Baseball for the Thinking Fan

Login | Register | Feedback

btf_logo
You are here > Home > Transaction Oracle > Discussion
Transaction Oracle
— A Timely Look at Transactions as They Happen

Monday, February 04, 2008

2008 ZiPS Projections for Diamond Mind 9 and Microsoft Excel, Build 1.1

The title is pretty explanatory. 

2008 ZiPS Projections for Microsoft Excel

2008 ZiPS Projection Disk for Diamond Mind 9

2/8 Update

The DMB disk has been updated with SG’s version with depth charts, plus a few new signings (Hinske, Miller, etc.) a couple of other minor league signings, Slayden moved to PHI, and projections for Ramon Nivar (256/295/331, Edgar Gonzalez (259/319/370), and John Hudgins (4.98).  Remember to let me know of additional player requests!

So, what’s done here?  Projections for 1036 pitchers and 1005 hitters.  If you play Diamond Mind, 40-man rosters (with the exceptions of players I don’t do like whathisface from Cuba and David Price) and as many non-roster invitees as I have been able to verify are on the team, either as a non-roster invitee, an offseason minor league signing, or a carryover player not eligible for minor league free agency.

SG has provided the 2008 schedule, so that’s done.

What’s then left is making any corrections, adding any players that you wish to see projected that I can reasonably do (some 300 players here did not have projections for the team-by-team rundowns), and eventually making rosters.  In particular, if any DMB players would like to contribute a manager tendencies profile,  it would be a big help and save time of the odious task of whittling the 40-man rosters down and allocating playing time for sims.

To import the 2008 disk into DMB, create a new empty database, select “Restore” and choose the zip file provided.


Creative Commons License


This work is licensed under a
Creative Commons Attribution-Noncommercial 3.0 United States License.

Dan Szymborski Posted: February 04, 2008 at 06:09 AM | 116 comment(s) Login to Bookmark
  Related News:

Reader Comments and Retorts

Go to end of page

Statements posted here are those of our readers and do not represent the BaseballThinkFactory. Names are provided by the poster and are not verified. We ask that posters follow our submission policy. Please report any inappropriate comments.

Page 1 of 2 pages  1 2 > 
   1. 1k5v3L Posted: February 04, 2008 at 06:59 AM (#2683068)
Awesome, Dan. Thanks for the hard work, wonderful projections, and great laughs at TO.
   2. Phil Coorey is a T-Shirt Salesman Posted: February 04, 2008 at 07:46 AM (#2683074)
Thanks dan - all good as always
   3. Danny Posted: February 04, 2008 at 07:58 AM (#2683075)
Thanks, Dan. Is the RC/27 park adjusted?
   4. shock Posted: February 04, 2008 at 08:19 AM (#2683082)
Thanks, Dan. You're the best.
   5. SG Posted: February 04, 2008 at 03:29 PM (#2683171)
Dan's the man.

I'll be setting up manager profiles/depth charts, and should have it done by the end of the week.
   6. Bring Me the Head of Alfredo Griffin (Vlad) Posted: February 04, 2008 at 03:33 PM (#2683173)
Sweet. Thanks, Dan.
   7. Ennder Posted: February 04, 2008 at 04:18 PM (#2683199)
Does using ZiPs and DMB end up counting defense twice?

I mean ZiPs itself shoots out it's projections including either projected defense from this year or actual defense from last year (not exactly sure how it works). DMB uses the defensive ratings to modify what happens on each play. Is that resulting in defense counting twice?

I ask because when running seasons to see how the NL Central comes out the Cubs(good defensive team) pitchers almost always beat their projections overall and the Brewers(bad defensive team) almost always fall short and as I understand it those projections themselves already included defense so this shouldn't be happening?
   8. CoastalFan Posted: February 04, 2008 at 04:40 PM (#2683214)
Dan, you SO rock - thank you!
   9. CoastalFan Posted: February 04, 2008 at 04:42 PM (#2683220)
Quick question - what fixes the error "unable to write to DBTteam file"?
   10. RoyalsRetro (AG#1F) Posted: February 04, 2008 at 04:50 PM (#2683229)
DAN IS AWESOME!!!!!!!!!!!
   11. Mooser Posted: February 04, 2008 at 04:56 PM (#2683234)
I agree with Ennder. If the ZIPS pitchers projections for ERA and Hits etc. already account for the projected defence behind them, then defence will be counted twice. The projections need to be defence neutral to prevent this.
   12. galaxieboi Posted: February 04, 2008 at 05:08 PM (#2683241)
Dan rules!

