive projection: 7 runs.
UZR Average for the 06-08 seasons: 3.235
Probabilistic Model of Range Runs: N/A (hurt; did not meet the 1000 balls in play minimum used by David Pinto).
Justin Inaz's Defense Runs: +6.7

Average of defensive metrics: +5.645

WAR: 1.855
Market Value on 1 year deal: $9.275 million.

Cesar Izturis

Marcels wOBA projection: .290
CHONE: 7
UZR 06-08: 2.59
PMR: 11.69
Justin: 14.7

Average: 8.995

WAR: +1.139
Market Value: $5.696

Adam Everett

Marcels wOBA: .280
CHONE: 19
UZR 06-08: 18.02
PMR: 4.59
Justin: -0.5

Average: 10.28

WAR: 0.769
Market Value: $3.845

Nick Punto

Marcels wOBA: .299
CHONE: 3
UZR 06-08 (just his SS numbers used): 25.275
PMR: +5.74
Justin: +0.6

Average: 8.65

WAR: 1.549
Market Value: $7.749

Some notes:

-The UZR was a weighted average. I weighted 2008 at .45, 2007 at .35, and 06 at .20.

-The WAR is calculated by giving each player 595 PA's for 2009, which equates out to roughly 85% playing time -- or the amount that a normal starting fielder plays out of a 162 game season. This is a bit of an assumption for those on this list that have had health problems in the past season or two (Jack Wilson, Adam Everett). I also included an addition of 0.5 wins for a positional adjustment of these guys playing SS.

-Market Value is just WAR multiplied by 5, as that's the going rate for a 'win' above replacement level.

This method puts the players in this order: Jack Wilson, Nick Punto, Cesar Izturis, and Adam Everett.

Wilson costs a prospect and one that should be pretty good (likely one of the Tigers top 5 not named Rick Porcello), due to the Pirates apparently picking up a large chunk of the deal. No thanks.

Nick Punto will come cheap. He fared better in this method than I anticipated and has posted wOBA's from 2005-2008 of: .282, .324, .260, .324. He seems to have a drop off after getting closer to the league average each year. If this pattern holds true, he's likely to drop into the lower .290's -- that'd put him on par offensively with Cesar Izturis.

Izturis, also would come cheap monetarily. He's more glove than bat, but has been steady with the wOBA's since 2005: .274, .258, .273, .292. He's projected for .290 but the transfer to the AL from the NL might be enough to keep him below that projection. Also working in his favor is that he's 28 and could be on another year of an upswing before decline sets in.

Everett's by far the worst bat of the four potential options. His defense has always overcome such shortcomings at the dish, but he's coming off a couple of injury-plagued seasons in which his elite defense is being questioned as a thing of the past.

These options are brutal. There's no other way to put it. I think that I like Cesar Izturis on the 1-year deal, but I say that with immense trepidation. I think I can expunge Jack Wilson from this discussion simply because he costs players rather than money. With that said, the other three aren't exactly names you expect on World Series contenders. I can get behind picking up any of the remaining three, as it's just a matter of who's the cheapest.

Thursday, December 4, 2008

More SS Solutions; Everett and Izturis

Well, Billfer mentioned Cesar Izturis and Adam Everett as other options that have been linked to the Tigers. Well, lets look at how they stack up, in dollars, with Wilson and each other.

Adam Everett

The hands-down best defensive shortstop in baseball just a couple seasons ago, a couple injury-plagued seasons have left his once +40-something plays above-average defense down to just +18 plays in 2007 (he wasn't on the Fielding Bible leaderboards in 2008). As the very definition of an all-glove, no-bat Shortstop, this is not good news. Marcels has his wOBA projection at .280 in 299 PA's. On the very surface, that comes out to, using the method in my previous post, -14.3 runs above average (RAA). That then converts over to being 0.638 Wins Above Average (WAA) offensively. Again, that's just 299 PA's. If we give him just 75% of the playing time (full time starters are normally around the 85% range) because of his injury troubles, that puts him at -25 runs above average offensively, which is -0.391 WAA.

