Yesterday we looked at a projected outcome to the 2011 season from the Giants offense using a new projection used by fangraphs.com called RotoChamp. Let's take a look at what the pitching might do:
Starters:
Tim Lincecum: 17-8, 2.93, 217 IP, 75 BB, 242 K's.
Matt Cain: 16-9, 3.25, 220 IP, 68 BB, 176 K's.
Jonathan Sanchez: 11-10, 4.05, 184 IP, 93 BB, 196 K's.
Madison Bumgarner: 11-9, 3.84, 175 IP, 41 BB, 137 K's.
Barry Zito: 9-9, 4.44, 180 IP, 78 BB, 136 K's.
Relievers:
Brian Wilson: 3-3, 2.48, 72 IP, 26 BB, 87 K's, 37 Saves.
Sergio Romo(Bill James): 5-2, 2.29, 59 IP, 15 BB, 64 K's.
Jeremy Affeldt: 2-2, 3.20, 53 IP, 25 BB, 48 K's, 4 Saves.
Santiago Casilla(Bill James): 4-4, 3.86, 77 IP, 37 BB, 76 K's.
Ramon Ramirez: 3-3, 3.72, 69 IP, 28 BB, 50 K's.
Javier Lopez(Bill James): 3-3, 4.11, 57 IP, 25 BB, 36 K's.
Guillermo Mota(Bill James): 3-3, 3.49, 49 IP, 19 BB, 38 K's.
Jeff Suppan(Bill James): 4-6, 5.17, 94 IP, 33 BB, 50 K's.
Dan Runzler(Marcel): 3-2, 3.86, 42 IP, 18 BB, 40 K's.
If you take the Starters plus the 7 most likely relievers, you get a record of 90-65 which leaves 7 games unaccounted for. If you add in Runzler you get to 93-67 with just 2 games unaccounted for. Of course, it's highly unlikely that a team goes a whole season using only 13 pitchers. W-L are the toughest to project. The tendency of all projections systems is to regress to .500 because they are so unpredictable. I think Giants fans would not be too surprised or unhappy with a 93-94 win season. Given the projections we are seeing for the Giants offense, if you take W-L out of the pitching projections, the Giants could easily do better than 93-94 wins.
Of the specific projections here, I think Sanchez is a bit pessimistic. Lincecum and Cain are about right on ERA but maybe a bit optimistic on W-L. Romo is very optimistic. Lopez is pessimistic and Ramirez optimistic. I think I would flip-flop their ERA's.
What do you think?
Subscribe to:
Post Comments (Atom)
too many variables
ReplyDeletewhat hasnt been taken into account, is the amount of innings all these guys threw last season
bochy was very good keeping the pen arms fresh, im just a little worried about the starters
im hoping that he limits all their pitch counts in the early going
Lopez sucks look at his career stats, that's right where he will probably be. Ramirez is correct, Mota is very optimistic that would be in line with his career years which makes no sense as to why they think he will have a career year, Romo is very optimistic but it's possible he can do that, they're playing it safe with Wilson and Lincecum as that is right where they fall if you average the last two seasons, Cain was a lazy projection as that is right in line with what he always does, Casilla and Affeldt I'll buy that, Bumgarner is not getting any credit, nor is Sanchez or Zito. They project Zito will have a much worse year than he did the last two years and that Sanchez will be much worse than last year, i don't buy either.Also, Runzler is much better at Ks that what they gave him, i also don't buy what they projected him. All in all, very interesting but i don't agree with them or think that they are credible. We'll see how the season goes.
ReplyDeletePlus, Mota has already been released just recently.
ReplyDeleteThe poor starters projection is because they are relying on FIP and xFIP to project Bumgarner, Sanchez, and Zito, though that usually mean Cain is up too, perhaps Paapfly sold him on Cain. I think they will all do better and that will up the win total.
I have been saying that I think mid-90's win can be expected and I would not be surprised by wins around 100, and these projections do not change my mind.
OGC - Where did you hear that mota had been released? I thought that he signed a minor league deal with an invite to camp. Was this recent news? Thanks.
ReplyDeleteOK, maybe this was a mistake. I just thought it would be fun to look at how the season might play out. It's actually more optimistic than most projections. Have you ever looked at a typical fan's projections for their own team? They usually add up to some all-time record for HR's or ERA or Wins or something like that, just ridiculously unrealistic.
ReplyDeleteOGC,
I'll flat out predict Lopez ends up with a lower ERA than Ramirez. Not sure what's wrong with projecting Cain to keep doing what he's been doing. That's not lazy, that's just common sense. I agree that they are pessimistic on Sanchez but every other projection system is too.
