 |
View Old Forum Thread
Old Forum Index » Baseball Beta Testing » Beta News » New Gold Glove Formula
Admin
This is mainly just an FYI in case you wonder why the vote totals look different.
Tyson
New Gold Glove Formula
November 03, 2009 at 05:11PM View BBCode
I've turned on a new gold glove formula for Beta that the BLL wishes to use. Results should start showing tomorrow. Khazim is going to pop in and test it.This is mainly just an FYI in case you wonder why the vote totals look different.
Tyson
khazim
November 03, 2009 at 05:32PM View BBCode
I have an excel spreadsheet set up using the fielding stat export so I can compare the two. Were there any changes you needed to make to the presented formula to work within the Sim D universe?
Admin
Tyson
November 03, 2009 at 05:35PM View BBCode
No, I was able to use the formula as supplied by Geoff:Basically I set a soft cap at 1,098 innings played (roughly 75% of the season). Beyond that they player drops from 1 inning = 1 adjusted inning to 10 innings = 1 adjusted inning. This should prevent A+ health players from running away with the award.
Adjusted innings = IF (total innings >=1098,1098+(total innings played - 1098)*.1), and IF (total innings<= 729,(total innings+((729-total innings)*.9), else Adjusted innings = Total innings
(Adjusted innings*((FPCT*1000)+(plus/minus*2)+CS-SB+DP-PB))/(122 games * 9 innings)
Comments or questions are appreciated.
This formula all worked for all positions, even catchers and pitchers.
Tyson
tworoosters
November 04, 2009 at 05:09PM View BBCode
Wow some dramatic differences, under Khazim's formula [url=http://beta.simdynasty.com/player.jsp?id=44675]Madrid[/url] and [url=http://beta.simdynasty.com/player.jsp?id=39675]Myette[/url] were both GG outfielders whereas under Dwight's (default) system they finish as essentially negative fielders (below 1,000 points).
khazim
November 04, 2009 at 05:52PM View BBCode
Well I need to dig into pitching because a pitcher with 1 TC shouldn't be winning a Gold Glove.
khazim
November 04, 2009 at 05:58PM View BBCode
In fact, I'll go one further. A player on the waiver wire shouldn't win the Gold Glove period.
khazim
IP = 9.33333
TC = 1
PO = 0
A = 1
DP = 0
+/- = 0
FPCT = 1.000
this should result in a GG number of 598.39
I think I see where the issue might be.
Did I include the adjusted innings for being under 50% of a season's innings played?
What I used to calculate adjusted innings in Excel is:
=IF(Innings played >=1098,1098+((Innings played-1098)*0.1),IF(Innings played <=729,(Innings played+((729-Innings played)*0.9)), Innings played))
Using this, if you play zero innings, your Gold Glove number starts at 656.1
November 04, 2009 at 06:16PM View BBCode
and according to my spreadsheet, his overall GG number is wrong.IP = 9.33333
TC = 1
PO = 0
A = 1
DP = 0
+/- = 0
FPCT = 1.000
this should result in a GG number of 598.39
I think I see where the issue might be.
Did I include the adjusted innings for being under 50% of a season's innings played?
What I used to calculate adjusted innings in Excel is:
=IF(Innings played >=1098,1098+((Innings played-1098)*0.1),IF(Innings played <=729,(Innings played+((729-Innings played)*0.9)), Innings played))
Using this, if you play zero innings, your Gold Glove number starts at 656.1
tworoosters
November 04, 2009 at 06:23PM View BBCode
Which formula is the Gold Glove page currently showing ?
khazim
November 04, 2009 at 06:24PM View BBCode
It doesn't look like it is, let me dig as little deeper. I'm at work so it may take an hour or so for me to pull the numbers for my own spreadsheet (designed them prior to fielding stats being able to be directly exported.)
hobos
Sorry to critique the formula, but it may be worth noting that this will essentially reward whoever has the most innings played. Note that +1 on +/- will add 2/(9*122) ~ 0.0018, whereas 10 additional innings will add 0.1(1.000*1000)/(9*122) ~ 0.091.
