Strat-O-Matic Baseball 365

I'd like to run an experiment in which we try create an environment which will allow for the most accurate simulation of a given player card (i.e.. how close can we replicate Ruth's 1921 card where he hit .378 with 59 HRs and had a 1353 OPS).

Some thoughts I've had are:

First: I believe it would be necessary to have all 12 teams play in Turner (8/8 - 8/8) as these BP ratings represent the base values on all player cards?

Second: I'm pretty sure that SOMO salary structure is based upon an $80mil salary cap, so would that be the logical place to set the cap in this experiment?

Third: It think it will be necessary to structure the player set to represent things like Non-DH line-ups vs Non-DH pitchers, not allowing modern relief pitching against dead-ball hitters, etc.?

Any input, thoughts and/or ideas?

My understanding (from conversations with Strat in the 80s) was that the formula for creating the card cards is calculated to reproduce the actual season with the same ballparks, against the hitters or pitchers from that year. In other words, each season is a closed system and the only way to reproduce the results is to play that exact season: teams, cards ballparks and all. It's certainly very possible that Strat has changed that. They've always been somewhat secretive about their formula for creating the cards.