I just got my very first DMB on saturday so the timing couldn't be better. All these years I've been using APBA's computer game, so I'm stoked for a change.

With all the hard work you do Dan, how can we offer some financial 'thanks'??
   13. Dan Szymborski Posted: February 04, 2008 at 05:34 PM (#2683264)
I agree with Ennder. If the ZIPS pitchers projections for ERA and Hits etc. already account for the projected defence behind them, then defence will be counted twice. The projections need to be defence neutral to prevent this.

It doesn't quite work out this way because of how pitcher event tables are generated - the event table doesn't care about the earned runs requested at all and only cares about a pitcher's BABIP to a small degree, which I compensate for by being very conservative with my ratings. In other words, if Joe Schmo has a 3.40 ERA but his peripherals are that of a 4.60 pitcher, DMB's event table is already considering Schmo to be much closer to a 4.60 pitcher than a 3.40 ERA pitcher, not considering Schmo a 3.40 ERA pitcher and then modifying it from there with defense.
   14. Dan Szymborski Posted: February 04, 2008 at 05:35 PM (#2683266)
With all the hard work you do Dan, how can we offer some financial 'thanks'??

A paypal contribution to dan@baseballprimer.com or a contribution to the site as a whole by clicking the little button above the hot topics are always welcome.
   15. Mooser Posted: February 04, 2008 at 06:15 PM (#2683308)
It doesn't quite work out this way because of how pitcher event tables are generated - the event table doesn't care about the earned runs requested at all and only cares about a pitcher's BABIP to a small degree, which I compensate for by being very conservative with my ratings. In other words, if Joe Schmo has a 3.40 ERA but his peripherals are that of a 4.60 pitcher, DMB's event table is already considering Schmo to be much closer to a 4.60 pitcher than a 3.40 ERA pitcher, not considering Schmo a 3.40 ERA pitcher and then modifying it from there with defense.

Perfect, thank you
   16. Dan Szymborski Posted: February 04, 2008 at 11:29 PM (#2683611)
Always good when most of the work is completed!

My next task is my overhaul of minor league translations - I'd really like to dump linear relationships.
   17. harrball Posted: February 05, 2008 at 02:00 AM (#2683701)
Thank you Dan! SG, thanks in advance.
   18. Masticore317 Posted: February 05, 2008 at 02:44 AM (#2683714)
This is fantastic!!!

A few Padre suggestions - Ramon Nivar, Edgar Gonzalez, John Hudgins. Also, I believe Jeremy Slayden is with the Phillies, not SD?
   19. Dan Szymborski Posted: February 05, 2008 at 02:49 AM (#2683719)
A few Padre suggestions - Ramon Nivar, Edgar Gonzalez, John Hudgins.

Sure!

Also, I believe Jeremy Slayden is with the Phillies, not SD?

Oops, I think I added him to SD with Sinisi.
   20. Der Komminsk-sar Posted: February 05, 2008 at 02:49 AM (#2683720)
As always, thanks so much Dan.

My next task is my overhaul of minor league translations - I'd really like to dump linear relationships.
Glad to hear that...

I'll try to come up with a (partial) list of team changes as I go through this - my time is a little more limited than I'd like these days...
   21. Der Komminsk-sar Posted: February 05, 2008 at 02:50 AM (#2683721)
(oddly enough, slayden was the first one i noticed...)
   22. Chris Needham Posted: February 05, 2008 at 03:01 AM (#2683723)
An Ed Rogers/Pete Orr platoon at shortstop for the Nats? Yegods.

Of course when the alternative is Guzman...
   23. Dan Szymborski Posted: February 05, 2008 at 03:04 AM (#2683725)
Oh, I didn't make those manager profiles, I just asked DMB to generate quick profiles so I could have fun and sim a few seasons.
   24. Der Komminsk-sar Posted: February 05, 2008 at 03:06 AM (#2683727)
Chris Booker -> DC
Joe Borchard -> ATL
I think Matt Craig is still a CUB, right?
Abe Nunez (OF) -> DC
Luis Gonzalez -> FLA
Chris Aguila -> PIT
Dave Krynzel is still a Brewer, I believe.
   25. Chris Needham Posted: February 05, 2008 at 03:11 AM (#2683729)
When Ed Rogers and Pete Orr are involved, nobody is having fun.

