|
||||
|
04-13-2010, 06:52 PM | #1 |
Hall Of Famer
Join Date: May 2006
Posts: 3,625
|
Pre-Defined Draft Values
We have a pretty serious problem with pre-defined draft values that needs to be addressed. I'm not sure who developed these values, but I've been told that they were supposed to be based on runs created, runs saved, and player longevity.
However, when sorting the Master.csv by draft values (column 34), there are quite obvious errors or otherwise inexplicable problems. This is causing the AI to make poor draft decisions in historical leagues. Here are some actual examples from the CSV file. I defy anyone to explain and justify why these players would be valued and ranked in this order. Minnie Minoso - Draft Value 9145 (25th among ALL historical players!) 17 yrs, 1,835 G, .298 AVG, 336 2B, 83 3B, 186 HR, 1,023 RBI, 814 BB Fred McGriff - Draft Value 8039 (66th overall) 19 yrs, 2,460 G, .284 AVG, 441 2B, 24 3B, 493 HR, 1,550 RBI, 1,305 BB Jimmy Zinn - Draft Value 7467 (101st overall) 5 yrs, 18 G, 105.1 IP, 5.04 ERA Earl Caldwell - Draft Value 7300 (118th overall) 8 yrs, 200 G, 587.2 IP, 4.69 ERA Steve Garvey - Draft Value 6704 (186th overall) 19 yrs, 2332 G, 440 2B, 43 3B, 272 HR, 1,308 RBI, 479 BB Lefty Gomez - Draft Value 3371 (1702 overall) 14 yrs, 368 G, 2503 IP, 3.34 ERA Stan Musial - Draft Value 3751 (1346th overall) 22 yrs, 3,026 G, .331 AVG, 725 2B, 177 3B, 475 HR, 1,951 RBI, 1,599 BB Sandy Koufax - Draft Value 1792 (4042nd overall) 12 yrs, 397 G, 2324.1 IP, 2.76 ERA The Master.csv is filled with draft values that do not make sense. I'm sorry, but I can't accept Chuck Tanner having a higher draft value and being taken in a draft ahead of Clete Boyer, Ted Abernathy, and Sandy Koufax! By the way, OOTP gives Koufax a 1-star potential in the draft pool anyway, despite having complete scouting turned off! Thanks to this and his senseless draft value, he went in the 4th round of a 1954 draft, right after Herb Moford, who pitched 50 games in his career for an ERA of 5.03! I'm hoping others will chime in and we can figure out a way to fix this. |
04-13-2010, 06:54 PM | #2 | |
All Star Reserve
Join Date: Sep 2008
Location: Equestria
Posts: 808
|
Quote:
__________________
OOTP Resident Brony |
|
04-15-2010, 06:33 PM | #3 |
Hall Of Famer
Join Date: May 2006
Posts: 3,625
|
Am I out of my mind, or does it seem like virtually no one is actually bothered by this glaring, critical problem? Or at least not bothered enough to post a response or do anything about it? Judging by the silence here and in my other attempts to raise this to the attention of the OOTP crew, this must be deemed insignificant. But it is an absolute show-stopping problem for historical play.
Last edited by Charlie Hough; 04-15-2010 at 06:43 PM. |
04-15-2010, 08:43 PM | #4 | |
All Star Reserve
Join Date: Sep 2008
Location: Equestria
Posts: 808
|
Quote:
I was just pointing out that Fred McGriff is the only one that possibly belongs up there. That being said, i am biased so... Either way, i always wanted to use this feature but never found rhyme or reason why. I'm sorry that it seemed i didn't care, because i do.