For example, AL 3B:
Crabtree 1374.3 IP, .962, 16 DP, -7 +/-
Cassidy 1035 IP, .957, 19 DP, +7 +/-
Scores:
Crabtree 986
Cassidy 902
Scores* (adjusted IP *(.960*1000)/(1098)):
Crabtree 992
Cassidy 904
Their actual FLD%, DP, and +/- had almost no effect.
Thus A+ health players will run away with the award. For AL 3Bs, who I examined, the top 6 players had decreasing health save for Andrews, who had a bad FLD% and wasn't far out of line (and the 7th, Cassidy, only played vs righties).
November 04, 2009 at 08:57PM View BBCode
The numbers, at least for non-pitchers, appear fine.Sorry to critique the formula, but it may be worth noting that this will essentially reward whoever has the most innings played. Note that +1 on +/- will add 2/(9*122) ~ 0.0018, whereas 10 additional innings will add 0.1(1.000*1000)/(9*122) ~ 0.091.
For example, AL 3B:
Crabtree 1374.3 IP, .962, 16 DP, -7 +/-
Cassidy 1035 IP, .957, 19 DP, +7 +/-
Scores:
Crabtree 986
Cassidy 902
Scores* (adjusted IP *(.960*1000)/(1098)):
Crabtree 992
Cassidy 904
Their actual FLD%, DP, and +/- had almost no effect.
Thus A+ health players will run away with the award. For AL 3Bs, who I examined, the top 6 players had decreasing health save for Andrews, who had a bad FLD% and wasn't far out of line (and the 7th, Cassidy, only played vs righties).
khazim
November 04, 2009 at 11:34PM View BBCode
None of the numbers matched up with what I had in my spreadsheet. For example, Rick Sevcik, with 1414 innings played, came up with a score of 1212.93 with an adjusted innings of 1129.6. The site lists his end of year points at 1,026. I'm not sure where the difference lies. I'll poke around with this some more. I'm sure I missed something vital.
khazim
Crabtree: 1374.333 innings = (1098+((1374.333-1098)*0.1)) or (1098+((276.333)*0.1) or 1098 + 27.6333 or 1125.6333
Cassidy: 1035 innings, which adjusts to 1035. Instead of there being a 339 inning differential, my formula knocks it down to 90.
Instead of comparing a player who played 75% of the innings of the other, we have are comparing an adjusted inning total where one (Cassidy) has 92% of the innings played.
November 04, 2009 at 11:45PM View BBCode
also, hobo, thank you for the critique, but you picked two players straddling the 75% breakpoint. Their adjusted innings, based upon what you listed woudl be:Crabtree: 1374.333 innings = (1098+((1374.333-1098)*0.1)) or (1098+((276.333)*0.1) or 1098 + 27.6333 or 1125.6333
Cassidy: 1035 innings, which adjusts to 1035. Instead of there being a 339 inning differential, my formula knocks it down to 90.
Instead of comparing a player who played 75% of the innings of the other, we have are comparing an adjusted inning total where one (Cassidy) has 92% of the innings played.
hobos
The website is using
(Adjusted innings*((FPCT*1000)+(plus/minus*2)+CS-SB+DP-PB))/(122 games * 9 innings)
If I did the math right, you appear to be using
(Adjusted innings*((FPCT*1000)))/(122 games * 9 innings)+(plus/minus*2)+CS-SB+DP-PB
This is where my critique was centered. The website is dividing all of the extra stats by 1098, making them worthless, so the website currently only uses the adjusted innings total and FLD%, thus all players are currently ranked by their health, as I said.
November 05, 2009 at 02:30AM View BBCode
khazim, the error lies in the formula Tyson is using. Based on that formula, all the numbers on the site are correct.The website is using
(Adjusted innings*((FPCT*1000)+(plus/minus*2)+CS-SB+DP-PB))/(122 games * 9 innings)
If I did the math right, you appear to be using
(Adjusted innings*((FPCT*1000)))/(122 games * 9 innings)+(plus/minus*2)+CS-SB+DP-PB
This is where my critique was centered. The website is dividing all of the extra stats by 1098, making them worthless, so the website currently only uses the adjusted innings total and FLD%, thus all players are currently ranked by their health, as I said.
