Intentional walk strategy is broken

[Dyzalot]

In over 3,000 PA in MLB, Billy Hamilton has been intentionally walked one time. In the 2020 OOTP season I just finished simming, Hamilton was intentionally walked 13 times in 615 PA. He was batting 8th in the lineup for most of the season. Why is it that the current chess AI's can calculate moves out to the hundreds but the current OOTP AI can't even see the value of pitching to a horrible batter with two outs and a runner in scoring position in order to make the pitcher lead off the next inning? And why is Hamilton the only one on the list that can't hit and bats 8th? Does the AI use a different strategy against the human controlled team than the others when simming?

[HerbD]

It's simple baseball strategy. He's hitting in front of the pitcher now. Before he was hitting leadoff for the Reds and Royals.

Sent from my Pixel 2 XL using Tapatalk

[CBeisbol]

Originally Posted by HerbD

It's simple baseball strategy. He's hitting in front of the pitcher now. Before he was hitting leadoff for the Reds and Royals.

Sent from my Pixel 2 XL using Tapatalk

Every team, in one league at least, has a player hitting in front of the pitcher. Yet, Hamilton has more walks than the rest, despite, probably, being a worse hitter.

[Sweed]

Originally Posted by CBeisbol

Every team, in one league at least, has a player hitting in front of the pitcher. Yet, Hamilton has more walks than the rest, despite, probably, being a worse hitter.

Doesn't that just suggest that he, by chance, came up in that circumstance (2 outs, runner in scoring position) more than other batters?

Kind of like how RBIs are not considered to be important because they are dependent on a situation out of the batter's control. This, in a similar way, is driven by a condition, no?

[Dyzalot]

Originally Posted by Sweed

Doesn't that just suggest that he, by chance, came up in that circumstance (2 outs, runner in scoring position) more than other batters?

Kind of like how RBIs are not considered to be important because they are dependent on a situation out of the batter's control. This, in a similar way, is driven by a condition, no?

Ok but that still means the strategy is broken. Hamilton is barely better than a pitcher. And it doesn't explain why the other four hitters on the list are all good hitters not hitting in the 8 spot. But yes, if the default is to just IBB the #8 hitter every time with two outs, a runner in scoring position and pitcher on deck then the strategy is broken. MLB managers don't do this. When is the last time a crappy #8 hitter led the majors in IBB? According to Baseball Reference, never? If not never then its at least been a few decades.

https://www.baseball-reference.com/l..._leagues.shtml

Those stats would imply that the AI isn't intentionally walking the good hitters enough if 13 IBB for a #8 hitter led both leagues.

[CBeisbol]

Originally Posted by Sweed

Doesn't that just suggest that he, by chance, came up in that circumstance (2 outs, runner in scoring position) more than other batters?

Kind of like how RBIs are not considered to be important because they are dependent on a situation out of the batter's control. This, in a similar way, is driven by a condition, no?

That's a possibility, yes.

Originally Posted by Dyzalot

Hamilton is barely better than a pitcher.

That's not really true
Your Hamilton had a 52 OPS+
Pitchers in 2019 had a -13 OPS+
That's a huge difference

[Questdog]

Originally Posted by Dyzalot

Ok but that still means the strategy is broken. Hamilton is barely better than a pitcher. And it doesn't explain why the other four hitters on the list are all good hitters not hitting in the 8 spot. But yes, if the default is to just IBB the #8 hitter every time with two outs, a runner in scoring position and pitcher on deck then the strategy is broken. MLB managers don't do this. When is the last time a crappy #8 hitter led the majors in IBB? According to Baseball Reference, never? If not never then its at least been a few decades.

https://www.baseball-reference.com/l..._leagues.shtml

Those stats would imply that the AI isn't intentionally walking the good hitters enough if 13 IBB for a #8 hitter led both leagues.

In 2019, which was not so long ago, 4 #8 hitters were in the top 10 in intentional walks in the NL: Maikel Franco, Carson Kelly, Alex Avila and Tucker Barnhardt.

And in real life, the Reds batted Billy Hamilton 9th a lot, after the pitcher...

P.S. Also from 1983 to 1985, a #8 hitter led the NL in intentional walks each season.

[BBGiovanni]

I wonder if the engine does what real pitchers often do, namely throw a wide one after getting to 3-0 rather than groove one, or "pitch around" rather than give up a hit. Obviously real pitchers don't do that, but if the engine thinks that way it may not care that Hamilton is a noodlebat with winged feet.

[Dyzalot]

Originally Posted by CBeisbol

That's not really true
Your Hamilton had a 52 OPS+
Pitchers in 2019 had a -13 OPS+
That's a huge difference

Pretty sure my pitchers were better than that. But yes, "barely" was an exaggeration.

[Questdog]

They A.I. will not consider the hitting quality of a pitcher compared to the #8 hitter when deciding whether to issue an intentional walk or not, unless the pitcher is a two-way player. This is a blind spot of the A.I, to be sure, but not a large one.

[Dyzalot]

Originally Posted by Questdog

In 2019, which was not so long ago, 4 #8 hitters were in the top 10 in intentional walks in the NL: Maikel Franco, Carson Kelly, Alex Avila and Tucker Barnhardt.

