Quote:
Originally Posted by CBLCardinals
Good question
I did auto calc, then selected the 2023 season in the dropdown box for statistical outputs.
From there I’m messing around with getting pitcher stamina closer to real life. Right now I am using “Low”, but the modifiers on 1.000.
The innings per start is getting really close to sample size of 2023 pitchers.
If you take someone like Kevin Gausman, he went about 6 IP per start in 2023.
With simulation he went 6.1 IP per start.
The overall innings totals are higher since he started more games in the predictions sim.
|
Thanks for responding. Just to be clear, I was not in any way conveying a comment or critique on either your process or results. I was curious because I find that (through v24, at least; bought but have not downloaded v25 yet) not running AutoCalc before even the first season of a new standard MLB save results in some significantly off-target league stats at the end of the year (particularly for batting average). I do not really understand why if a) the initial ratings are not engine generated but based on imported ZiPS (that, if anything, are conservative in variance compared to end-year MLB results due to the nature of using median projections) and b) the target league totals that determine LTMs are similarly a set of direct inputs, AutoCalc should be so necessary even for year one. It makes one (well, me, at least) question if the AutoCalc is being relied on as a “Make This Work” corrective in lieu of playtesting and subsequent optimization by the developers. While such a tool might be sufficient for realizing some broad target such as league totals and upper and lower individual outliers, I am (based on observation with OOTP and experiences more generally) skeptical that it does not come at a cost in terms of distributions and valuations of particular attribute profiles versus a more curated approach.
Anyway, you did not sign up for reading long winded expositions on overthought yet undercooked contemplations, so I will see myself out …