**UPDATE 5:25PM** I went through and added the value of each player in terms of 2008 Free Agent value, according to the FanGraphs and, I believe, Tom Tango's methodology. Our payroll, the the ether-world of theory, would be $144.07...as it stands right now. I'm sure there will be some corrections downwards, because we look too good right now.
***********
**UPDATE 3:30PM** Some adjustments to base running and fielding have been made, along with a new measure for SP ERA next year (per Sky's suggestion). The base running measures are from BPro and the SP ERA is a weighted measure of the last three years worth of each SP's tRA* (the weights were not uniform for Moehler and Wandy). The adjustments leave us at 82 projected wins, but I'm going to overhaul the wOBA's on offense and I believe Sky is looking into the nuts and bolts of his formulas.
The one thing that we're missing is more insight into what doesn't look/feel right about these projections. This is intended to be a community projection system...so we need community-wide input.
***********
SkyKalkman, of Beyond the Box Score, has issued a challenge for fans and bloggers of each team to come up with projected WAR—thereby projected records—for their teams this offseason. If you mozy on over to BTB, you'll see the current returns and it's quite impressive for it being the holidays and the commotion having just started two days ago.
After emailing Sky back and forth (read: pestering and pleading for help in the areas I was woefully ignorant in) Sky graciously allowed me to be a test dummy for a generic spreadsheet he created that has standardized values for positional adjustments and weights for WAR. I went through and averaged together Bill James' and Tom Tango's projections for offense; made some educated guesses, using UZR data from FanGraphs, for defense; kind of guessed on the base running effects (BPro and Bill James have some good stats to look at for truly quantifying these) and PA and playing times at positions; used tRA* from Stat Corner for SP ERA (Sky requested a luck/park neutral stat of some sort), and Bill James and Tom Tango averages of projected FIP for relievers. The current projections are rough to say the least.
However, as a community of diehard Astros fans we hold the kind of collective knowledge necessary to hone and refine this projection to near perfection. At the bottom of this post is a link to my filled in spreadsheet in Google Document format. Take a gander at it, mull it over, look at some projections, and then leave your corrections/additions in the comments. We'll keep playing around with it until it's gold. If you want your own copy of the spread sheet, hit me up by email (it's on my profile) and I'll send you a copy. This should be fun and slightly up lifting. So far I have the Astros pegged at about 79 wins next year and I don't have the bullpen or the starting rotation fully accounted for...but I also don't have the bench fully accounted for either...
Finally, look for the Astroiest Bracket to be released on New Years day. It'll be a 32 entity battle royale, with some fancy polls that allow you to vote on the first 16 match-ups all in one post.
As promised, the first pass at projecting the 2009 Houston Astros.
0 recs | 21 comments
A few initial comments just from glancing..
1. I would probably put Matsu’s fielding at zero or better, if I were forecasting. UZR has Matsui at +1 something for his career at 2d base. Matsui is hurt by bad results last year on most of the defensive systems, but PMR had him as a slight plus on defense.
2. I don’t know how Pence can be zero on baserunning, when Bill James’ metrics say he is terrible…basically the second worst Astros base runner…right next to Carlos Lee.
3. I would disagree strongly with any projection that has Oswalt with an ERA over 4. He has never had an ERA as high as 4 in his career. I am not a betting man, but that is one bet I might be inclined to take.
clack - December 26, 2008
just to elaborate on my comment about Oswalt...
I just ran a regression on Oswalt’s career BABIP, ERA, Fly/HR, and GB% numbers. Oswalt’s ERA has no correlation with BABIP (R squared is less than .01). His ERA has no correllaion with Fly/HR (R squared almost same as for BABIP). There is a negative correllation between Oswalt’s ERA and his GB % (-.30). The latter is a bit surprising, and is worth thinking about.
My point is that Oswalt’s performance in terms of results historically doesn’t vary with BABIP.
clack - December 26, 2008
Correct that...the negative correllation with groundball rate is in the expected direction.
clack - December 26, 2008
Noted and changed.
Also, I took a stab at the Base Running numbers using BPro’s stats (I don’t have a Bill James Account).
81.2 wins right now.
Stephen Higdon - December 26, 2008
Ouch
Pence second worst baserunner next to Carlos Lee.
ol Pete - December 26, 2008
It's mental mistakes.
Pence is great in one area…beating out infield singles. But he has bad stolen base success and makes too many blunders on the base paths. Lee probably does a better job on stolen bases, even though part of that is surprise.
clack - December 26, 2008
By the way, have you seen Fangraphs' WAR...
for the Astros?
http://www.fangraphs.com/winss.aspx?team=Astros&pos=all&stats=bat&qual=0&type=6&season=2008&month=0
I’m not saying they are necessarily right, but it might be worth comparing what they arrived at vs. your calculation to understand why they are different. (Obviously, one big difference is that they are measuring 2008 instead of a projected 2009.) The fangraphs calculation is interesting because it arrives at a monetary value….so you can see who is underpaid and who is overpaid.
clack - December 26, 2008
I have seen the FanGraphs WAR
I should/will add a column for calculating value later today.