Why would these projections change your mind about mid-90's wins? That's right where these numbers put them!
The guys who made up this system will tell you they fully expect that some of these players will do better than their projection, some will do worse. It's projection for crying out loud!
Projection systems, by their very nature, do not project breakout seasons and are suspicious of breakouts or outliers which only makes sense. They are always going to be quite conservative. I think this one actually takes more risks than most. I mean MARCEL pretty much thinks Wins are a crapshoot so goes for about .500 on almost everybody.
ReplyDeleteSorry OGC, I meant to put your initials above the 3'rd paragraph in my previous comment.
For you, sir:
ReplyDeletehttp://www.fangraphs.com/blogs/index.php/matt-cain-isnt-that-unique-after-all/
Just to continue the great Matt Cain debate
(and i still say Cain's going to post an ERA between 3.5 and 4.0...)
ReplyDeleteDavid,
ReplyDeleteI saw that article this morning and posted a response. Goes along with what i've been saying for a long time. It's not just Matt Cain, although he is the most striking example. The Giants have an organizational philosphy of HR aversion. They would rather walk a batter or two than throw one down the pipe and risk having it hit out of the park. Thanks for the link.
No problem. keep up the good writing.
ReplyDeleteBut i still think the STDEV is overblown using HF/FB rather than HR/OFFB, due to variance differentials b/w the 2
ReplyDeleteThanks DrB, I was wondering...
ReplyDeleteSorry if I did not phrase it the way I intended: put another way, these projections do not say anything that would suggest that mid-90's wins won't happen.
And I do understand the limitations of projection systems, I've dealt with forecasting for most of my career. I agree with all of your statements regarding projections.
However, I stand by my statement regarding those pitchers, I believe that Roto-Champ probably rely on FIP and/or xFIP in some way, fashion, or methodology, resulting in lowered projections for those players. Particularly if they are associated with Fangraphs.
Part of using projections is knowing the limitations of where they might fail. I believe that they fail for pitchers who don't fit the DIPS mode, and the Giantshave pitchers (Zito, Cain) who don't fit the theory. So I believe they will usually be forecasted for lesser performance than they should.
They also fail for pitchers who are not like they were when the projection systems is using that data as input into their forecast, like Sanchez. He has had issues in prior seasons that feed into the projection, like his lack of stamina in 2008 and his poor mechanics which hurt his first half of 2009, those feed into the projections too.
I also find that most systems do not do a good job of adjusting for age and the minors when handling pitchers like Bumgarner. I'll admit I don't know Roto-Champ's system, but based on what I see projected by a system I am familiar with and like - Baseball Forecaster/Ron Shandler - clearly Roto-Champ is not accounting for how well Bumgarner did in the minors in 2009 and 2010 and at his age, which Baseball Forecaster translated into great MLE's for Bumgarner, and thus project 3.47 ERA for 2011.
And for those assuming bias on my part, all I care about is the best projection to use, whether that means good or bad for my team's players, I would note the problem if I think the projection is much better than what I thought is a good and proper projection. All I care about is have the right perception of where my team stands, for good or bad.
My apologies about Mota too, I would have bet anything that I saw in a recent Mercury blurb that Mota was released. I can't find that now, so I must have been mistaken. Sorry!
No problem on the Mota thing OGC. I actually like his arm out of the pen in non pressure situations and would have been a little upset to see him gone.
ReplyDeleteDavid,
ReplyDeleteThanks for the props. I still think we have a lot to learn from PitchFx data. I believe that a lot of things we currently attribute to chance or luck will ultimately prove to have other explanations.
Dr B
OGC,
ReplyDeleteSince these projection systems are generally very conservative, I like to think of them as the worst case scenario, barring injury of course. In that light, I think the projections for the 2011 Giants, even Marcel, are quite encouraging.
I greatly agree with you. That's why I try to state nothing too definitely and note that when I do, it's personal belief and not fact.
ReplyDeletePitch F/X will teach a lot in the near future. I wish I had time to learn pitch f/x data charts and MySQL so I could read data points better rather than email my contemporaries to run diligence for me.
Inspired by our debates, I modified the xWHIP calculator's next release so you can change normalization/regression points in case you do not believe a player "fits the mold" of regression
DME,
ReplyDeleteYes, I think PitchFx data is going to prove too daunting for amateur analysts even if you are a Math Professor just because of all the time it will take just to enter the data points.
I'm afraid PitchFx is about to take sabermetrics back out of the hands of "armchair GM's" and put it back into the hands of people who do it full time for a living.
Anyway, at least people like me can sit back, look at the results and hopefully make constructive criticisms.