(I figured it was just the soft spot in your heart for your favorite Orioles SS of the future)
   26. Der Komminsk-sar Posted: February 05, 2008 at 03:15 AM (#2683730)
Matt Miller (P) -> BOS
Jeff Manship -> MIN
Josh Towers -> COL
Elizardo Ramirez -> TEX
Stephen Andrade -> back to TB
Mike Burns -> PIT
Mike Gallo -> TOR
Armando Galarraga (still TEX, supposedly about to be DET for Michael Hernandez)
I think Blake Johnson is still with KC.
Roger Deago -> TB
I think (but don't know) that Scott Tyler is still in the Marlins' org.
   27. Dan Szymborski Posted: February 05, 2008 at 03:19 AM (#2683731)
I should note that some of those aren't my fault as BA hadn't updated to the February 1st week of transactions at the time (though I did get some of them).
   28. Dan Szymborski Posted: February 05, 2008 at 03:22 AM (#2683734)
I did have Lizard and Matt Craig is still a cub (don't know why I didn't put him back).

Not sure about Scott Tyler - I know he was a minor league free agent and I can't find any reference to him signing anywhere.
   29. Dan Szymborski Posted: February 05, 2008 at 03:23 AM (#2683735)
Now, I'm probably not going to release a disk with these particular updates until I get SG's disk with the profiles (so I don't have to do the changes all over again unless he also sees this and does them).
   30. Der Komminsk-sar Posted: February 05, 2008 at 03:26 AM (#2683740)
Zach Hammes -> SD
Seanez is reportedly about to rejoin LAD.
Isn't Goleski still with CLE?
Kevin Hooper has been released.
Gregorio Petit's name is reversed.
   31. Der Komminsk-sar Posted: February 05, 2008 at 03:29 AM (#2683741)
I should note that some of those aren't my fault as BA hadn't updated to the February 1st week of transactions at the time (though I did get some of them).
Oh - I didn't mean anything at all derogatory (or expecting changes anytime soon), just trying to help...
   32. Dan Szymborski Posted: February 05, 2008 at 03:39 AM (#2683745)
Oh - I didn't mean anything at all derogatory (or expecting changes anytime soon), just trying to help...

I didn't mean you said anything at all derogatory! Just wanted an excuse just in case someone thinks I'm excessively sloppy! Minor league transactions are a horrific pain in the ass.
   33. galaxieboi Posted: February 05, 2008 at 03:45 AM (#2683749)
Just got home and downloaded it. Sweet t*ts.
   34. Der Komminsk-sar Posted: February 05, 2008 at 04:37 AM (#2683780)
Minor league transactions are a horrific pain in the ass.
Damn straight. Even BA misses 'em occasionally. I wish milb.com would keep team's offseason's rosters/transactions up-to-date, especially pre-Rule 5 time (yeah, I care about minor league R5 eligibles, what about it?).

Oh, just to be clear - I didn't think you thought I was being negative, I was just making sure we're cool. Awkwardly. :)

A few requests (thanks for Scott Patterson, by the way): Roy Corcoran (SEA), Diory Hernandez (ATL), John Raynor (FLA), and Danny Ray Herrera (TEX).
   35. HowardMegdal Posted: February 05, 2008 at 05:25 AM (#2683806)
May higher powers bless you and provide you coats of various colors, Dan.
   36. Masticore317 Posted: February 05, 2008 at 12:44 PM (#2683864)
I didn't mean you said anything at all derogatory! Just wanted an excuse just in case someone thinks I'm excessively sloppy! Minor league transactions are a horrific pain in the ass.


There really should be some better way of tracking minor league transactions.
   37. CoastalFan Posted: February 05, 2008 at 04:45 PM (#2684004)
this is SOOOO better than Christmas....holy CRAP Hampton is winning games for Atlanta in my sim runs - 17 - 4 last time through! And Pedro + Santana + Maine = STRESS PUPPIES for NL East hitters...dang...
   38. Dan Szymborski Posted: February 05, 2008 at 04:48 PM (#2684009)
this is SOOOO better than Christmas....holy CRAP Hampton is winning games for Atlanta in my sim runs - 17 - 4 last time through! And Pedro + Santana + Maine = STRESS PUPPIES for NL East hitters...dang...