__________________
OOTP Resident Brony |
|
04-15-2010, 08:54 PM | #5 | |
Hall Of Famer
Join Date: Apr 2009
Location: Toronto, ON
Posts: 6,123
|
Quote:
You are correct sir. It is broken, and the only solution unfortunately (for now) is to turn it off. The crazy thing is that the other two things that were supposed to be exclusive to the historical database.odb file (historical pitches and positional experience) worked flawlessly for my games. Only the draft values were fubared when modding. I share your frustration. Garlon and Spritze created that part of the Master.csv/Historical Database.odb based on career Runs Created or some such stat and it has not worked. My suggestion is a gentle ping in their general direction, remembering that a) They are volunteers who do this as a hobby and have contributed a great deal to this game and b) Just about everything that's ever been introduced to the game has fallen short of what everyone around here has come to expect, but that eventually it makes its way up to the level of other features of the game that came before it. Yes, some other things are probably still broken in 11 (despite the as yet mostly empty technical support section), but the fact is there isn't another game with this much depth and options that allows you to play in career mode. My suggestion would be for you to purchase the WAR database (1871-2009) from this site and maybe write over the draft values with career WAR or RAR (forgive me Garlon and Spritze , it's just a temporary fix) in order to produce more realistic draft results, although to be fair IRL there ain't no such thing as a surefire prospect. You could also do what I do and turn off the draft values thingy and write over the names, nicknames, nationalities...anything that gives away a player's identity and use 20,000 (about 17,500 actually) or so fake names (generated and sent to your e-mail box for free at this site) in the firstname; lastname columns so you have no idea who is who and play with ratings turned off so you're completely blind (i.e. closer to the level of the AI). Understandably, that is not everyone's cup of tea, but it gets rid of the Monday morning quarterback draft syndrome, whereby you basically know how things will turn out for each player in the draft, while the AI is somewhat clueless. Unfortunately, it also gets rid of the ability to play historical games with your childhood heroes - just as in real life there are trade-offs, and it's up to you to decide what's acceptable for you. You can always change the database back to its original form once you've taken a game as far as you're going to take it to see how your favourite players did. While we're on the subject could somebody possibly fix things up so that historical gamers who play with modded databases can utilize this potentially great and equalizing feature if and when it is fixed? Thankyou from an intermediate computer user who wouldn't have the first clue how to do so. |
|
04-15-2010, 08:55 PM | #6 |
Hall Of Famer
Join Date: Dec 2002
Location: The OC
Posts: 6,347
|
I don't even know what this feature is or what it does, although certainly if that's the all-time ranking of those players most of them are way off.
That said, Steve Garvey's rank looks pretty darn close. He's probably a little overrated - I doubt Garvey is in the top 200 all-time, considering there are about 200 people in the Hall of Fame - but still pretty damn close. Garvey is definitely ranked higher than, say, 250.
__________________
Looking for an insomnia cure? Check out my dynasty thread, The Dawn of American Professional Base Ball, 1871. |
04-15-2010, 09:31 PM | #7 | |
Hall Of Famer
Join Date: Apr 2009
Location: Toronto, ON
Posts: 6,123
|
Quote:
|
|
04-15-2010, 09:35 PM | #8 |
Hall Of Famer
Join Date: Dec 2002
Location: The OC
Posts: 6,347
|
It's funny, because Garvey was one of the few players of the 1980s who was constantly referred to as a future Hall of Famer. Yeah, he was always overrated back in the day, but I just don't think I buy him as #500. Maybe top 250 is too high, I suppose.
Anyway, it's kind of beside the point. No worries.
__________________
Looking for an insomnia cure? Check out my dynasty thread, The Dawn of American Professional Base Ball, 1871. |
04-15-2010, 09:47 PM | #9 |
Hall Of Famer
Join Date: Apr 2009
Location: Toronto, ON
Posts: 6,123
|
It is supposed to eliminate the problem of very good historical players falling to the human player in the later rounds of the amateur draft, but as Charlie Hough has eloquently demonstrated it can't possibly do that given how it is set up now, unless it is supposed to be a reflection of the player's ability level based on how he played in his first 2 or 3 seasons. That's not the impression I got from discussions I saw about it. I thought it was supposed to be a reflection of the player's entire career. The checkbox is located in Game Setup > League Setup > Rules > Amateur Draft Rules between the Enable Amateur Draft checkbox and the Amateur Draft Date line. As stated above for now definitely leave it unchecked as it needs work.
|
04-15-2010, 09:58 PM | #10 |
OOTP Historical Czar
Join Date: Dec 2001
Location: Bothell Wa
Posts: 7,254
|
Draft Value is based on OWP and DWP and RC/27 and Career Length. I believe that is short for Offensive (Batters) and Defensive (pitchers) Winning Pcts. So far I have not been able to find the formulas that produce these values in the statistical neutering program but if I delete the old values from the output table the neutering process produces new ones. The new values do NOT match the ones that are in the game currently. I do not know if these new values are "better" but they are different.
