|
Amaroq -> RE: BIG Feature Request (7/21/2006 11:13:03 PM)
|
quote:
ORIGINAL: KG Erwin
Simply put, let's expand the real-player historical leagues to allow for full minors. Yes, this does require fictional players to be included, BUT, as the seasons progress, the fictional scrubs will be weeded out.
My desire is for 50-man rosters in a historical real-roster league.
Does anyone else support this request?
Shaun already knows my feelings on this - I completely agree, that's exactly what I would want to play.
...
Historically, you could do this with PureSim 2005 - if you played 50-man rosters in 1901, with about 20 'real' players for each of the 16 teams, it would generate 30 'super scrubs' - players who never develop beyond the '10 point scrub' level - their maximum value in any hitting or pitching category was 10 on the 100-point scale.
If you played this way, by 1904, you had enough 'real' players that every team was pretty solid - there might be 3 to 5 'scrubs' left in some minor league systems, but most players were 'real'.
However, it had a secondary problem: if a player made his big-league debut in 1904 in real life, he would typically not be quite big-league quality in game terms - usually he was about three years of development time away from 'the show'. You'd give him that development time before bringing him up to the majors - and sometimes might leave guys in the minors four or five years just because you have better players ahead of them - but they would still decline and retire at the 'correct' historical ranges, so most players had careers three to four years shorter than was realistic, resulting in noticeably low all-time/career records.
Shaun removed the feature for PS'07 for performance reasons: if you did the 1901 setup, when 1902 came, ALL of those ten-point-scrubs became free agents. The AI's would get to a point during the 1902 free agency where there were about 500 free agents, and each team would i.d. that it needed to sign one, and try to sign the same guy... leaving you 499, which the AI approached in the same way. It took a long time, and wasn't easily fixable, so Shaun put in the 'real players == 35-man roster' restriction.
...
I quite enjoyed the PS'05 experience - I think if Shaun could approach this by simply fixing the AI bug - perhaps the fix would be as simple as making the length of contract for 10-point-scrubs be based on the round they were drafted in: a scrub taken in rounds 0-30 gets a 4-year contract, 31-40 gets a 3-year contract, 41-50 gets a 2-year contract, and 51-60 gets a 1-year contract (assuming a 60-man roster limit; adjust those distributions for 50, 55, and 70) - then that would suffice for the majority of users.
In the long-term, I'd like to see an *option* for minor-league support where P.S. imports 'real' players typically at either age 17-19 (high school) or ages 21-23 (collegiate).
I tried using a modified Lahman DB from the OOTP community which addressed this, but it wasn't compatible with PS - whatever DB entry PS is using to determine debut year wasn't updated appropriately in the DB.
The 'quick' way to do this would be to simply import all players three years prior to their big-league debut date. Its rough, and hacky, and would lead to things like drafting Nolan Ryan at age 16 [;)] but I think it would give you a reasonable sense of it. Among other things, the 1901 'initial draft' would include all players from 1901-1904 - which would mean that the 50-man rosters were nearly completely full, dramatically reducing the number of 'scrubs' available.
Even better would be creating some form of mapping table, e.g:
Real P.S
Debut Debut
--------------
<18 N
18 17
19 18
20 18
21 19
22 19
23 21
24 21
25 22
26 22
27 23
28 23
>28 N
That would at least 'feel' like most players were available either out of high school, or out of college. It would still have some problems: you'd draft Dave Winfield at age 19, rather than at age 21 - but it would get some things right, as you'd draft Nolan Ryan as an 18-year-old.
Best would be to combine that with 'transaction info', as you can find at baseball-reference.com, to discover when players were first available in real baseball. Even if that information isn't entirely available from the Lahman DB, it is available on Retrosheet, where you can find Winfield in the 1973 draft at Retrosheet: Transactions, 1973. Since PureSim already has to do some pre-Lahman processing on the DB, you could make correlating the retrosheet data with the Lahman data part of that.
...
If you ran one step further with that, you could even use the Retrosheet data in a 'historical re-creation association' format, using it to guide teams to make the actual roster moves that they made historically. For example, we can see that on April 17, 1925, the New York Giants traded Rosy Ryan to the Boston Braves in exchange for Tim McNamara.
That sort of detail has never made sense from the perspective of a GM-game, where roster transactions are a huge part of the game... but it might make sense for the 'historical manager' perspective: if I want to manage the team with the exact players that the team had on that day, using the real historical schedules, etc - can I, through my in-game managerial talent alone, break the Curse of the Bambino more quickly than the real Red Sox did?
|
|
|
|