Admin
Tyson
November 05, 2009 at 05:14AM View BBCode
khaz, can you let me know what the formula should be? The one Geoff gave me appears to be incorrect.Tyson
khazim
[Edited on 11-5-2009 by khazim]
November 05, 2009 at 01:52PM View BBCode
The first formula is identical to the one in my original spreadsheet, Tyson. That should work fine. That's why I was thrown off by the end of season numbers not matching the ones from the spreadsheet, not to mention the guy with 9 innings played winning the Gold Glove. In 15 seasons of using this over in the Dennis Eckersly league, I never had a guy with that few innings played even show up on the radar of a Gold Glove award.[Edited on 11-5-2009 by khazim]
khazim
November 05, 2009 at 05:33PM View BBCode
whatever the issue was, it seems to be tracking correctly now.
khazim
I've got a quick check sheet built which lets me check the formula as often as I like in about 30 seconds or so.
November 05, 2009 at 07:04PM View BBCode
Are you still having the Gold Gloves process once per day?I've got a quick check sheet built which lets me check the formula as often as I like in about 30 seconds or so.
Admin
Tyson
November 05, 2009 at 07:17PM View BBCode
Yes, it is once a day. I can run it for you again if you give me notice though, it will be out of whack the next time a game runs.Tyson
khazim
November 05, 2009 at 07:31PM View BBCode
Is there any way you could rig it to run at 5PM ET? It would put me in a great spot to make the check and not miss it by a game or two.
Admin
Tyson
November 05, 2009 at 07:44PM View BBCode
It will run at 4:55PM Eastern. This will give you until 5:48PM Eastern before the next game runs.Tyson
khazim
NL
P: Dick Roser (601.942), Ray Cohen (601) - I have Cohen well down the list, but the numbers are close (601.213)
C: Susan Encarnacion (616.146), Susan Encarnacion (615)
1B: Terry Paredes (616.292), Terry Paredes (616)
2B: Ed Scott (618.711), Ed Scott (618)
3B: Joe Hickman (608.547), Joe Hickman (608)
SS: Reb Valle (606.864), Wiley Reich (606) - I have Reich (606.511)
LF: Jerry Castleman (615.692), Erin Shaw (615) - I have Shaw in 2nd (615.059)
CF: Glenn Boggs (617.886), Glenn Boggs (617)
RF: Chris Freeman (616.364), Chuck Judd (616) - I have Judd in 2nd (616.213)
This is just one league, and the numbers are darn close. The analyst in me screams that this looks like a rounding issue.
Tyson, do you know if your numbers are being rounded prior to the end result?
November 05, 2009 at 11:20PM View BBCode
okay, as of 5PM ET, 11/05/09 GG leaderboard comparison Mine on the left, the board's on the right:NL
P: Dick Roser (601.942), Ray Cohen (601) - I have Cohen well down the list, but the numbers are close (601.213)
C: Susan Encarnacion (616.146), Susan Encarnacion (615)
1B: Terry Paredes (616.292), Terry Paredes (616)
2B: Ed Scott (618.711), Ed Scott (618)
3B: Joe Hickman (608.547), Joe Hickman (608)
SS: Reb Valle (606.864), Wiley Reich (606) - I have Reich (606.511)
LF: Jerry Castleman (615.692), Erin Shaw (615) - I have Shaw in 2nd (615.059)
CF: Glenn Boggs (617.886), Glenn Boggs (617)
RF: Chris Freeman (616.364), Chuck Judd (616) - I have Judd in 2nd (616.213)
This is just one league, and the numbers are darn close. The analyst in me screams that this looks like a rounding issue.
Tyson, do you know if your numbers are being rounded prior to the end result?