Tom Tango has just released his yearly projections via his projection system Marcel. You can find them on FanGraphs — individual player pages or sortable leaderboards — or you can download the entire .zip file on Tango’s own website. Marcel is a very basic projection system that uses 3 years worth of data, with the most recent production weighted the heaviest, some add age adjustments, and regression to the mean.
The Bill James projections are already out, but I’m not crazy about them and Marcel is another system to check out. With that in mind, I’ve collected projections for hitters that could see some time on the 2009 Giants team. I then used the wOBA scores and calculated bRAA — batting runs above average — and RV700 scores. If you’re unfamiliar with these statistics, I’ll explain them briefly blow. Alternatively, you can check out StatCorner’s glossary for definitions of these statistics and many others.
wOBA (Weighted On-Base Average) is a linear weights statistic devised by Tom Tango. This is a great introduction to wOBA. wOBA takes the series of events that can happen when a hitter steps to the plate (1B, 2B, 3B, HR, RBOE, BB) and weights each event accordingly. HR’s are worth more than 2B’s, and 2B’s are worth more than 1B’s. It’s very intuitive and easy to follow, think of it as a more refined OPS. League average wOBA tends to be around .338 — I’ve used .332 for my calculations below because offense was down in 2008 and Marcel assume it will also be down in 2009. Hitters with a .400+ wOBA are considered very good hitters and ones with wOBA of .300 or less, are considered poor hitters.
bRAA (Batting Runs Above Average) is how many runs above average a hitter is credited for by hitting alone. It’s computed by taking RV/PA (Run Value per plate appearance above average) and multiplying by how many PA’s a hitter has accumulated over a season. PA’s are also projected for our hitters, but what would happen if he hit for a full season? That’s where the next statistic comes into play.
RV700 (Run Value per 700 PA’s) is RV/RA multiplied by 700 PA’s, essentially a full season for a hitter. This lets us see how much a hitter would contribute if he batted for 700 PA’s.
Now, let’s check out the numbers:
Click to enlarge
I’m still on a color-coding table kick, so forgive me if the colors drive you crazy. The colors are pretty self-explanatory, green indicates a good score and red indicates a poor score. Also, keep in mind that these numbers — bRAA, wOBA — aren’t park adjusted and I’ve made no positional adjustments. You could probably fiddled with the numbers a little if you wanted to.
Thoughts
- The top 5 hitters on the Giants by bRAA are: Pablo Sandoval (5.54), Fred Lewis (3.74), Nate Schierholtz (3.07), Conor Gillaspie (3.02), Aaron Rowand (2.99). Probably the most surprising hitter of the bunch was Conor Gillaspie who Marcel projected to get 204 PA’s in ’09 and hit for a line of: .278/.355/.444. Playing time is hard to project for these sort of things, but even Gillaspie’s wOBA of .349 is pretty favorable. But, you’ll notice that his reliability score is very low indicating that Marcel is taking a shot in the dark with Conor’s projection.
- On the other hand, I think Marcel under-projected Sandoval’s playing time who looks to be a large part of the Giants team in 2009. Marcel projected Sandoval to only get 277 PA’s but his wOBA of .355 is the best among Giants hitters. His projected bRAA was 5.54 runs above average for a hitter but if he batted for close to 700 PA’s, it would be +14 runs above average with the bat. That’s one of the reasons that position is important for Sandoval. If he’s a catcher putting up +14 runs above average with the bat, then he’s above average for his position and solid, maybe a borderline All-Star. The offensive requirements at 1B/3B are much tougher. Pablo’s value plays best at catcher.
- Going back to Pablo for a second, Marcel has his slash-line at: .310/.355/.472 with a BABIP of .341. Pablo had a huge BABIP of .361 in ’07 but Marcel only knocks about 20 points off in the projection. I thought it was interesting that Marcel is predicting Pablo to K more in ’09 as compared to ’08. His K% jumps 5% from 9.7 to 14.7.
- Molina takes a slight dip in this projection. His wOBA of .338 in ’08 is almost 20 points higher than his projected wOBA of .318 for ’09. Don’t worry too much because a .318 wOBA catcher is still most likely average for his position (I haven’t ran the wOBA numbers for catchers in the NL, so correct me if that’s wrong) and he will be a 34-year-old catcher with weight issues heading into next year. Molina is a bit past the age when catchers go poof, so he might be more durable than we think, but he should still be watched closely.