Hey! He's only supposed to get 3 starts!
   39. SimaTime Posted: February 05, 2008 at 05:15 PM (#2684026)
Dan, it looks like Myers is still projected as a reliever/closr. Would it be possible to have him changed to a starter? Thanks!
   40. OBSLCommish Posted: February 06, 2008 at 10:37 AM (#2684596)
Thanks for all the hard work, Dan! Our sim league definitely appreciates it. Any chance we could get some projections for OF Josh Flores (Houston)? Mucho appreciado, amigo.
   41. Comic Strip Person Posted: February 06, 2008 at 05:50 PM (#2684786)
So, is there any way that SG could put up a link to his macros again this year? I've actually got Excel this year, and I want to try running and tabulating a bunch of seasons. If there is such a way: Thanks, SG.

And, thanks, Dan. This is really cool, and I enjoy it every year.
   42. galaxieboi Posted: February 06, 2008 at 09:06 PM (#2684969)
Hey, I got a question. How do you set up DMB to run a crap load of seasons? Is there a way to do it without manually restarting each time? Thanks a lot everyone...
   43. SG Posted: February 06, 2008 at 11:38 PM (#2685175)
TNGW/galaxiboi, here's a link to my tools to run Diamond Mind for 100 iterations.

http://www.replacementlevel.com/stuff/sg/RunDMB.zip
   44. Comic Strip Person Posted: February 07, 2008 at 12:19 AM (#2685232)
SG, let me just reiterate: you're great. Thanks. I bug you for this every year, and I've failed to make it work in Open Office. This year, I actually have Excel, but could never imagine how to come up with what you've done.
   45. Masticore317 Posted: February 07, 2008 at 01:54 AM (#2685317)
All of your hard work is very, very much appreciated!
   46. galaxieboi Posted: February 07, 2008 at 02:00 AM (#2685321)
Thanks a lot, SG! If you or Dan are ever in town (Seattle) I'd like to buy ya a beer.
   47. Pat Rapper's Delight Posted: February 07, 2008 at 02:11 AM (#2685327)
Over on the DMB forum, one of the guys (Diesel) uses a scripting tool called AutoIt to manage running several iterations of a season and saving the stats. I adapted it for stuff I run and it works pretty nicely. Don't know if that's what SG in ATL uses or not.
   48. HowardMegdal Posted: February 07, 2008 at 02:19 AM (#2685331)
If my sample of one season is to be believed, the Mets will open by sweeping the Marlins. This, of course, assumes that I am the manager.
   49. galaxieboi Posted: February 07, 2008 at 02:30 AM (#2685342)
The first several seasons I ran had the Mets killing in the NL. I was running some different Mariner lineups and never touched any of the other teams how they came from Dan.
   50. cardsfanboy Posted: February 07, 2008 at 02:48 AM (#2685353)
where is Juan Gonzalez? :)
   51. Jeff K. Posted: February 07, 2008 at 07:31 PM (#2685795)
Hey Dan. The file doesn't really display right in Office 08 beta (for Mac, of course.) Just so you know. It's sized all wrong when you open it, and zooming in makes the cells not big enough.
   52. Dan Szymborski Posted: February 07, 2008 at 07:36 PM (#2685802)
Jeff, you can't just do select all and change the font size or cell size? I really have no idea how to diagnose this one since I don't have a Mac! All I know is that it looks fine for me through Excel 07. Anyone else here using Office 08 for Mac?
   53. Jeff K. Posted: February 07, 2008 at 08:05 PM (#2685834)
Jeff, you can't just do select all and change the font size or cell size?

You can, but it's a PITA and still doesn't look right.

Anyone else here using Office 08 for Mac?

Highly doubtful. It's in beta. The only reason I have a copy is that they pre-released beta to University tech staffs. I'm just saying it now because it's likely to be the same thing once 08 hits release, which is in March.
   54. SG Posted: February 07, 2008 at 08:35 PM (#2685861)
First set of depth charts are done, and can be downloaded here:

http://replacementlevel.com/stuff/sg/zips2008.zip

If anyone sees anything on their favorite team that doesn't look right please let me know so I can make these more realistic.
   55. FMReader Posted: February 07, 2008 at 09:51 PM (#2685960)
I've run the 2008 season with DMB and the ZiPS data 100 times. Here are the results:

team div wc p/o wins
New York Yankees 79.0% 12.0% 91.0% 98.6
New York Mets 78.5% 10.8% 89.3% 98.2
Atlanta Braves 17.5% 46.8% 64.3% 91.5
Chicago Cubs 67.8% 4.3% 72.2% 90.7
Boston Red Sox 16.5% 28.5% 45.0% 89.5
Los Angeles Angels 62.3% 3.5% 65.8% 88.9
Detroit Tigers 43.8% 10.0% 53.8% 88.0
San Diego Padres 47.5% 2.1% 49.6% 87.8
Cleveland Indians 42.3% 8.5% 50.8% 87.5
Oakland Athletics 34.8% 3.5% 38.3% 85.7
Toronto Blue Jays 4.5% 22.0% 26.5% 85.5
Arizona Diamondbacks 26.5% 8.6% 35.1% 84.8
Philadelphia Phillies 4.0% 12.6% 16.6% 84.1
Milwaukee Brewers 22.5% 6.0% 28.5% 83.9
Los Angeles Dodgers 19.5% 4.0% 23.5% 83.5
Minnesota Twins 11.3% 5.0% 16.3% 81.7
Tampa Bay Rays 0.0% 5.0% 5.0% 79.4
Colorado Rockies 4.5% 2.0% 6.5% 76.8
St. Louis Cardinals 7.3% 0.0% 7.3% 76.3
Kansas City Royals 2.5% 1.0% 3.5% 75.9
Washington Nationals 0.0% 1.5% 1.5% 74.8
San Francisco Giants 2.0% 0.0% 2.0% 74.6
Cincinnati Reds 1.8% 0.3% 2.2% 73.0
Seattle Mariners 1.3% 0.0% 1.3% 71.9
Houston Astros 0.5% 1.0% 1.5% 71.7
Texas Rangers 1.5% 0.0% 1.5% 71.2
Baltimore Orioles 0.0% 1.0% 1.0% 70.6
Chicago White Sox 0.0% 0.0% 0.0% 70.0
Pittsburgh Pirates 0.0% 0.0% 0.0% 68.0
Florida Marlins 0.0% 0.0% 0.0% 66.1
   56. RobertMachemer Posted: February 07, 2008 at 11:58 PM (#2686075)
Is the Red Sox's average of 89.5 wins a function of Schilling's (possibly) being out? Or was that not yet included? Replacement Level Yankees got quite different results from its 100 iterations.
   57. RobertMachemer Posted: February 08, 2008 at 12:00 AM (#2686080)
Ah, I see I can't edit my post. ####.

Forget that post. Different question: is ZIPS really that much more pessimistic on the Red Sox than CAIRO? I was under the impression that ZIPS was pretty positive about the 2008 Red Sox.
   58. SG Posted: February 08, 2008 at 12:08 AM (#2686088)
I got similar results to FMReader with my ZiPS runs using the same depth charts that I used for CAIRO, although I had the Red Sox averaging 92 wins not 89.5. By my calculations, losing Schilling probably only costs the Red Sox a win or so.

I haven't compared the two projections all that closely yet, although the few I eyeballed were pretty close.

Sometimes the projections on a spreadsheet don't quite play out that way in Diamond Mind for whatever reasons.
   59. tyler Posted: February 08, 2008 at 01:37 AM (#2686138)
I'm pretty surprised that the A's are at 86 wins. Also a little surprised at how bad the White Sox are projected to be.
   60. FMReader Posted: February 08, 2008 at 03:42 AM (#2686218)
I fed the ZiPS projection disk into into DMB without any changes. You can fuss with manager profiles in the game forever. I suspect that CAIRO made some educated guesses in this area. In some instances, I agree more with CAIRO's results. I would have thought the Red Sox would come out ahead of the Yankees. For instance, the projection disk has Chamberlain, Hughes, and Kennedy all in the starting rotation and not Mussina. Realistically, Girardi will probably start Mussina and put Chamberlain in the bullpen. As an A's fan, I hate to say that I suspect the CAIRO number will prove more accurate.
   61. hub Posted: February 08, 2008 at 12:21 PM (#2686362)
I am having trouble getting the file to work. I followed the directions yet when trying to 'restore', DMB gives the following error message: "Unable to build database indexes". Anyone have a workaround? Thanks.
   62. Dan Szymborski Posted: February 08, 2008 at 03:13 PM (#2686403)
I am having trouble getting the file to work. I followed the directions yet when trying to 'restore', DMB gives the following error message: "Unable to build database indexes". Anyone have a workaround? Thanks.

