|
||||
|
|
OOTP 19 - Historical Simulations Discuss historical simulations and their results in this forum. |
|
Thread Tools |
11-03-2018, 01:23 PM | #21 |
Hall Of Famer
Join Date: Jun 2004
Posts: 3,736
|
I just conducted another fielding test on the 1906 season. I calculated and edited the fielding ratings for every player based on the plus/minus data and formulas for translating these into ratings. For players with low sample size fielding statistics in 1906 (less than 50 A for infielders or less than 100 PO for OF) I then went to their 3yr values and if that was still too low I then used their career values. This way the player fielding ratings would not be subject to sample size problems.
Here was the result of the first test of this for the Cubs: Real Cubs Defensive Efficiency: .736 Simulated Cubs Defensive Efficiency: .737 Real League Defensive Efficiency: .695 Simulated League Defensive Efficiency: .695 I am testing this season because this particular Cubs team is the most extreme defensive team relative to the league since 1901. Their defensive efficiency was 41 points higher than league average. |
11-03-2018, 02:48 PM | #22 |
Hall Of Famer
Join Date: Apr 2009
Location: Toronto, ON
Posts: 6,123
|
You are doing amazing work Garlon, even though it is going largely unnoticed. Keep plugging along, and maybe eventually we'll see the fruits of your labour in the game.
__________________
My corrected FaceGen IDs .zip file here: https://drive.google.com/file/d/1oRd...usp=share_link OOTP post re-FG IDs here: https://forums.ootpdevelopments.com/...postcount=3198 My DB which restores Fed Leaguers here: https://drive.google.com/file/d/1ZoN...B2GCcULxt/view Instructions for the DB: https://forums.ootpdevelopments.com/...07&postcount=9 |
11-03-2018, 03:52 PM | #23 |
Hall Of Famer
Join Date: Jun 2004
Posts: 3,736
|
Thank you. I have the fielding ratings calculated in this same way for every season 1871-2017 in a file. I have a 1yr base, like the one above, that will use 3-yr or career if the player sample size is too low for the season. I have a 3yr base that uses career when the sample size is too low for the 3yr span. And I have a Career base as well.
|
11-03-2018, 04:35 PM | #24 | |
Hall Of Famer
Join Date: Apr 2009
Location: Toronto, ON
Posts: 6,123
|
Quote:
__________________
My corrected FaceGen IDs .zip file here: https://drive.google.com/file/d/1oRd...usp=share_link OOTP post re-FG IDs here: https://forums.ootpdevelopments.com/...postcount=3198 My DB which restores Fed Leaguers here: https://drive.google.com/file/d/1ZoN...B2GCcULxt/view Instructions for the DB: https://forums.ootpdevelopments.com/...07&postcount=9 |
|
11-03-2018, 10:11 PM | #25 |
OOTP Historical Czar
Join Date: Dec 2001
Location: Bothell Wa
Posts: 7,254
|
OOTP has been using the same bill james plus minus system garlon is proposing since OOTP 14 or 15. I don't remember exactly since it was so long ago. every year it has gotten more sophisticated as newer research has come to light. Last year total zone/total fielding runs above average was added. The goal for OOTP20 would be to add something new in addition to what is already in place.
This proposal is just a rehash of really old stuff and has been in place for years. Sorry to burst your bubble. The implementation can be looked at for enhancements but the data is already in the game.
__________________
It's madness, madness, I tell you! For the love of God, don't do it! |
11-03-2018, 10:37 PM | #26 |
Hall Of Famer
Join Date: Jun 2004
Posts: 3,736
|
The data is not in the game properly. The ratios in your fielding file are way off. The formulas that OOTP uses to translate such ratios and weight them into ratings in the game is flawed as well.
In the file you have it has Andruw Jones 1998 RF1 as 1.78. Sorry, but no outfielder is making 78% more plays than anyone else. He made 413 PO that season. 1.78 means he made 186 PO above average. I have him at 1.12, and making 43 PO above average. Your file has Joe Tinker 1908 at SS with RF1 as 1.45. This is not possible as well. He made 570 assists as SS that year and according to this formula the average SS playing his innings should have made 393 assists, so he apparently made 177 assist above average...sorry, but that's just not possible. I have him at 1.14 that season, an astounding 68 assists above average. For Joe Tinker in 1906 you have his Error rating as 1.54. He fielded .944 at SS while the league fielded .933. So per 1000 total chances he made 56 errors while the league average was 67. 67/56 = 1.196 This can be corrected and we can add my calculations to the DB and keep these if you want, but there is something seriously not right about how you have calculated them. I know you said that you didn't use team Defensive Efficiency, which I did. And I think you are using some other methodology for adjusting for groundball tendency of teams. Last edited by Garlon; 11-03-2018 at 11:55 PM. |
11-04-2018, 12:50 AM | #27 |
OOTP Historical Czar
Join Date: Dec 2001
Location: Bothell Wa
Posts: 7,254
|
you are completely right in so far as ratings do not work the way you think they do.
there is no one to one relationship between ratings and assists for instance. sorry but it is just not currently so. you are beginning with assumptions you wish were true but are not. we have discussed this privately for months and I have even published the current formulas in use for you but I guess you just can't process the information right now. maybe later?
__________________
It's madness, madness, I tell you! For the love of God, don't do it! |
11-04-2018, 01:48 AM | #28 |
Hall Of Famer
Join Date: Jun 2004
Posts: 3,736
|
I never said there is a one to one ratio between ootp range ratings and assists. For every 5 points that you raise Range ratings you will get about a 1% increase in Defensive Efficiency. So if you set outfield range to an average of 120 on the 250 scale for the league and a player has a range of 145 OF Range, you should expect that player to end up with about 5% more PO than average as every 5 points of OF Range above/below average will yield about a 1% increase in performance.
For errors there is a proportional relationship in the game. If you set the league average to 120 on the rating scale a player with a 240 will make half as many errors and a player with a 60 will make twice as many errors. OF Arm is proportional as well. As the game is right now you will not get realistic fielding results on a per player basis. The LTM's that control the rest of the statistics somewhat hide what is happening with fielding, but there are problems because of how the ratings are translated. I can tell you that the game right now cannot properly rate players at multiple infield or outfield positions. |
11-04-2018, 12:06 PM | #29 |
OOTP Historical Czar
Join Date: Dec 2001
Location: Bothell Wa
Posts: 7,254
|
My mistake, this diatribe is about the player editor apparently. I thought you were having issues with fielding ratings in a base file.
I have no idea about how the player editor works and what happens when you muck about in there. I'd suggest just don't but I am sure that won't stick. BTW all the base file ratings are per position not lumped together but I am sure you already know that. But I guess you are not referring to that anyway. If possible please be clear as to specifically what you are referring to as all the switching of references are sort of hard to follow.
__________________
It's madness, madness, I tell you! For the love of God, don't do it! |
Bookmarks |
Thread Tools | |
|
|