- Marcel essentially has Fred Lewis holing steady from his ’08 numbers. His projected line is: .279/.350/.426. Lewis’ BABIP was a talking point among Giants fans last season and Marcels adjusts it some by knocking it down 20 points to .347. Lewis loses a touch of power but his OBP stays the same. If Lewis could repeat his ’08, or at least come close to it in, the Giants would be very happy.
- The most futile Giant with a bat in his hands? Say hello to Mr. Ivan Ochoa. His bRAA of -6.97 runs was the worst among projected Giants. If Ochoa somehow accrued 700 PA’s in a season, he’d be -18 runs by his bat alone. Ouch. Brian Bocock isn’t too far behind him with his score of -4.49 batting runs. I was surprised to see Bocock’s projection crediting him with hitting 5 HR’s in 240 PA’s.
- I was expecting Burriss’ projection to be a lot more pessimistic than the Bill James projection, but Marcel was even more optimistic than Bill James was. Marcel projects Manny to have an wOBA of .334. I have to admit, I still have some problems with this Burriss projection. It’s adding 20 points of wOBA to his ’08 score and it’s even projected him to hit 5 HR’s. I’ll take the under on this projection but it’s nice to dream.
- Marcel has Rowand pretty much staying the same ast last year with a wOBA of .338, but he should play a little above average in 2009 because scoring is projected to be down — a league wOBA of .332.
Interesting stuff, like all projection systems the numbers are fun to look at but your mileage may vary. I’m most looking forward to the ZiPS and PECOTA projections for the Giants hitters and when they come out, I might do something similar like I’ve done here with Marcel.
Comment Starter: Any projections catch your eye?
Very interesting. Kind of shocking to see not only Gillaspie, but Ishikawa and Rowand in the green, particularly since many have pillored Rowand. As well, in the negative sense, to see Randy Winn in yellow and Molina in red.
And Ishikawa actually had a wOBA of .346 in 2008′s Marcel’s, higher than Lewis’s .344 (thanks for link to Marcels!). Which I would say is pretty interesting given that Ishikawa did not have what any would call a good season in 2006 and 2007 in the minors in AA, so the projection was strictly based on how well he had done previously in
A-ball and below. In fact, in spite of how well he did in 2008 at all higher levels, his projected wOBA is actually lower than in the 2008 projections.
The above data looks like a pretty good justification to do what the Giants are doing now, putting Pablo at 3B with 1B perhaps filled internally with Ishikawa, meanwhile force Burriss and Frandsen to compete for 2B while (hopefully) getting a player with green across for SS.
Yes, very interesting, I would like to see you do something similar with ZiPS and PECOTA.
OGC,
I forgot to mention in my post and I should have but Marcel doesn’t look at minor league data, it only looks a MLB data. So, one of the reasons that Ishikawa is projected “decently” — still on the very low end for a 1B — is that he did well in his brief run in the majors last year. Same for Burriss.
The Gillaspie projection is basically a wild guess. The 3% reliability score on that says so, I had Tango tell me that projections with reliability scores under 60% should be thrown out.
I’d like to see the Giants play Pablo behind the dish because a +14 catcher is way more valuable than a +14 1B or 3B, where the offensive requirements are raised.
Overall, these are still some weak-ish projections for Giants hitters. Pablo is probably the most interesting even if his reliability score is below the 60%.
And also, the color-coding is just for Giants hitters. I expect if I tossed them in with all the other hitters in the majors (and figured out battings runs, etc.) we’d have way more Reds than Greens because everyone else is most likely going to be hitting better than us, which tilts the scale some.
I wanted to give some context of how Giants hitters are rating among other MLB hitters when using RV700.
Pablo Sandoval was projected to +14 batting runs. Here’s a brief list of other +14 hitters: Russell Martin, Mike Fontenot, Travis Snider, and Hunter Pence. Also, you must note that Pablo’s reliability score is rather low, so we should take his numbers with a grain of salt.
The next best hitter among 700 PAs on the Giants is Nate Schierholtz. Again, his reliability score is quite low so take his numbers with some skepticism. Nate was a +8.5 hitter and some of his company is: Ian Stewart, Josh Phelps (I missed him the first time around), Orlando Hudson, Fernando Tatis, and Ryan Garko.
Lewis was +3 and his grouping includes: Jason Kubel, David DeJesus, Jacoby Ellsbury, and Greg Norton.
Once again, these aren’t adjusted for postion so just keep that in the back of your mind. A +5 SS is much better than a +5 1B.
And if anyone is wondering, Albert Pujols is the highest ranked hitter with a +55 batting runs above average when extrapolated to 700 PA’s. He’s the only hitter to project into the +50′s.