Are you using DMB 9 with the latest patch? I have no idea what the problem could be otherwise - I've never had this happens when using restore.
   63. Bad Doctor Posted: February 08, 2008 at 04:45 PM (#2686466)
The difference for the Braves is pretty striking too, especially given that the rap on ZiPS, as I recall (sorry, Dan), has been it foresees a few too many breakouts in kids and a few too many dropoffs for old guys. If you told me the Braves came out 10 wins higher in one of the two forecasts, I would have assumed it would have been CAIRO.
   64. Dan Szymborski Posted: February 08, 2008 at 04:51 PM (#2686469)
The difference for the Braves is pretty striking too, especially given that the rap on ZiPS, as I recall (sorry, Dan), has been it foresees a few too many breakouts in kids and a few too many dropoffs for old guys.

Heh, half the people think I'm way too pessimistic on young players and half the people think I'm way too optimistic on young players.
   65. hub Posted: February 08, 2008 at 08:50 PM (#2686661)
Are you using DMB 9 with the latest patch? I have no idea what the problem could be otherwise - I've never had this happens when using restore.


I am using DMB 9c.
   66. hub Posted: February 08, 2008 at 08:54 PM (#2686665)
Nevermind, Dan. It was a 'vista' problem that I resolved. Thanks again for your projections and hard work.
   67. Dan Evensen Posted: February 09, 2008 at 05:18 PM (#2687132)
Thanks Dan! J00 Iz UB3r-L33t!!!!!11one
   68. Dan Szymborski Posted: February 09, 2008 at 10:50 PM (#2687329)
Nevermind, Dan. It was a 'vista' problem that I resolved. Thanks again for your projections and hard work.

Oh, if it was a Vista problem, I could've helped. Vista problems are pretty easy to fix - all you have to do is insert your XP disc and follow the directions on-screen!
   69. Ray (RDP) Posted: February 09, 2008 at 11:48 PM (#2687343)
Random question: How is DMB's internet play? Does it work pretty well? And does it enable one to set up a league with some real managers and some computer managers, and play out a season over the internet?
   70. Keith Law Posted: February 09, 2008 at 11:54 PM (#2687346)
Oakland Athletics 34.8% 3.5% 38.3% 85.7
Seattle Mariners 1.3% 0.0% 1.3% 71.9


If true, there will be a GM opening in Seattle by the All-Star Break.
   71. harrball Posted: February 10, 2008 at 02:37 AM (#2687428)
all you have to do is insert your XP disc and follow the directions on-screen

Dan, you're right, and it is always a great idea to upgrade a PC. You get better performance, wider software support, and more devices work with it! That's why more Vista users are upgrading to XP every day.

If true, there will be a GM opening in Seattle by the All-Star Break.

Run it before the trade, after the trade - it makes no difference, the Mariners and the Orioles both suck.
Baltimore Orioles 0.0% 1.0% 1.0% 70.6
   72. Masticore317 Posted: February 11, 2008 at 01:50 PM (#2688011)
Thanks for the update, Dan! FYI, David Freese is a Cardinal now (Edmonds trade).
   73. rollers333 Posted: February 12, 2008 at 04:51 AM (#2688568)
Thanks for all the hard work Dan!

Anyone know if there is a quick and easy way to combine this spreadsheet and the PECOTA spreadsheet to compare projections and remove players that don't have projections for both systems? Any help would be appreciated...
   74. dr. bleachers Posted: February 12, 2008 at 04:58 AM (#2688573)
Anyone know if there is a quick and easy way to combine this spreadsheet and the PECOTA spreadsheet to compare projections and remove players that don't have projections for both systems? Any help would be appreciated...

I'll second this. Also, I've got the macro running fine, but I'm having trouble figuring out where to put the (sorry, from memory at the moment - I'm on a different computer) column_g and column_p files go for the last step. I've tried pretty much everything. I'm assuming this is how you get the &#xpl;ayoffs numbers to fill in?

And all this work is truly awesome, Dan. SG, you too. Very cool.
   75. dr. bleachers Posted: February 12, 2008 at 05:00 AM (#2688574)
Ugh, I wish I could edit here. That should be "% playoffs". Also, hopefully this sort of thing isn't annoying. I'm new to Diamond Mind, and some of this isn't quite as intuitive for me as I'd hoped, for whatever reason.
   76. PS is probably going to lose his mind Posted: February 12, 2008 at 06:43 AM (#2688624)
I was looking for some way to combine ZiPS and PECOTAs to form some composite rating. I was going to add some CHONEs, but their spreadsheet was down.

I haven't made it yet, but there's the issues of distributing PECOTAs... I'd either need proof of ownership, and distribute such a document by hand, or build it with the PECOTAs deleted, allowing people to paste things in later...

