Protection in a line up

[Baseball Ray!]

I'd be interested in getting feedback from anyone willing to share. I am currently doing a save that is tracking the MLB career of Dale Murphy, starting with the 1978 season (even though his best years didn't occur until the 80's).

However, what I am finding as I play through 1978 another star is emerging - Bob Horner. The lineup also has some other decent hitters including Murphy (his first full year in the majors), Jeff Burroughs, and Gary Matthews. Right now, I have Horner batting 5th, but I'd like to bump him to 3rd (making sure he gets an AB in the 1st inning among other things - he's leading the team in almost all offensive categories).

So, in contemplating this, it got me thinking as important in today's game (2020) the strategy of having protection hitting behind your sluggers so they don't get pitched around, etc., in OOTP, does the AI for the opposing team make this same assessment of a teams lineup, but in reverse? Or, does the AI simply evaluate each batters contact/power/eye numbers and throw balls and strikes on that basis? For example, if Gary Matthews is hitting in front of Bob Horner, Matthews would be pitched to differently by the AI, versus if he was hitting in front of say, 1 1/2 star Darryl Chaney (who I just sent down).

I'm ignoring obvious cases like the batter hitting before a pitcher is due up and first base is open, so the batter is intentionally walked to get to the pitcher.

Thanks,
Steve

[Dyzalot]

I find the AI to be pretty "unsophisticated" in this area. It will sometimes intentionally walk a weak, lefty hitter with a lefty pitcher on the mound in order to get to the pitcher's spot without taking into consideration that there's a right handed hitting power hitter on the bench ready to hit for that pitcher. It seems like most "in game" decisions by the AI are almost done in a vacuum, with not much thought about what might come next. It is possible though that the scheduled next hitter has some impact on how often the AI might employ the intentional walk or "pitch around" strategies.

[Baseball Ray!]

Thanks for your comments - I thought this might have generated many more. With the sophistication of machine learning & AI in business and technology, I wonder just how far this can be harnessed to reproduce the realism of in-game management.

With the current AI, I have sometimes scratched my head when the opposing AI manager would bring in a lefty reliever against a lineup of right handed batters (it was not a ninth inning closer situation). I will admit I did not check afterwards what the strength of their BP was at the time of the switch.

[Dyzalot]

Originally Posted by Baseball Ray!

Thanks for your comments - I thought this might have generated many more. With the sophistication of machine learning & AI in business and technology, I wonder just how far this can be harnessed to reproduce the realism of in-game management.

With the current AI, I have sometimes scratched my head when the opposing AI manager would bring in a lefty reliever against a lineup of right handed batters (it was not a ninth inning closer situation). I will admit I did not check afterwards what the strength of their BP was at the time of the switch.

I think the problem is the way people use the term "AI". In OOTP, it is simply a set of rules that must be followed per the people who programmed those rules. There is no "machine learning" or advanced AI concepts such as you see with Elon Musk using an AI to beat top League of Legend players. Basically the "AI" in OOTP is really just a finite set of "if/thens". I imagine it would take a complete overhaul of the game engine to incorporate some of the advanced AI concepts that are being used by top developers and programmers these days.