Beef With the AI's Logic

Zeno · 08-13-2010, 11:05 PM

Overall, I am very pleased with the game. It's a lot of fun and gets better every year, but I have an issue with the way the AI manages it's lineups.

This has happened in a number of situations, but I will use the one that inspired me to post this on the forum.

In this particular simulation, Ryan Madson was the closer for the Phillies in 2010. He compiled 49 saves in 51 attempts and finished the season with a 0.93 ERA and a 0.84 WHIP. A career year for even the most elite of closers.

Off-season comes around and the Phillies make some moves. I always look at every roster on opening day to see any trades or signings I might have missed.

To my surprise, Madson was no longer the closer but was the setup man. The closer was Michael Wuertz who ended 2010 with 16 saves, 4.21 ERA and 1.40 WHIP.

The game relies too heavily on ratings. Even if Madson's abilities deteriorated, there is no way any organization in the history of baseball would not have him as their closer. The past success would instantly trump any performance deficiency.

I'm not saying that the ability should not factor into a team's decision when there are borderline calls, but when a player has such a spectacular season, the teams should be forced to make ignorant decisions that any GM would make in real life.

fspikec · 08-13-2010, 11:57 PM

Quote:

Originally Posted by Zeno

Overall, I am very pleased with the game. It's a lot of fun and gets better every year, but I have an issue with the way the AI manages it's lineups.

This has happened in a number of situations, but I will use the one that inspired me to post this on the forum.

In this particular simulation, Ryan Madson was the closer for the Phillies in 2010. He compiled 49 saves in 51 attempts and finished the season with a 0.93 ERA and a 0.84 WHIP. A career year for even the most elite of closers.

Off-season comes around and the Phillies make some moves. I always look at every roster on opening day to see any trades or signings I might have missed.

To my surprise, Madson was no longer the closer but was the setup man. The closer was Michael Wuertz who ended 2010 with 16 saves, 4.21 ERA and 1.40 WHIP.

The game relies too heavily on ratings. Even if Madson's abilities deteriorated, there is no way any organization in the history of baseball would not have him as their closer. The past success would instantly trump any performance deficiency.

I'm not saying that the ability should not factor into a team's decision when there are borderline calls, but when a player has such a spectacular season, the teams should be forced to make ignorant decisions that any GM would make in real life.

You can customize the AI's evaluation and then make the ratings based of that evaluation. It can be weighted based on 4 different fields (Ratings, Current year stats, last year stats and 2 years ago stats.) for instance, most stats only players use 0/67/22/11 or some variation of. I personally use 5/62/22/11.

http://www.ootpdevelopments.com/documentation/ootp11/index.php?page=game_options

robc · 08-14-2010, 12:06 AM

Quote:

Originally Posted by fspikec

You can customize the AI's evaluation and then make the ratings based of that evaluation. It can be weighted based on 4 different fields (Ratings, Current year stats, last year stats and 2 years ago stats.) for instance, most stats only players use 0/67/22/11 or some variation of. I personally use 5/62/22/11.

http://www.ootpdevelopments.com/documentation/ootp11/index.php?page=game_options

I believe someone posted that is used for trade and minor league promotion / demotion, but not lineups.

ShaneCarson · 08-14-2010, 02:40 AM

Quote:

Originally Posted by robc

I believe someone posted that is used for trade and minor league promotion / demotion, but not lineups.

Says right there in the manual...

Quote:

AI Options and Player Evaluation Options
AI options control how the game's artificial intelligence selects team lineups and evaluates players.

Zeno · 08-14-2010, 03:21 AM

Ah, thanks! I misunderstood those settings. I changed them up a bit and got the Phillies to use Madson as their closer. Great success! You're a trove.

robc · 08-14-2010, 08:42 AM

Quote:

Originally Posted by ShaneCarson

Says right there in the manual...

