OOTP Developments Forums

OOTP Developments Forums (https://forums.ootpdevelopments.com//index.php)
-   OOTP 20 - General Discussions (https://forums.ootpdevelopments.com//forumdisplay.php?f=3956)
-   -   AI Too Heavily Favoring Defense (https://forums.ootpdevelopments.com//showthread.php?t=308264)

majesty95 11-13-2019 04:15 AM

AI Too Heavily Favoring Defense
 
4 Attachment(s)
So, I have a single season sim runnign where we've added past or future stars to our 1987 replay. The issue is, some of the players are not playing. For instance, Shoeless Joe is getting benched for Daryl Boston. While less ridiculous but still unreasonable is Robin Venture playing behind Donnie Hill. There are others I'm sure. I just haven't gone through each team eyet.

The only reason I can gather is that the AI is over evaluating fielding. If they were reasonably close in ratings I could buy it but they're trading a 7.5 WAR player for a 0.7 WAR player and a 4.7 WAR player for another 0.7 one. I get the AI doesnt know their real life WAR but they can see the ratings. The only logical reason is the defense but even that isn't logical with as many wins as they are giving up...

I guess I could take the time to force start all of these players but A. that's time consuming and B. I've always seen the AI plays those guys until they drop. Like they won't substitute for them to keep them fresh until they damn near dead. There's no setting to control this that I know of. Do I have any other options? I tried changing the AI eval settings to all kinds of things and none seem to make a difference...

majesty95 11-13-2019 04:21 AM

2 Attachment(s)
This one makes even less sense. Really no sense. The AI is starting Roy Smalley at DH over Rod Carew...

David Watts 11-13-2019 09:43 AM

I have a random debut league going right now and the New York Giants benched Derek Jeter for the entire 1903 season in favor of Jonathan Villar. Jeter is a dreadful 36 defensively at SS, while Villar is I think either a 50 or 60. I'm using one year recalc, so I will be very interested to see what happens next season. I think Villar recalced to his best MLB season for 1903, 285, 19 home runs and 62 stolen bases. Next season he should crumble, so I'm guessing Jeter will be back starting.

koohead 11-13-2019 11:38 AM

I'm not in front of the game at the moment and can't find a screenshot online but is there anything in Manager Profiles that may impact lineup decisions? What if you changed from Traditional to Saber or vice versa?

majesty95 11-13-2019 11:42 AM

Quote:

Originally Posted by koohead (Post 4559790)
I'm not in front of the game at the moment and can't find a screenshot online but is there anything in Manager Profiles that may impact lineup decisions? What if you changed from Traditional to Saber or vice versa?

There are no coaches. Everything is setup by default by the AI

Reed 11-13-2019 12:14 PM

That is very curious and I wish I had an answer. My wild guess would be some strange things happen when you mix players from various years. Hopefully someone will have a reasonable explanation.

koohead 11-13-2019 01:13 PM

Quote:

Originally Posted by majesty95 (Post 4559792)
There are no coaches. Everything is setup by default by the AI

I guess I was thinking of the "strategies" sliders.

majesty95 11-13-2019 01:19 PM

Quote:

Originally Posted by koohead (Post 4559845)
I guess I was thinking of the "strategies" sliders.

Meaning going into each team? I guess theoretically I could do that. Just seems like you shouldn't have to in order to get a Hall of Famer to start over a 2-star lol

actionjackson 11-13-2019 07:27 PM

I have to admit that when I first read the title of the thread, I was like "Yes...Now the Mark Belangers of the world will be set free from being stapled to the bench, and the full value of their contributions will be accounted for...WOOO-HOOO!!!"...And then I opened up the thread, and...What the?

The irony of the Ventura/Hill one is that Ventura consistently wins gold gloves when he's in my leagues because he was such a great fielder. The dude could hit too.

I remember the hype surrounding Daryl Boston when he first came up. He had awesome tools, but he'd be a mediocre 4th or 5th OF option on most teams no? Shoeless Joe was on his way to Cooperstown (even though the Hall didn't exist back then) before the scandal hit in 1919. Hill was a decent backup infielder/utility player, but not a starter, and definitely not over Robin Ventura. Smalley was a decent hitter with very good pop for a SS, but we're talking about Rod freakin' Carew here.

None of these make any sense really, so I'd like to know what's going on as well. I doubt it's anything in the settings because Jackson, Ventura, and Carew are clearly superior players that are getting jobbed. Gonna have to keep an eye out for this one myself.

My only question is were Boston, Hill, and Smalley III originals on their teams, while Jackson, Ventura, and Carew were the ringers that arrived in the time machine? Could that be messing with the AI? It shouldn't be because in each case the ratings make it crystal clear who the better player is. That's all I've got for ya. I'm completely flummoxed.

CBeisbol 11-13-2019 09:13 PM

Quote:

Originally Posted by majesty95 (Post 4559692)
So, I have a single season sim runnign where we've added past or future stars to our 1987 replay. The issue is, some of the players are not playing. For instance, Shoeless Joe is getting benched for Daryl Boston. While less ridiculous but still unreasonable is Robin Venture playing behind Donnie Hill. There are others I'm sure. I just haven't gone through each team eyet.