But, we know that his bread is buttered by his glove. The CHONE defensive projections have him at +19 runs defensively. We add those to his RAA over the course of 75% playing time (525 PA's; -25 RAA), and you get -6 RAA. Convert that WAR, puts him at 1.43 WAR. Add on 0.5 wins for being a SS, and his overall WAR becomes 1.93 -- close to Jack Wilson's 1.98 WAR we discovered earlier. This WAR of 1.93 equates out to being worth $9.65 million on the open market for a 1 year deal. Coming off of an injury riddled last couple of years, his asking price will be drastically slashed -- maybe even into the $4.5 range, or roughly half of his value based on his Marcels Projections.

The pros: Can be had cheap. 2008 salary was just $2.8 million. Doesn't cost anything in terms of players. No long term commitment.

The cons: Health is an issue, and his seemingly declining defense is even more of an issue. Still an above-average defender, his bat comes close to completely negating his defensive worth.

Cesar Izturis

Izturis had 414 PA's with the St. Louis Cardinals in 2008 and posted a .292 wOBA -- his highest since the .313 wOBA he posted with the LA Dodgers in 2004. Marcels projects a .290 wOBA over the course of 421 PA's, which is roughly 60% of the playing time. We'll put him on equal footing to Everett above, and give him 525 PA's -- or just 75% of the playing time. His .290 wOBA would put him at -20.54 runs above average over the course of 525 PA's in 2009. His Marcels projection puts him at just +7 with the glove -- not the projection that Everett has. Izturis finished 2008 at +19 in the Fielding Bible. Those +19 plays equate out to 14 runs saved when converted using Chris Dials conversions. That total is pretty much corroborated by Beyond the Boxscore and

Evaluating Jack Wilson

There's been rumors that a deal for Jack Wilson was done between Detroit and Pittsburgh, and that the deal was never finalized and that they're still talking. Whether it's done, not done, or not even being talked about is only mildly important until we know the value of Jack Wilson.

So, let's evaluate him. His Marcels projection has him posting a .311 wOBA in 418 plate appearances in 2009.

First off, the 418 PA's seems rather low. That's because Marcels is built on a 3 year sample for players which are weighted at 5/4/3. He was hurt last year with a leg injury and, thusly, had just 330 PA's. From 2005-07 he averaged 589 PA's. Lets average that with his 2008 330 PA's and we get 460 PA's, which is roughly 66% playing time.

So, let's calculate his worth.

To turn wOBA into a runs number, you just have to do the following:

wOBA - league average wOBA / 1.15 * PA

Very simple. The League average wOBA will always be the league average OBP. 2008 in the AL, it was .335. For Wilson, it looks like this.

.311-.335/1.15*460=-9.6. That is 9.6 runs below average. His CHONE defensive projection for 2009 is 7 runs to the good defensively. That would make him -2.6 runs above average (RAA).

Now, 10.5 is the relationship between runs and wins, so -2.6/10.5=-0.248 Wins Above Average (WAA). Replacement level is defined at being 2 wins below average, so we add 2 to that and get his Wins Above Replacement (WAR) of +1.752.

Now, that +1.752 WAR is the not the same coming from a shortstop as it is from a 1st baseman. So, we must use a positional adjustment. Tom Tango (I'm leaning heavily on his methods for this evaluation), has positional adjustments laid out. For a SS, we ad 0.5 wins. Our total now comes to 2.252 WAR.

This offseason, the going rate for a 'win' is roughly $5 million. So, 2.252*5=$11.26 million.

However, we don't expect someone who has been so durable over his career to play just 66% of the time. Let's adjust that to 85% (the typical amount of a full time starting position player) and we get 595 PA's. His new offensive runs projection equals out to:

.311-.335/1.15*595=-12.42 RAA. Add in the +7 defensive projection, and you get -5.42 RAA. Convert to WAR and you get 1.48 WAR. Add in the positional adjustment and you get 1.98 WAR. The going rate for that would value him at $9.9 million.