I'll see what I have to work with, anyway.
   77. Dr. Vaux Posted: February 12, 2008 at 07:33 AM (#2688640)
If you run sims using the composite rating and publish the results, BP wouldn't have a legitimate legal quarrel with you; you wouldn't have made the PECOTAs available in a form that was decipherable.
   78. hub Posted: February 12, 2008 at 11:09 AM (#2688659)
Yuniesky Betancourt is labelled as hitting 'L' in the ZiPS database. He bats 'R'.
   79. Masticore317 Posted: February 12, 2008 at 12:33 PM (#2688662)
Also, I believe Nook Logan stopped switch hitting last year - he's now solely a right handed batter.
   80. rollers333 Posted: February 12, 2008 at 01:12 PM (#2688663)
I was looking for some way to combine ZiPS and PECOTAs to form some composite rating. I was going to add some CHONEs, but their spreadsheet was down.

I haven't made it yet, but there's the issues of distributing PECOTAs... I'd either need proof of ownership, and distribute such a document by hand, or build it with the PECOTAs deleted, allowing people to paste things in later...

I'll see what I have to work with, anyway.



This is exactly what I'm also looking to do, but facing the challenge of how to delete the players that don't have projections for all 3 systems...
   81. SG Posted: February 12, 2008 at 04:06 PM (#2688779)
Also, I've got the macro running fine, but I'm having trouble figuring out where to put the (sorry, from memory at the moment - I'm on a different computer) column_g and column_p files go for the last step. I've tried pretty much everything. I'm assuming this is how you get the pl;ayoffs numbers to fill in?


The columnG-ColumnP files should go in the same folder as the spreadsheet and the standings reports.
   82. PS is probably going to lose his mind Posted: February 12, 2008 at 05:06 PM (#2688832)
#77, not having DMB, I'd be working on the actual spreadsheet that compiles such a report. I'd run such a thing by having sheets containing the projections, and a series of lookups, weighting and them appropriately. Obviously, some names need to be adjusted.

However, if people know I'm weighting PECOTA:ZiPS:CHONE as say, 4:3:2, 'project' a guy as hitting .280, and you know his CHONE is .275 and his ZiPS .283, you can pretty much reverse engineer the PECOTAs back from them, which is what I want to avoid.

(Also, CHONEs seem to be back up, which wasn't the case yesterday. Go figure.)

The only way to get around that is release a spreadsheet that has the PECOTA sheet blank, but all the formulas live. So, people who have access to PECOTAs can paste them in, and those who don't can 'settle' on ZiPS:CHONE 3:2.

Also, I can count 'this player exists' as a stat. That would allow you to filter out those who don't appear in all lists.
   83. Ennder Posted: February 12, 2008 at 09:34 PM (#2689161)
I have PECOTA and would love a sheet like that so there is definitely an audience for your work!
   84. PS is probably going to lose his mind Posted: February 12, 2008 at 11:16 PM (#2689309)
In retrospect, I'm seeing a few things that may cause issues.

Not all databases show all the same stats. Some are moot, such as saves, but some, such as BFP, HBP and others aren't available in all databases. I can probably do some tricks to fill in missing stats from existing data, but we'll have to see what's missing after.

Not all database are up to date, so teams can be off. I can take a Marcel approach and say those are rare enough that regression to the mean covers the difference. Making adjustments on the fly is usually beyond what a spreadsheet should be able to do without macros. (And I use OOo, so they're probably not compatible.)
   85. Luke Gofannon Posted: February 13, 2008 at 07:40 PM (#2690143)
I'm among those who worry that DMB, with its new owner, is headed in the direction of shutting down its production of new season disks and projection disks for use in its stand-alone version. Since the game engine itself is already so good and the community of players so knowledgeable, I'm hoping there will be good "homebrews" to take the place of company-produced disks if Dayne Myers shuts down production of further seasons for DMB's computer version. If such a shutdown were to occur, do you think it's likely you will still do these yearly ZiPS files to be used with the DMB game engine?
   86. Dan Szymborski Posted: February 13, 2008 at 08:06 PM (#2690210)
If such a shutdown were to occur, do you think it's likely you will still do these yearly ZiPS files to be used with the DMB game engine?

I think that would make it more likely I do the projection disk as long as people are still using their DMB.
   87. Luke Gofannon Posted: February 13, 2008 at 08:16 PM (#2690237)
"I think that would make it more likely I do the projection disk as long as people are still using their DMB."