The only reason I can gather is that the AI is over evaluating fielding. If they were reasonably close in ratings I could buy it but they're trading a 7.5 WAR player for a 0.7 WAR player and a 4.7 WAR player for another 0.7 one. I get the AI doesnt know their real life WAR but they can see the ratings. The only logical reason is the defense but even that isn't logical with as many wins as they are giving up...

I guess I could take the time to force start all of these players but A. that's time consuming and B. I've always seen the AI plays those guys until they drop. Like they won't substitute for them to keep them fresh until they damn near dead. There's no setting to control this that I know of. Do I have any other options? I tried changing the AI eval settings to all kinds of things and none seem to make a difference...

Did you do any tests to see if, somehow, the defensive player actually created more wins?

That's an idea that's been bandied about here recently.

majesty95 11-13-2019 10:01 PM

Quote:

Originally Posted by CBeisbol (Post 4560062)
Did you do any tests to see if, somehow, the defensive player actually created more wins?

That's an idea that's been bandied about here recently.

I haven’t. Some circles will say defense is overrated. It started with Moneyball. Remember Hatterberg at 1B? I can’t see how they’d make up three wins in defense. And in the case of Carew, no defense to factor in...

I suppose it could be that they are indeed the ringers, but there are guys on other teams actually playing. Not every ringer is getting sat. It’s weird too bc I’ve run these before on previous versions and never had an issue.

Dyzalot 11-13-2019 10:59 PM

Hatteberg worked at first base because defensive ability is almost irrelevant at that position. The difference in wins saved by a premier defensive first basemen and a guy like David Ortiz is quite minimal. The Scott Hatteberg experiment doesn't work at any other position except maybe left field. That doesn't disqualify defense being important at other positions. Saying all of that, I tend to agree with the OP that the AI seems to be kind of messed up here in choosing who plays.

majesty95 11-14-2019 03:56 PM

So what do you guys think? My only option at this point to force start them?

CBeisbol 11-14-2019 04:17 PM

Quote:

Originally Posted by majesty95 (Post 4560301)
So what do you guys think?

That you should sim a few hundred (or thousand) seasons with each as the starter to see what the actual difference in wins is.

Anyone 11-14-2019 04:44 PM

Quote:

Originally Posted by CBeisbol (Post 4560306)
That you should sim a few hundred (or thousand) seasons with each as the starter to see what the actual difference in wins is.

Of course, if a team wins more with Daryl Boston starting than with Shoeless Joe, then it would mean the game sim was overvaluing defense in outcomes rather than just in managers' choosing who to start.

Wingedlion14 11-14-2019 05:35 PM

Well defense is important. An extreme example here, but in a game of mine the Indians had JD Martinez play 150 games in RF. He had 6 WAR on offense, and -5.5 on defense. They basically spent $23 million and gave up prospects to get a barely replacement level guy, all because they played him in the field.

Of course, that’s an extreme scenario because JD Martinez was rated 20/80 in RF, and they had a below-average CF too which didn’t help either.

CBeisbol 11-14-2019 06:04 PM

Quote:

Originally Posted by Anyone (Post 4560309)
Of course, if a team wins more with Daryl Boston starting than with Shoeless Joe, then it would mean the game sim was overvaluing defense in outcomes rather than just in managers' choosing who to start.

Well, there are a few options, right?

1) These players produce fewer wins and the AI is overvaluing these players for whatever reason. Maybe defense. Maybe scouting. Etc..

2) these players actually do produce more wins.

majesty95 11-14-2019 06:05 PM

You guys are so focused on defense or defending the game or whatever your agenda is that you're completely overlooking the fact that there is ZERO reason why Roy Smalley should be playing DH over Rod Carew. I would also argue that the advantage that Hill has over Ventura defensively is minimal. Lastly, don't the star levels take into account defensive ratings?

CBeisbol 11-14-2019 06:12 PM

Quote:

Originally Posted by majesty95 (Post 4560338)
You guys are so focused on defense or defending the game or whatever your agenda is that you're completely overlooking the fact that there is ZERO reason why Roy Smalley should be playing DH over Rod Carew.

Scouting? R/L splits? ???
Quote:

Lastly, don't the star levels take into account defensive ratings?
Yes. But surely you've identified some players that perform beyond (or below) their overall rating

Anyone 11-14-2019 06:43 PM

Quote:

Originally Posted by CBeisbol (Post 4560336)
Well, there are a few options, right?

1) These players produce fewer wins and the AI is overvaluing these players for whatever reason. Maybe defense. Maybe scouting. Etc..

2) these players actually do produce more wins.

Yes, but my point is if it's #2 there's a big problem with the sim making that so, because there is probably no knowledgeable baseball writer or fan in the country, from the most sabermetric-oriented to the most traditional, who, if one ignores the Black Sox Scandal (which isn't part of player ratings so one can ignore it), would not agree that Joe Jackson was a far more valuable player than Daryl Boston.


All times are GMT -4. The time now is 07:50 PM.

Powered by vBulletin® Version 3.8.10
Copyright ©2000 - 2024, vBulletin Solutions, Inc.
Copyright © 2020 Out of the Park Developments