The manual may be wrong. Look at this post by SteveP: http://www.ootpdevelopments.com/boar...ml#post3019856

He is pretty knowledgable about the game. Also, see that thread for an experiment that seems to prove this.

ShaneCarson · 08-14-2010, 12:45 PM

Quote:

Originally Posted by robc

The manual may be wrong. Look at this post by SteveP: http://www.ootpdevelopments.com/boar...ml#post3019856

He is pretty knowledgable about the game. Also, see that thread for an experiment that seems to prove this.

Well that's not good. They might want to get that changed in the manual ASAP or fix it so it DOES have an effect on lineups.

SteveP · 08-14-2010, 12:46 PM

Quote:

Originally Posted by ShaneCarson

Says right there in the manual...

The AI options tab has two relevant components: (1) sabermetric vs traditional; and (2) the AI Player Evaluation weightings.

The first is related to lineups, as the text there says. The second has no effect on lineups. So the manual is correct, just a bit ambiguous in the wording.

ShaneCarson · 08-14-2010, 12:50 PM

That's incredibly ambiguous. I'm not sure anyone would pick up on that.

MizzouRah · 08-14-2010, 10:16 PM

Does anyone favor Sabermetric over Traditional?

Killing Time · 08-15-2010, 12:16 AM

Quote:

Originally Posted by shanecarson

well that's not good. They might want to get that changed in the manual asap or fix it so it does have an effect on lineups.

+1

MizzouRah · 12-30-2010, 05:40 PM

Here is something that makes me scratch my head..

I'm 15 games into the 2012 season and the hitting leader right now is Arizona's 22 year old CF Faust Pegoretti, who is hitting 0.480 so far this season, he has 24 hits in 50 ABs. (and great ratings)

A 28 year old OF who has never played for the big club comes back from the disabled list and they promote him to the majors and stick this kid back into AAA. The guy they promoted played mainly AA ball last year.

This move makes no sense to me at all..

First guy is the 0.480 hitter..

He's replaced for this guy?

PSUColonel · 12-30-2010, 07:22 PM

Try weighting the ratings at a zero. I think I have this years stats at 67( which gets prorated) and then last years at 30 and two years ago the rest. Works very well in my league

I use traditional but as I said, I weight ratings at zero. I'm still convinced they are taken into account a little even at this setting

These setting work well for me, but remember this also... Your manager and scout don't always have the same opinion of a player. I this slot when a player has bad ratings but good stats..I don't usually see it the other way around...unless a really good player is having a really bad year.

I also put scouting accuracy on very low

MizzouRah · 12-30-2010, 09:15 PM

In this case though.. Faust wins in both ratings AND stats, should be a no-brainer for the CPU?

Faust should be starting in RF for Arizona, because they don't even have Tyler starting, he is a bench player and the RF they have playing now isn't close to Faust.

I'm sure there is a reason the CPU AI chose to do this, I just wonder what it was?

GEP · 12-31-2010, 09:59 AM

Quote:

Originally Posted by MizzouRah

I'm sure there is a reason the CPU AI chose to do this, I just wonder what it was?

Is Tyler out of options, maybe?

MizzouRah · 12-31-2010, 01:27 PM

Quote:

Originally Posted by GEP

Is Tyler out of options, maybe?

Nope, he has 3 option years left.

I put Faust on my shortlist and will continue to follow his 2012 season.

Heck, I might even try to trade for this guy.

MizzouRah · 12-31-2010, 10:17 PM

As I move along, what has taken place is actually pretty cool.

The next day, Arizona options Tyler Graham to AAA and signs FA OF Vernon Wells.

Faust is only hitting .219 in AAA after 8 games and his replacement on the big club is hitting .449 so far.

Very early on, but it will be interesting to follow along.