The big difference (and I’m sure that there are many smaller ones) is that FanGraphs has their model set to fill a position in 600 PA, while Sky, Tango, and many other people set their models for 700 PA to fill the season. So it’s going to over value some guys, because they get to fill a greater majority of the season.
The other major difference, I’m sure, is that these defense values are just averages from years past, so Berkman takes a little knock, etc. But maybe we should more confidently project defense from a our players…but we need some discussion on it all to come to a consensus.
BTW: Thanks for all your comments…they’ve already made the projections much more thought out…not just mechanical printouts from some model.
Stephen Higdon - December 26, 2008
I mentioned Matsui above...
and I notice that CHONE projects Kaz at +1 on defense at 2d base. The CHONE projection for Boone’s defense at 3d base is -13 (yikes!).
clack - December 26, 2008
I haven't taken a look at it since the season ended,
so I’m working from recall here, but didn’t Berkman put in a pretty good defensive season this past year? I seem to remember that going into BTB’s discussion of Pujols vs. Lance for MVP. The FanGraphs WAR chart gives him a -12 run projection on defense. I suppose that depends on whether or not you think last season was a fluke for Berkman (just as with his improved baserunning), but I tend to think that, while his defense might not be as good next year, I don’t think it will return to his previous levels.
Only_A_Lad - December 26, 2008
I can't find WAR projections on FanGraphs
Will you post the link?
Stephen Higdon - December 26, 2008
No projectied WAR are on Fangraphs currently.
Fangraphs only shows WAR for 2008 and past years on team batter’s page, under projections. (I show the link in my earlier post for Astros’ 2008 WAR.) One of the articles says that they are working on putting up a projected WAR page for 2009.
clack - December 26, 2008
Berkman is +11 or so runs on defense, on Fangraph's WAR.
The -12 is a positional adjustment. It is supposed to reflect the lower value of first base vs. other positions. So Berkman, Pujols, and all other first basemen would get the same positional adjustment (or at least pro rata based on playing time). This is a somewhat different approach to adjusting for positions. Some systems, like BP’s, adjust offensive stats to reflect position differences. Tango and others have attempted to quantify the difference in defensive value between positions, meaning that 12 runs saved on defense at the shortstop position is more valuable than 12 runs saved on defense at first base.
clack - December 26, 2008
You're right on
FanGraphs is adjusting for the defensive skill of the position and the expected offensive production of a 1B.
Sky actually has a really good article on trying to adjust for defense and offense between positions at BTB.
Stephen Higdon - December 26, 2008
yeah, I mis-read that one
whoops
Only_A_Lad - December 26, 2008
Cool stuff.
Evidently my estimate of offense outs is way off. Need to look into that. High-OBP teams get a lot more PAs than low-OBP teams and I was trying to estimate that via wOBA. Huh.
I really like using tRA* as a neutralized ERA measure. But I’d look at the past few years, maybe weighted 50/30/20%. And make sure to multiply by .92 to get it onto the ERA scale, not RA.
Great start and I hope this sparks discussion about each individual piece of every player’s game, to make the projections the best they can be.
I prefer CHONE, PECOTA, and ZiPS to Bill James and Marcel (the last of which is pretty pointless for players without much MLB experience). FYI.
Oh, and you didn’t pester me at all.
Sky Kalkman - December 26, 2008
I'll go back through and add CHONE to it all
And PECOTA when it’s released. This was just a quick and dirty way of getting something on the spreadsheet to see how it all played out.
I’m in the process of coming up with tRA* weighed averages and I’ll replug those in.
Thanks again for all your help Sky, I think this has the potential to get really big in the future.
Stephen Higdon - December 26, 2008
Cool.
If you or anyone else here has interesting ideas on how to draw threads like this together from all 30 teams and make it useful, please let me know.
I love the ability for you and your readers to make changes throughout the off-season as information is uncovered.
And I’m looking forward to seeing what the average win projection is across all MLB teams. It should be 81, but I’m pretty sure most folks will be at least a bit optimistic, even if not on purpose.
Sky Kalkman - December 26, 2008
I've already found myself battling that
It’s hard to be completely objective, but hopefully once everyone’s cards are the table we can start making adjustments.
Stephen Higdon - December 26, 2008
Alberto Arias?
Where’s LaTroy Hawkins?
entropic soul - December 27, 2008
Duh!
See, this is why everyone’s gotta pitch in…help my dumbass out.
Stephen Higdon - December 28, 2008
You must Login with your SB Nation account and be a member of The Crawfish Boxes to post a comment.