AI Lineups and Pitchers Who Can Hit

[Charlie Hough]

It's been a long time since I've run a historical simulation in the deadball era, but I just noticed something strange involving AI lineups and certain pitchers who can hit well. I'm playing a game in 1916, and the Red Sox have two pitchers with excellent and decent batting ratings, respectively: Babe Ruth and Carl Mays.

The AI bats Ruth 9th, much like Red Sox manager Bill Carrigan did in real life. This is fine because managers in the deadball era generally batted the pitcher 9th, regardless of how well he could hit. This was the case with Ruth as well as other pitchers who could hit better than some of their teammates, such as Walter Johnson.

However, here's where it gets strange. Carl Mays, who is also on the Red Sox and is a decent hitter, was just placed 8th in the starting lineup rather than 9th. But his batting ratings are nowhere near as good as Ruth's.

Normally, the AI bats Mays 9th, but apparently it batted him 8th because the backup CF was playing, and he is a considerably weaker hitter than Mays. Thus, the AI placed the CF in the 9th spot. But if it's going to base lineup decisions on batting ratings, then why wouldn't the AI place Ruth 8th or higher in the same situation? He's always better than every other hitter in the lineup, including the player who bats 8th. But the AI always bats Ruth 9th in the lineup, usually behind a hitter who is only half as good as he is.

This might be to protect against a subsequent relief pitcher batting higher in the lineup, but the AI could do a double-switch to prevent this. It just seems strange that the AI would do this in one case but not the other.

Has anyone else seen the AI do something like this with pitchers who can hit well? This is a minor anomaly, so it probably doesn't warrant a fix, but it still seems oddly inconsistent.

[Brad K]

AI for a different team may have made decisions you consider to be better.

[Syd Thrift]

There’s also the notion, which I don’t think is borne out by the data but it exists, that in some cases hitting your worst hitter 8th is better than hitting them 9th. The reasoning is that the only time you can guarantee that your lead off hitter will lead off an inning is the first one and so having a second lead off hitter in the 9 hole can kick start the heart of your order by getting more players in scoring position for them. I know Tony LaRussa loved that crap. I think the data say that the #9 hitter gets the fewest PAs, a lot fewer over the course of the season over your 2 hole hitter for instance, and as such you really shouldn’t play games like that.

But it is a strategy and it does get employed, or at least it did when pitchers hit.

[Charlie Hough]

Originally Posted by Brad K

AI for a different team may have made decisions you consider to be better.

It's not really a matter of whether the decision is better or worse. It just appears to be inconsistent with how the AI is coded to evaluate batters and handle pitchers in the lineup. It appears to be doing something rather inconsistent or self-contradictory. Of course, human managers can do the same thing, but they're not hard-coded by developers.

For me, it's not a question of whether the strategy exists in real life or makes sense. It already does on both counts, but it's weird that the AI would do it in the case of Mays but not apply it where it makes even more sense with Ruth.

[luckymann]

Are you referring to v23 here Charlie?

[Syd Thrift]

Originally Posted by Charlie Hough

It's not really a matter of whether the decision is better or worse. It just appears to be inconsistent with how the AI is coded to evaluate batters and handle pitchers in the lineup. It appears to be doing something rather inconsistent or self-contradictory. Of course, human managers can do the same thing, but they're not hard-coded by developers.

For me, it's not a question of whether the strategy exists in real life or makes sense. It already does on both counts, but it's weird that the AI would do it in the case of Mays but not apply it where it makes even more sense with Ruth.

Because the strategy is not about Ruth vs Mays, it’s about the #7 or #8 hitter.

[Charlie Hough]

Originally Posted by Syd Thrift

Because the strategy is not about Ruth vs Mays, it’s about the #7 or #8 hitter.

Okay, so what's the strategy? In one case, it's placing its weakest hitter 8th with its best hitter 9th, and, in the other case, it's placing its weakest hitter 9th and a better hitter 8th. What's the strategy behind it? What logic is it using to make this determination? And why is it inconsistent?

As for luckymann's question, yes, this is with OOTP 23.

[Syd Thrift]

No, like I said, it’s using 9th as a second lead off slot.

[luckymann]

Originally Posted by Charlie Hough

Okay, so what's the strategy? In one case, it's placing its weakest hitter 8th with its best hitter 9th, and, in the other case, it's placing its weakest hitter 9th and a better hitter 8th. What's the strategy behind it? What logic is it using to make this determination? And why is it inconsistent?

As for luckymann's question, yes, this is with OOTP 23.

I suggest then that it might have something to do with the AI settings for two-way players and the Ohtani Rule. Have you got the Ohtani Rule on or off?

[Charlie Hough]

Originally Posted by Syd Thrift

No, like I said, it’s using 9th as a second lead off slot.

So it's going to put its worst hitter in that second leadoff spot, with no regard for the batting ratings of the pitcher who then bats 8th, and no regard for whether the field player batting 9th has a good OBP, speed, or baserunning? That's what it did when it batted Mays 8th in the lineup. If it actually wanted to treat the 9th spot as a second leadoff slot, it would have batted Mays in that slot, because he was a much better hitter than the field player and also had good speed and baserunning. The whole thing seems as if it's applying a strategy based on the positions of the players and not their ratings.

[Charlie Hough]

Originally Posted by luckymann

I suggest then that it might have something to do with the AI settings for two-way players and the Ohtani Rule. Have you got the Ohtani Rule on or off?

There is no DH in the deadball era, so the Ohtani rule is greyed out in the league settings. The box is checked, but it's greyed out and wouldn't be applied anyway.

[David Watts]

Very strange. Whenever I've seen a OOTP manager use the bat the pitcher 8th strategy it does so every single game of the season. I've never seen it do so for just one pitcher on the staff.

[Brad K]

In my current save back in the 70s my AI manager wanted Freddie Patek to lead off. I value OBP over speed for lead off so I did the player strategy deal and made Patek bad 9th.

[Charlie Hough]

Originally Posted by David Watts

Very strange. Whenever I've seen a OOTP manager use the bat the pitcher 8th strategy it does so every single game of the season. I've never seen it do so for just one pitcher on the staff.

That's more of what I would expect, but, in this case, the AI only did this for one game. It was when there was a reserve outfielder who was going to make a rare start and would have potentially hit in the 8th spot. This is why I'm wondering if there is some logic where the AI will make this move if the player who would normally be placed in the #8 slot is below a certain ratings threshold.

In all the other box scores I've reviewed for the Red Sox where Mays was pitching, it didn't do this in the other cases. I haven't sampled all the box scores for the season thus far, but it doesn't appear to have done this with any other Red Sox pitcher either.

[luckymann]

Originally Posted by Charlie Hough

There is no DH in the deadball era, so the Ohtani rule is greyed out in the league settings. The box is checked, but it's greyed out and wouldn't be applied anyway.

Yes, of course - silly me.