Jack Wilson's contract pays him $7.25 in 2009 with an $8.4 million option in 2010 with a $600,000 buyout clause. The difference between his value and his actual salary is $2.65 million in savings. His current contract is paying him approximately like a 1.45 WAR player, and we've valued him at 1.98 WAR. Depending on the prospects going to Pittsburgh, which I can't imagine being high-end given that Wilson's coming off a bad, injury-plagued campaign in 2008, I would like a deal for Jack Wilson, contingent on the prospects heading the other way.

Thursday, November 6, 2008

Tigers Defense ... Again

David Pinto has started unveiling his Probabilistic Model of Range data over at Baseball Musings. Basically he takes the data from Baseball Info Solutions and using some criteria winds up with some numbers that he releases to judge their defense/range.

So far, Pinto has released the data for Shortstops and for 2nd baseman. Using some simple conversions, we can turn this data into a runs total. Lets start with the Shortstop position since it was the most scrutinized amongst Tigers fans, and it was the first one that David released over at Baseball Musings.

The shortstop data included a total of 39 who qualified for this data. But before we get into the actual individual performances, let's look at how the Tigers did as a team, in comparison to the rest of the league:

Go to the shortstop link above to see the hard data, as I'm just going to be posting Team, Plays Made (where Plays Made are the Defensive Efficiency Ratio extrapolated out to the average number of balls in play for teams, which was 4,395), Predicted Plays Made (where you take the predicted DER and extrapolate it out ove 4,395), Difference between the two, and then, finally, Runs. I apologize in advance for my lack of chart-making skills. These are incredibly spaced out and I haven't fixed them, but hopefully you get the gist.

Team

Plays Made

Predicted

Difference

Runs

Brewers

558

532

26

20

Giants

510

488

22

17

Marlins

523

505

18

13

Angels

527

514

13

10

Cardinals

554

541

13

10

Red Sox

497

492

4

3

Phillies

558

554

4

3

Braves

567

563

4

3

Diamondbacks

488

483

4

3

Cubs

527

523

4

3

Astros

510

510

0

0

Athletics

488

488

0

0

Rangers

505

505

0

0

Dodgers

563

563

0

0

Indians

527

527

0

0

White Sox

545

545

0

0

Royals

505

505

0

0

Rays

505

505

0

0

Orioles

519

523

-4

-3

Rockies

567

571

-4

-3

Pirates

541

545

-4

-3

Blue Jays

497

501

-4

-3

Twins

549

558

-9

-7

Yankees

497

505

-9

-7

Nationals

523

536

-13

-10

Mariners

466

479

-13

-10

Padres

519

532

-13

-10

Reds

479

497

-18

-13

Tigers

501

527

-26

-20

Mets

505

532

-26

-20

Now, as we see, the Tigers had a lot of problems fielding the ball at Short this year. In a previous entry, I concluded that Edgar Renteria wasn't the main culprit (Ramon Santiago was very bad by my methods). But, being 20 runs in the bad is never good -- no matter the category. That equals out to roughly 2 wins worth.

Before we get to the individual numbers, a quick aside on the numbers. The runs totals are a bit skewed. To put everyone on an equal footing, I took their Defensive Efficiency Ratios that you can see on the link to the actual PMR stats (please, go check out Baseball Musings for support since David Pinto purchases this data and pays a fee to publicly put it online) and multiplied that by 4000. I do this becasue 4000 balls in play is roughly equal to 145-150 games played for an average fielder. To give an example, Marco Scutaro's DER was 0.128. You take the 0.128*4000 and get 512 plays made. He only made 173 on the season, but his pace was for about 512 if he played in 145-150 full games. You then take his Predicted DER which was 0.116*4000 and you get 464. Given the balls hit to him, he was on pace over the course of 145-150 games to field 464 of them. 512-464=48. You then plug the 48 into the 2B/SS conversions (the link posted way above at the top of this post) and get a runs total. Marco Scutaro actually made 16 more plays than predicted, since he was only in the field for 1352 balls in play. So, a lot of the guys at the top of the positions are guys with great rates, however, their actual defensive contributions were much lower than the runs totals below, but, if you extrapolated their defense out to a full season's worth of chances, this is what they were on pace to produce in the field.

