2003 Park Factors (October 1, 2003)
Courtesty of Robert Dudek...
--posted by TangoTiger at 05:08 PM EDT
Posted 5:42 p.m.,
October 1, 2003
(#1) -
Patriot
Me too, me too!
These are based on 5 yrs when applicable and use some regression factors suggested a long time ago by MGL.
http://f1.grp.yahoofs.com/v1/UEB7P1bJtCjOJ14uT6Xqpb2VOxevTjMytN_f3KnWuBdrh1VVf_WRVr4FukHN-vjEcr61vM0oH2A0U3xpc-aomBuqXK4/PF.xls
Posted 7:02 p.m.,
October 1, 2003
(#2) -
Tangotiger
Patriot, with yahoo groups, you should right-click the file name, do a copy of the URL address, adn post that address. They randomize URLs making this inaccessible.
Posted 7:09 p.m.,
October 1, 2003
(#3) -
Patriot
Thanks.
Maybe this would be easier though. Click on "Park Factors". This page also has an explanation of how they are figured.
http://gosu02.tripod.com/osusaber/id19.html
Posted 7:30 a.m.,
October 2, 2003
(#4) -
studes
(homepage)
Patriot, I think you need to fix the links on your page as well. They're the same gooblebygook as the original link you posted here. Can't wait to see them; I didn't know you had posted 2003 stats.
Thanks
Posted 10:50 a.m.,
October 2, 2003
(#5) -
Patriot
Oh, ok. They work on my machine, but I guess it makes sense that they wouldn't work for anybody elses. I'll get on it :)
Posted 11:16 a.m.,
October 2, 2003
(#6) -
Patriot
(homepage)
If this doesn't work, I give up.
Posted 12:02 p.m.,
October 2, 2003
(#7) -
studes
(homepage)
Patriot, the links work great. Thank you!
BTW, how do you derive park factors from Doug's data? I don't see any park breakouts on his page. Or are you using another source?
Posted 12:09 p.m.,
October 2, 2003
(#8) -
Patriot
Oh yeah, I should cite the data I used. I got the previous 4 years data from KJOK's excellent database which is floating around on the net somewhere(can't remember right now). And the 2003 data was from MLB.com.
Posted 12:30 p.m.,
October 2, 2003
(#9) -
tangotiger
KJOK's great parks database can be found at the baseball-databank yahoo group.
Posted 3:16 p.m.,
October 2, 2003
(#10) -
Rally Monkey
Robert, are you able to break down the Expos's park factor by games in Montreal and in Puerto Rico?
Posted 3:18 p.m.,
October 2, 2003
(#11) -
Rally Monkey
Are the Rockies still playing with soggy baseballs? That's the lowest park factor for them I've ever seen.
Posted 3:52 p.m.,
October 2, 2003
(#12) -
bob mong(e-mail)
(homepage)
I have a general park factors question:
With the unbalanced schedule and the outlier status of Coors field, is it accurate to compare AL park factors directly to NL park factors?
For example, check this out:
Park Factors of NL West:
Year: LA PF, SD PF, SF PF, COL PF, ARI PF
1988: 97, 98, 96 (Dodger Stadium, Qualcomm/Jack Murphy Stadium, Candlestick)
1989: 98, 100, 97
1990: 97, 101, 96
1991: 98, 103, 97
1992: 98, 102, 94 (due to three-year average, first effects of Coors are felt here)
1993: 96, 103, 96, 120 (Coors opens - note that, due to 3-year average, the full effect of Coors is not yet felt)
1994: 93, 97, 95, 116 (Full effect of Coors)
1995: 91, 96, 96, 128 (Note steady decline of Dodger Stadium)
1996: 92, 95, 96, 129
1997: 92, 93, 98, 123 (interleague play begins, first effects of the BOB are felt here)
1998: 93, 91, 96, 119, 101 (BOB opens. Full effects of BOB not yet felt)
1999: 96, 96, 89, 129, 96 (Full effects of BOB are felt)
2000: 92, 91, 91, 131, 102 (first effects of unbalanced schedule are felt: notice drop in all pitchers' parks. PacBell opens)
2001: 90, 91, 91, 122, 106 (Unbalanced schedule begins. Full effects not yet felt)
2002: 91, 92, 91, 121, 108 (This year is two-year average of 2001 & 2002. Full effects of unbalanced schedule are felt).
Notes:
Dodger Stadium had no changes in dimension between 1988 and 2002 (all ballpark info from ballparks.com). Yet, its park factor dropped from 97 or 98, between 1988 and 1991, to 90 or 91, between 2001 & 2002.