I suspect the value creation process may need to start with a "tabla rosa" and I probably missed that step when I added this new DV process to the previously existing statistical neutralization process. That or something like it might be why the numbers are off. They are being added to rather than replaced by is my guess. I do not know this for a fact but something like this is definitely at play. Unfortunately my paying job is taking all my time right now but I should be able to review and fix this for one of the patches. At the moment it is too big a job to trace the code without having the necessary time to do it. There is nearly a gigabyte of code and data to sift and I didn't name the process "draft value" and it wasn't until today that I found the OWP output table and figured out that was at least a part of the DV process. Now that I know at least a portion of the subprocess (and what I named it 2 years ago) it should make finding the rest easier. Anyway....Sorry if some of you are creeped out I've been using it as it stands and really I personally kind of like it that not all the sure things are bunched at the very top of round 1. I find it interesting to be able to grab a pretty fair country player in round 5. Maybe thats just because I play rather than sim though. Different strokes for different folks. |
04-15-2010, 10:29 PM | #11 | |
Hall Of Famer
Join Date: Apr 2009
Location: Toronto, ON
Posts: 6,123
|
Quote:
|
|
04-15-2010, 10:42 PM | #12 |
Hall Of Famer
Join Date: May 2006
Posts: 3,625
|
Spritze, thanks very much for your response. I still wasn't clear about who had actually worked on these values, so it's good to know that you're the person. If there is anything that I or anyone else can do to help, don't hesitate to ask. I'm just hoping for a solution and not looking to beat anyone up for it.
By the way, I play out my games and don't sim through, so I probably play the game similarly to your style. But I don't like seeing great players drop unrealistically to later rounds, thus giving the human manager easy pickings to stock an organization with future stars. Others seem to have similar concerns. The fact that you're now seeing some conflicting values when you produced new output may confirm my suspicion, which is that some aspect of the formula used to produce these numbers went awry or values were scrambled by an output or sorting process. Some of the values seem perfectly fine and are in an appropriate sequence. Others seem nonsensical. I hope it's not a problem with the formula itself, although I had suspicions about longevity at first. The two weakest players I used in my examples above are both players who were out of the majors for a number of seasons but reappeared later. Their actual total number of seasons is low, but those seasons were spread out over a much longer time. So I wondered if the beginning and ending dates of their careers might have caused them to get a skewed effect based on supposed longevity. The only problem with that theory is that some of the other examples were players who had even longer careers based on number of seasons and the starting and ending dates of their careers. So they should be rated higher. My only other theory is that some of the values may have been scrambled by an erroneous partial sort or another function that may have changed some of the values. Last edited by Charlie Hough; 04-15-2010 at 10:54 PM. |
04-15-2010, 11:08 PM | #13 |
OOTP Historical Czar
Join Date: Dec 2001
Location: Bothell Wa
Posts: 7,254
|
Keep in mind that this number is produced AFTER neutralization and INCLUDES gap seasons. So that player with the career gap gets credit for all the years he played BASEBALL not just his major league career.
|
04-15-2010, 11:26 PM | #14 |
Hall Of Famer
Join Date: Apr 2009
Location: Toronto, ON
Posts: 6,123
|
|
04-16-2010, 12:17 AM | #15 |
OOTP Historical Czar
Join Date: Dec 2001
Location: Bothell Wa
Posts: 7,254
|
Found the formulas!!!! Yay!!!!!
|
04-16-2010, 12:28 PM | #16 |
Hall Of Famer
Join Date: Apr 2009
Location: Toronto, ON
Posts: 6,123
|
|
04-16-2010, 02:28 PM | #17 |
OOTP Historical Czar
Join Date: Dec 2001
Location: Bothell Wa
Posts: 7,254
|
Ya know, roughly if a player is in the top 4,000 he is most likely a first round draft pick, top 8,000 = second round etc. So I guess I'm not quite understanding the challenge. Keep in mind 139 years worth of players are grouped together so that means 1-139 might be #1 overall draft picks.
The Koufax example is still a first round pick or close and he only had 5 good years and the rest pretty much stunk. Looks like Garlon's formulas are pretty dang good to me, they might not match Houghs expectations but I think the reason the formulas are a bit vague is probably to merely put players in large groups. That said I see some things that could be added to help sort out players a little bit better but I'm not sure it would really change things all that much. But since this is being looked at might as well try a change or two. I'll use the various Spritze db's I'll make on Sunday as test beds. Can't hurt. ALSO draft value has zero, nada, nothing, whatsoever to do with how OOTP calculates a players potential. So don't mix these two things together please. Last edited by Spritze; 04-16-2010 at 02:31 PM. |
04-16-2010, 05:22 PM | #18 | |
Minors (Triple A)
Join Date: Mar 2009
Posts: 245
|
Quote:
|
|
04-16-2010, 05:36 PM | #19 | |
Hall Of Famer
Join Date: Apr 2009
Location: Toronto, ON
Posts: 6,123
|
Quote:
|
|
04-16-2010, 05:38 PM | #20 | |
OOTP Historical Czar
Join Date: Dec 2001
Location: Bothell Wa
Posts: 7,254
|
Quote:
Last edited by Spritze; 04-16-2010 at 05:43 PM. |
|
Bookmarks |
Thread Tools | |
|
|