And in real life, the Reds batted Billy Hamilton 9th a lot, after the pitcher...

P.S. Also from 1983 to 1985, a #8 hitter led the NL in intentional walks each season.

So never then? Hamilton led the ML in my sim. And he led it with a number that is about half the number that the average ML leader had over the past decade. I'm not even sure if the problem was that Hamilton got walked too much. But certainly the good hitters around the league weren't given enough free passes. At least if the top five in my sim were all #8 hitters then there'd be some kind of pattern to fix. I just find it really strange that the leader was the #8 hitter on the human controlled team and the rest of the top five were all top hitters from AI controlled teams. I'm not even sure what to make of that. It is just weird.

[Questdog]

P.S. Alex Avila only batted 8th in 38 games and was walked 7 times intentionally in 2019. That is a higher rate of IBB per game than the League Leader, Cody Bellinger. Maikel Franco batted 8th only 69 times, but was walked 15 times, a rate much higher than Bellinger's. In fact, he only missed by 2 of tying Bellinger for the League lead.

[Questdog]

Originally Posted by Dyzalot

So never then? Hamilton led the ML in my sim. And he led it with a number that is about half the number that the average ML leader had over the past decade. I'm not even sure if the problem was that Hamilton got walked too much. But certainly the good hitters around the league weren't given enough free passes. At least if the top five in my sim were all #8 hitters then there'd be some kind of pattern to fix. I just find it really strange that the leader was the #8 hitter on the human controlled team and the rest of the top five were all top hitters from AI controlled teams. I'm not even sure what to make of that. It is just weird.

One reason could be that you are the only team in your league to bat the same player consistently at #8.

[Dyzalot]

Originally Posted by Questdog

P.S. Alex Avila only batted 8th in 38 games and was walked 7 times intentionally in 2019. That is a higher rate of IBB per game than the League Leader, Cody Bellinger. Maikel Franco batted 8th only 69 times, but was walked 15 times, a rate much higher than Bellinger's. In fact, he only missed by 2 of tying Bellinger for the League lead.

Avila can actually hit though.

One reason could be that you are the only team in your league to bat the same player consistently at #8.

That's possible although I'd be surprised if no other NL teams had a consistent #8 hitter. Probably the bigger "discrepancy" is just how little other hitters get intentionally walked compared to actual MLB. From 2010 to 2019 the average MLB leader had 20+ IBB. Hamilton led my sim with 13. Probably should be seeing more IBB walks for the guys like Betts, Trout, Bellinger, etc.

[polydamas]

We need a larger sample size

[Questdog]

Originally Posted by Dyzalot

Avila can actually hit though.

Not like Bellinger.... Yet, he was walked more often.

Originally Posted by Dyzalot

That's possible although I'd be surprised if no other NL teams had a consistent #8 hitter. Probably the bigger "discrepancy" is just how little other hitters get intentionally walked compared to actual MLB. From 2010 to 2019 the average MLB leader had 20+ IBB. Hamilton led my sim with 13. Probably should be seeing more IBB walks for the guys like Betts, Trout, Bellinger, etc.

Check out the total IBB rate in your league compared to real baseball. Checking the league leaders against one another is not a reliable comparison because of the small sample size.

[CBeisbol]

Originally Posted by Questdog

One reason could be that you are the only team in your league to bat the same player consistently at #8.

I think you've cracked the code

8th place hitters led the majors in IBBs in 2019

[Dyzalot]

Originally Posted by Questdog

Not like Bellinger.... Yet, he was walked more often.

What I meant by that is if Hamilton was as good a hitter as Avila, it wouldn't have shocked me as much.

Originally Posted by Questdog

Check out the total IBB rate in your league compared to real baseball. Checking the league leaders against one another is not a reliable comparison because of the small sample size.

673 IBB in 188,517 PA for my 2020 sim. Without looking at actual MLB numbers, I'm going to assume that is quite low. I mean the NYY were only issued 13 IBB for the entire season. That can't be even close to realistic. Oakland was only issued 9. My team, the SF Giants, led the MLB with 44 intentional walks. Seeing as how we only won 68 games and ranked outside the top ten in the NL in every offensive category except stolen bases, that seems weird.

[Dyzalot]

Originally Posted by CBeisbol

I think you've cracked the code

8th place hitters led the majors in IBBs in 2019

Yeah but look at the totals. Looks like there were more than twice as many IBB issued by MLB in 2019 compared to my sim. And the fact that my crappy hitting team led the majors doesn't jive either.

[CBeisbol]

Originally Posted by Dyzalot

673 IBB in 188,517 PA for my 2020 sim. Without looking at actual MLB numbers, I'm going to assume that is quite low.

IRL was 753 in 134767.
A little low
Not quite low

It may be following a trend of fewer IBBs

I mean the NYY were only issued 13 IBB for the entire season. That can't be even close to realistic.

The IRL Yankees had 18
https://www.baseball-reference.com/l...=MLB&year=2019

Oakland was only issued 9.

Oakland 17

My team, the SF Giants, led the MLB with 44 intentional walks

The Giants had 26