Jack Murphy Stadium (later Qualcomm) had no changes in dimension between 1988 and 2002 (except for very minor changes to the RF foul line: from 327 to 330 in 1996). Yet it went from a mild hitters' park, 1988-1993, to a severe pitchers' park, 2000-2002.
Both of these parks are in southern California, where temperature extremes and inclement weather are rare - it seems unlikely that this change could be from more rainy/cold weather.
Candlestick (later 3com) had no changes in dimension between 1988 and 1999 except for minor changes, also to the RF foul line: from 335 to 330 in 1991, and then to 328 in 1993. It seems to have had a fairly stable park factor.
My hypothesis, which I can't back up, more than you just read, is this: Coors Field, and the unbalanced schedule (especially in combination) skews the park factors of NL West clubs so much that they cannot be directly compared to AL clubs (or possibly to NL clubs in other divisions). In other words, just because PacBell had a PF of 91 in 2002 does NOT indicate that it was more of a pitchers' park than Safeco Field in 2002 (PPF=94).
Comments? Does this make sense to anyone else?
Posted 4:43 p.m.,
October 2, 2003
(#13) -
dlf
Bob, Not that it affects the larger point, but the first few years for the Rockies were played in old Mile High Stadium, not Coors Field.
Posted 5:00 p.m.,
October 2, 2003
(#14) -
bob mong
Bob, Not that it affects the larger point, but the first few years for the Rockies were played in old Mile High Stadium, not Coors Field.
Thanks for the reminder, my mistake.
Posted 11:22 a.m.,
October 3, 2003
(#15) -
FJM
Bob: You could test your hypothesis simply by removing all Rockies games, home and away, from your data base and recomputing the PF's. You should still have enough data to look at for the NL West teams on a 3-year basis, except for the Rox themselves of course.
A couple other things you might want to check. BOB has the reputation, rightly or wrongly, of being a much better hitters park with the roof open. Also, since the Questec system was installed there, umpires supposedly won't give pitchers borderline strikes there that they can still get in non-Questec parks. (Both of these observations courtesy of Curt Schilling.)
Finally, I have heard many times that the Southern Cal. parks play very differently day vs. night. Can you do the split?
Posted 2:08 p.m.,
October 3, 2003
(#16) -
KJOK(e-mail)
Bob - certainly, the unbalanced schedule will somewhat skew the park factors for every team, and in a division with parks at both extremes, the unbalanced schedule will make both extremes skew towards being even more extreme.
To "factor out" this impact, you probably need to use a method such as this one:
http://www.highboskage.com/STATINFO.HTM
Posted 11:40 a.m.,
October 7, 2003
(#17) -
studes
(homepage)
Slghtly off topic, but Pete Simpson has created an awesome spreadsheet of all players' base data in 2003. He's formatted it extremely well for anyone who'd like to play with the data. It rivals Patriot's spreadsheets, and it does contain James' version of park factors (at least, as laid out in Win Shares).
You can download it at www.baseballgraphs.com/winshares/
Posted 12:30 a.m.,
October 9, 2003
(#18) -
stasis
Thanks for the link, studes.
I thought it was interesting that Gagne was the most valuable pitcher in the NL this year according to WS and that Brandon Webb was only the 11th most valuable (and 63rd overall).
Posted 10:13 a.m.,
October 9, 2003
(#19) -
studes
(homepage)
Yes, Gagne is very highly rated by Win Shares. I actually think that's appropriate. He had an amazing years in a lot of high leverage situations. If you estimate his leveraged innings (see Tango's work) you can approximate his Win Shares total.
Obviously, Webb was not in the majors for the entire year.
Posted 5:57 p.m.,
October 10, 2003
(#20) -
FJM
It's always hard to compare starters and closers. So let's compare Gagne to Smoltz. Does anyone believe EG contributed 56% more to LA than JS did to the Braves? He threw 28% more innings, so that explains half the difference. Where's the other half?
While we're on the subject, does anyone believe Pudge was the 9th best catcher in the NL defensively?
Posted 6:52 p.m.,
October 10, 2003
(#21) -
studes
(homepage)
FJM, Gagne rates more highly than Smoltz for a number of reasons (this is off the top of my head; I've got the spreadsheet at home):
- You're right about the innings.
- Gagne's Compenent ERA was lower. He allowed less hits than Smoltz (in more innings) and he had almost twice as many K's as Smoltz. His BB rate was higher than Smoltz. Win Shares does adjust ERA for relievers, based on Earned Run components, and Gagne was much lower than Smoltz.
- I think it's an accident of the methodology, but Gagne rates higher because he depended less on his fielders than Smoltz did.