08-13-2010, 11:05 PM	#1
Zeno Minors (Rookie Ball) Join Date: Apr 2010 Posts: 23	Beef With the AI's Logic Overall, I am very pleased with the game. It's a lot of fun and gets better every year, but I have an issue with the way the AI manages it's lineups. This has happened in a number of situations, but I will use the one that inspired me to post this on the forum. In this particular simulation, Ryan Madson was the closer for the Phillies in 2010. He compiled 49 saves in 51 attempts and finished the season with a 0.93 ERA and a 0.84 WHIP. A career year for even the most elite of closers. Off-season comes around and the Phillies make some moves. I always look at every roster on opening day to see any trades or signings I might have missed. To my surprise, Madson was no longer the closer but was the setup man. The closer was Michael Wuertz who ended 2010 with 16 saves, 4.21 ERA and 1.40 WHIP. The game relies too heavily on ratings. Even if Madson's abilities deteriorated, there is no way any organization in the history of baseball would not have him as their closer. The past success would instantly trump any performance deficiency. I'm not saying that the ability should not factor into a team's decision when there are borderline calls, but when a player has such a spectacular season, the teams should be forced to make ignorant decisions that any GM would make in real life.

12-30-2010, 07:22 PM	#13
PSUColonel Hall Of Famer Join Date: Feb 2002 Posts: 13,141	Try weighting the ratings at a zero. I think I have this years stats at 67( which gets prorated) and then last years at 30 and two years ago the rest. Works very well in my league I use traditional but as I said, I weight ratings at zero. I'm still convinced they are taken into account a little even at this setting These setting work well for me, but remember this also... Your manager and scout don't always have the same opinion of a player. I this slot when a player has bad ratings but good stats..I don't usually see it the other way around...unless a really good player is having a really bad year. I also put scouting accuracy on very low Last edited by PSUColonel; 12-30-2010 at 07:37 PM.

08-14-2010, 03:21 AM	#5
Zeno Minors (Rookie Ball) Join Date: Apr 2010 Posts: 23	Ah, thanks! I misunderstood those settings. I changed them up a bit and got the Phillies to use Madson as their closer. Great success! You're a trove.

08-14-2010, 12:50 PM	#9
ShaneCarson All Star Reserve Join Date: Oct 2007 Posts: 728	That's incredibly ambiguous. I'm not sure anyone would pick up on that.

08-14-2010, 10:16 PM	#10
MizzouRah Hall Of Famer Join Date: Dec 2001 Location: Troy, Mo Posts: 6,266	Does anyone favor Sabermetric over Traditional?

12-30-2010, 05:40 PM	#12
MizzouRah Hall Of Famer Join Date: Dec 2001 Location: Troy, Mo Posts: 6,266	Here is something that makes me scratch my head.. I'm 15 games into the 2012 season and the hitting leader right now is Arizona's 22 year old CF Faust Pegoretti, who is hitting 0.480 so far this season, he has 24 hits in 50 ABs. (and great ratings) A 28 year old OF who has never played for the big club comes back from the disabled list and they promote him to the majors and stick this kid back into AAA. The guy they promoted played mainly AA ball last year. This move makes no sense to me at all.. First guy is the 0.480 hitter.. He's replaced for this guy?

12-30-2010, 09:15 PM	#14
MizzouRah Hall Of Famer Join Date: Dec 2001 Location: Troy, Mo Posts: 6,266	In this case though.. Faust wins in both ratings AND stats, should be a no-brainer for the CPU? Faust should be starting in RF for Arizona, because they don't even have Tyler starting, he is a bench player and the RF they have playing now isn't close to Faust. I'm sure there is a reason the CPU AI chose to do this, I just wonder what it was?

12-31-2010, 10:17 PM	#17
MizzouRah Hall Of Famer Join Date: Dec 2001 Location: Troy, Mo Posts: 6,266	As I move along, what has taken place is actually pretty cool. The next day, Arizona options Tyler Graham to AAA and signs FA OF Vernon Wells. Faust is only hitting .219 in AAA after 8 games and his replacement on the big club is hitting .449 so far. Very early on, but it will be interesting to follow along.