Wow, that is good to hear. Thanks for taking the time to answer, and, as always, good stuff.
   88. Boriole Forester Posted: February 13, 2008 at 08:35 PM (#2690287)
Baltimore Orioles 0.0% 1.0% 1.0% 70.6
Chicago White Sox 0.0% 0.0% 0.0% 70.0
Pittsburgh Pirates 0.0% 0.0% 0.0% 68.0
Florida Marlins 0.0% 0.0% 0.0% 66.1


WTF do we have to do to get the #1 pick, trade Brian Roberts?

-Orioles Fans
   89. Boriole Forester Posted: February 13, 2008 at 08:36 PM (#2690288)
P.S. Thanks, Dan.
   90. resultsDisoriented Posted: February 13, 2008 at 10:45 PM (#2690487)
66: a little help with how you fixed the vista error? i'm getting it here too.
   91. Cheer and boo and raise a hullabaloo Posted: February 15, 2008 at 01:57 AM (#2691560)
So, I popped it into DMB and ran one season. The San Francisco Giants won the West with a 90-72 record. How rare of an occurrence is that? Replacement Level has the Giants making the playoffs 0 out of 100 times. They did it apparently by having a very good pitching staff, not-the-worst offense, and luck (San Diego's offense and pitching were both better).
   92. Jeff K. Posted: February 15, 2008 at 02:23 AM (#2691575)
I'm also looking to combine ZIPS, CHONE, Marcel, and PECOTA. I don't subscribe to BP anymore, so PECOTA is going to be tough.
   93. Jeff K. Posted: February 15, 2008 at 02:25 AM (#2691579)
Speaking of, is there a link for CHONE 2008? I only see 2007.
   94. Jeff K. Posted: February 15, 2008 at 02:26 AM (#2691582)
Oh, never mind, I'm an idiot.
   95. Jeff K. Posted: February 15, 2008 at 03:12 AM (#2691615)
If anyone cares, I'm in the middle of putting together ZIPS, CHONE, and Marcel into one Excel workbook, with an average combining all three. I'll leave worksheets for PECOTA for those who have it. Unless and until I do so, I'll leave the combination at the three free systems.
   96. Jeff K. Posted: February 15, 2008 at 03:34 AM (#2691633)
By the way Dan, for as much #### as I give you, you know I appreciate the hell out of what you do. And tell the milk measurer I said hi.
   97. Jeff K. Posted: February 18, 2008 at 05:46 AM (#2693441)
Okay, so here's a workbook with ZiPS, Marcel, CHONE on separate worksheets:

Link

That's my work, and it's only the three systems, with everything lined up as the same columns.

And here, thanks to the work of Dan Werr (and Chris Dial), is a two-page worksheet with all the projections on equal rows. Only players projected in two or more systems are on it, because if you aren't making CHONE (2000+ hitter projections!) plus one of Marcel and ZiPS, then I don't care:

Link 2.

Use the first if you want to set up your own spreadsheet/combination. Use the second if you want everything on one. I might post my sheet with the weighted averages, but that will mostly for my own league and not much else.

I think Dial may be getting the worksheet put into database form, so I'll keep you posted on that.
   98. Jeff K. Posted: February 18, 2008 at 05:49 AM (#2693446)
"Two-page worksheet" should read "two-worksheet workbook".

And it's in Office 2003 format.
   99. Dan Szymborski Posted: February 18, 2008 at 06:14 AM (#2693460)
By the way, just for those that think ZiPS is a super-pessimistic projection system, CHONE has 174 players with 500 at-bats or more and has that group at 271/339/432 while ZiPS has that exact same group as 272/340/431. CHONE has 92 pitchers with 162 innings or more and they combine for a 4.24 ERA while ZiPS combines them for 4.28.

Sorry, it's just I'm still getting more e-mails for "Oh my God, why is ZiPS the most pessimistic projection system in the history of the world. Don't you love America?" than every other query about ZiPS combined.
   100. Jeff K. Posted: February 18, 2008 at 07:41 AM (#2693509)
than every other query about ZiPS combined.

I posted a query in the ZiPS FAQ thread...
Page 1 of 2 pages  1 2 > 

You must be Registered and Logged In to post comments.

 

 

<< Back to main

BBTF Partner

Support BBTF

donate

Thanks to
Ray (RDP)
for his generous support.

Bookmarks

You must be logged in to view your Bookmarks.

Syndicate

Page rendered in 0.6876 seconds
66 querie(s) executed