Yes, the catcher fielding rankings often look screwy. Pete probably got the most questions about those. You're welcome to e-mail him and ask him why Rodriquez ranked so low. I'll try and look it up over the weekend.
Posted 7:00 p.m.,
October 10, 2003
(#22) -
Patriot
Another annoying inconsistency in WS. It is supposed to be straight value, but it uses ERC for relievers(in concert with RA/ERA). I understand that Bill might think that since relievers sometimes do bad and give up runs but aren't charged with them ERC has some use even in a value method for relievers. But now days, you have the inherited runner stats which can help you with that. In many other places in WS, Bill has no problem using different info for different eras, why not there?
Posted 7:46 p.m.,
October 10, 2003
(#23) -
studes
(homepage)
I agree, Patriot. I'm really hoping to use this winter to try and improve the Win Shares methodology, improving it in some spots and "modernizing" it in other spots. Now that we have a year's worth of raw data, we can play "what if" scenarios to play with last year's results and see if we can get them to better match what we know today.
My long-term plan for next year is to keep posting Win Shares in-season, but also post a newer version of Win Shares based on our winter research. I know this may be a complete waste of time in the end, but at least I'll have some fun doing it.
Posted 7:51 p.m.,
October 10, 2003
(#24) -
Patriot
If that's the case, would you like a laundry list of complaints :)?
Posted 8:00 p.m.,
October 10, 2003
(#25) -
studes(e-mail)
(homepage)
Yeah, I know I'm asking for it. Once I get a little organized, I'd love a laundry list of complaints (though I think I know some of them already, from reading previous fanhome and primer threads).
BTW, if anyone wants it, I've got the complete 2003 Win Shares calculations, with all player data loaded (thanks to Pete) in an Excel spreadsheet. The calculations are built in, so if you're a glutton for punishment, you can start messing with the Win Shares calculations and come up with your own version. Just let me know if you'd like a copy of the spreadsheet. Warning: it's not "user friendly" so you'll have to figure out how it all comes together. I found it fairly easy to navigate, though.
Posted 8:01 p.m.,
October 10, 2003
(#26) -
Patriot(e-mail)
I'll take one please.
Posted 9:12 a.m.,
October 11, 2003
(#27) -
studes
(homepage)
Did some research on Rodriquez this morning, looking at Pete's spreadsheets. A good comparison is Rodriquez to Ausmus (4 Win Shares to 9 Win Shares). Comparing the two:
- They played similar number of innings (1132 to 1158 for I-Rod and Ausmus)
- I-Rod had more errors (8 to 3)
- I-Rod allowed more passed balls (10 to 3)
- I-Rod allowed less stolen bases and they had similar SB% (40/60 for IRod and 68/105 for Ausmus). Win Shares only looks at CS% on the position level, not absolute SBs. It does look at absolute SBs allowed for individual fielders.
So, while the record is mixed, it is certainly hard to justify the position that Ausmus contributed twice as many fielding wins to the Astros as I-Rod did to the Marlins.
I think the biggest issue is that Win Shares assigns defense Win Shares to a position first, then to individual players. And while Rodriquez is great, Redmond is a pretty bad catcher, and this brings down the fielding Win Shares claim points available to Marlin catchers.
So there are a few wrinkles we could play with here. The stolen base methodology, for one. And definitely the position/individual splits. This, in fact, may be the place to introduce "loss shares," to ensure that lousy fielders don't bring down good fielders.
Posted 11:52 a.m.,
October 11, 2003
(#28) -
Pete(e-mail)
Dave,
You're at least partially right about Redmond dragging down Rodriguez. The only place I can see an impact is in stolen bases.
Marlin catchers other than Rodriguez [Castro and Redmond] were terrible at stopping stolen bases (30 of 35 successful), while other Astro catchers were just slightly better than Ausmus (18 of 29). If we assume that the Marlins' catchers have the same SB% allowed as Rodriguez, Rodriguez's fielding win shares would increase by about 2.
However, Ausmus does rank ahead of Rodriguez in other, not so obvious, areas. Namely:
- Ausmus had more assists (76 to 47). [Other Astro catchers had 11 while other Marlin catchers had 13.]
- Astro catchers had more non-strikeout putouts (50 to 28).
- By combining the above two factors with team catcher errors (Astros 7 [4 by other catchers], Marlins 10 [2 by other catchers]), we get the catcher non-strikeout error percentage. The Astros' was much better than the NL average (.049 vs. .081) while the Marlins' was worse (.102).
- This causes a difference of about 2 fielding win shares between Ausmus and Martinez.
- There is no evidence that this is caused by the other catchers.
Pete