So, how does PMR rate the individuals? Well...

Player

Plays Made

Predicted

Difference

Runs

Marco Scutaro

512

464

48

36.19344

Omar Vizquel

452

416

36

27.14508

Mike Aviles

476

444

32

24.12896

Maicer Izturis

464

444

20

15.0806

Jed Lowrie

432

412

20

15.0806

J.J. Hardy

500

484

16

12.06448

Erick Aybar

500

484

16

12.06448

Alex Cora

492

476

16

12.06448

Cesar Izturis

520

504

16

12.06448

Jack Wilson

512

496

16

12.06448

Bobby Crosby

452

440

12

9.04836

Jimmy Rollins

512

500

12

9.04836

Jason Bartlett

472

464

8

6.03224

Hanley Ramirez

472

464

8

6.03224

Juan Castro

460

452

8

6.03224

Luis Rodriguez

480

472

8

6.03224

Yunel Escobar

528

520

8

6.03224

Nick Punto

552

544

8

6.03224

Adam Everett

528

520

8

6.03224

Orlando Cabrera

500

496

4

3.01612

Ryan Theriot

472

468

4

3.01612

Miguel Tejada

464

464

0

0

Jhonny Peralta

472

472

0

0

Michael Young

468

468

0

0

Julio Lugo

444

448

-4

-3.01612

Angel Berroa

520

524

-4

-3.01612

Derek Jeter

448

456

-8

-6.03224

Stephen Drew

440

448

-8

-6.03224

Cristian Guzman

484

492

-8

-6.03224

Yuniesky Betancourt

428

440

-12

-9.04836

John McDonald

432

448

-16

-12.0645

Troy Tulowitzki

520

536

-16

-12.0645

Edgar Renteria

464

488

-24

-18.0967

Jose Reyes

456

480

-24

-18.0967

Khalil Greene

460

488

-28

-21.1128

Tony F Pena

440

468

-28

-21.1128

Brendan Harris

428

460

-32

-24.129

Jeff Keppinger

416

448

-32

-24.129

David Eckstein

412

452

-40

-30.1612

Here, we see Edgar Renteria is rated at -18.09 runs defensively, rounded to just -18. In my post in August about the Tigers Defense, I had Renteria at -5.56 runs. I updated all of my numbers (Here's that spread sheet [note: this spreadsheet link is NOT the PMR numbers]) and Edgar finished at -11 runs defensively. So, a difference of just 9 runs between my numbers and converting David Pinto's numbers to runs.

Well, the Shortstop position obviously needs an upgrade and that'll be something to look towards in a future post.

As for 2nd base, here's how the Tigers rated as a team:

Team

Plays Made

Predicted

Difference

Runs

Marlins

532

505

26

19.88377

Reds

510

488

22

16.56981

Diamondbacks

585

563

22

16.56981

Phillies

527

505

22

16.56981

Angels

549

536

13

9.941886

Cubs

527

514

13

9.941886

Rockies

545

536

9

6.627924

Indians

541

532

9

6.627924

Tigers

488

479

9

6.627924

Athletics

532

523

9

6.627924

Twins

488

483

4

3.313962

Blue Jays

554

549

4

3.313962

Brewers

514

510

4

3.313962

White Sox

532

532

0

0

Orioles

483

483

0

0

Yankees

563

563

0

0

Cardinals

492

497

-4

-3.31396

Red Sox

523

527

-4

-3.31396

Astros

475

479

-4

-3.31396

Mariners

585

593

-9

-6.62792

Royals

545

554

-9

-6.62792

Braves

527

536

-9

-6.62792

Pirates

440

448

-9

-6.62792

Rangers

505

514

-9

-6.62792

Nationals

461

470

-9

-6.62792

Giants

435

448

-13

-9.94189

Rays

488

505

-18

-13.2558

Mets

483

501

-18

-13.2558

Padres

470

497

-26

-19.8838

Dodgers

497

532

-35

-26.5117

The Tigers really picked it at the 2nd base position, helping to off-set the sub-par defensive play at shortstop, picking up 6.62 runs at 2B. However, that means the Tigers middle infield as a team wound up at -13, if you round the 6.62 runs up to 7 runs. That is not good, however, the middle infield "cost" the Tigers right around 1 win defensively.

As for individuals, the Tigers were obviously led by Placido Polanco. I had Polanco at the end of the year at +17.3 runs defensively, rounded to 17. How did he do in PMR? Well...

Player

Plays Made

Predicted

Difference

Runs

Adam Kennedy

484

444

40

30.1612

Mike Fontenot

484

444

40

30.1612

Marco Scutaro

536

508

28

21.11284

Emilio Bonifacio

396

368

28

21.11284

Chase Utley

484

460

24

18.09672

Howie Kendrick

528

504

24

18.09672

Dan Uggla

484

464

20

15.0806

Joe Inglett

528

508

20

15.0806

Placido Polanco

444

428

16

12.06448

Brandon Phillips

464

448

16

12.06448

Asdrubal Cabrera

516

500

16

12.06448

Juan Uribe

496

480

16

12.06448

Orlando Hudson

520

508

12

9.04836

Clint Barmes

480

468

12

9.04836

Mark Ellis

496

488

8

6.03224

Alexi Casilla

440

432

8

6.03224

Dustin Pedroia

480

476

4

3.01612

Rickie Weeks

452

448

4

3.01612

Brian Roberts

448

448

0

0

Robinson Cano

512

512

0

0

Jose Lopez

552

552

0

0

Kaz Matsui

428

428

0

0

Luis Castillo

428

428

0

0

Sean Rodriguez

484

484

0

0

Mark Loretta

464

464

0

0

Alexei Ramirez

480

484

-4

-3.01612

Mark Grudzielanek

516

520

-4

-3.01612

Tadahito Iguchi

444

448

-4

-3.01612

Freddy Sanchez

400

408

-8

-6.03224

Ian Kinsler

476

484

-8

-6.03224

Jamey Carroll

456

464

-8

-6.03224

Kelly Johnson

484

496

-12

-9.04836

Mark DeRosa

480

492

-12

-9.04836

Eugenio Velez

376

392

-16

-12.0645

Jeff Baker

472

492

-20

-15.0806

Akinori Iwamura

444

468

-24

-18.0967

Felipe Lopez

436

460

-24

-18.0967

Aaron Hill

476

500

-24

-18.0967

Ray Durham

392

424

-32

-24.129

Aaron Miles

440

472

-32

-24.129

Brendan Harris

396

428

-32

-24.129

Edgar Gonzalez

448

484

-36

-27.1451

Alberto Callaspo

452

488

-36

-27.1451

Damion Easley

424

464

-40

-30.1612

Jeff Kent

440

484

-44

-33.1773

Here we see Placido Polanco rates well, just like in my method. PMR converted to runs has Polanco at +12.06 runs defensively, rounded to 12 runs. So, my method and PMR differ by 5 runs -- not bad, if I say so myself.

So what can we conclude from these numbers? It would behoove the Tigers to improve the middle infield defense as a whole, namely the shortstop position. However, doing it at the cost of grabbing an all glove, no bat player is not a quality option as we're carrying two weak bats already with Brandon Inge at 3rd and Dusty Ryan and whatever veteran we pair with him behind the plate. However, someone like an Omar Vizquel for a late inning defensive sub (+8.54 runs in my methods, +12 runs of actual production this year via PMR converted to runs) wouldn't be a bad option to round out the bench. Though, it's more likely a reserve infielder will be Ramon Santiago or Ryan Raburn.