So, uh, how's the AI? - Page 2

84CubsFan · 02-26-2007, 12:24 AM

Bump this up to the top, because I care. And because other people are starting to ask about it again.

How is the gameplay in these latest builds? Are the computer GMs good enough for us to suspend our disbelief when unrealistic things happen?

The Wolf · 02-26-2007, 12:59 AM

Quote:

Originally Posted by 84CubsFan

Are the computer GMs good enough for us to suspend our disbelief when unrealistic things happen?

It would be unrealistic if some unrealistic things didn't happen.

PotatoPeeler · 02-26-2007, 02:45 AM

Quote:

Originally Posted by The Wolf

It would be unrealistic if some unrealistic things didn't happen.

i.e Alfonso Soriano, Carlos Lee, and Gil Meche.

Kemp · 02-26-2007, 03:57 AM

Does anyone know what kind of scouting the computer teams do? I know in 2006, the computer teams weren't scouting internationally but I was wondering how they'll be scouting the feeder leagues.

I'd hate to see the computer drafting horribly because they weren't scouting the feeder leagues properly. Also, do the computer teams have a good understanding of league equivalencies? For example, if you create 5 feeder leagues of varying levels of equivalencies, does the computer know to focus more on the higher equivalency leagues?

Markus Heinsohn · 02-26-2007, 04:05 AM

Quote:

Originally Posted by Kemp

Does anyone know what kind of scouting the computer teams do? I know in 2006, the computer teams weren't scouting internationally but I was wondering how they'll be scouting the feeder leagues.

I'd hate to see the computer drafting horribly because they weren't scouting the feeder leagues properly. Also, do the computer teams have a good understanding of league equivalencies? For example, if you create 5 feeder leagues of varying levels of equivalencies, does the computer know to focus more on the higher equivalency leagues?

All that matter to the CPU is the potential of players, which are not influenced by the feeder leagues MLEs. They only influence the current ratings.

Kemp · 02-26-2007, 04:08 AM

Quote:

Originally Posted by Markus Heinsohn

All that matter to the CPU is the potential of players, which are not influenced by the feeder leagues MLEs. They only influence the current ratings.

That's interesting, I didn't think that was the case. Do the Player Creation Modifiers impact potential ratings?

I'm just wondering how to emulate multiple feeder leagues of varying levels of (potential) ability.

rasnell · 02-26-2007, 05:46 AM

I add another factor to what I consider the AI -- statistical accuracy. Historic simmers are going to note how much better OOTP 2007 is when it comes to season and career leaderboards if you use the new feature of player recalc and auto league modifiers.

Because of the work on the play-by-play text, there are a lot more variables to make the outcomes more fun.

A lot of stat tracking and charting have been done by several beta testers, most notably RonCo, to make sure of aging, progression and even down to very specific issues of SB, WP, Home Runs Allowed by pitchers, etc.

I think roster management is much better, free agents keep reducing their demands over time until they're eventually signed so you don't have that awful list of unsigned superstars when the season begins, but there are still weaknesses that are being tested hard right now.

I still think it's too easy to outwit the AI during drafts. But you have a much greater challenge from the AI in the vastly improved trade system, especially if you use the settings of Hard, Favor Prospects.

Sign and release is still causing aggravations, but that is being focused on hard and there's four more weeks.

For me, the easiest way to judge whether the AI is different and tougher is this. I could have an inaugural draft and immediately build a World Series champion and then dominate for a decade.

I've never been able to do this in OOTP 2007. At best, it takes three years of significant moves, shuffling, free agent signing and good rookie drafts to have a shot at post-season. And it is very tough to dominate.

I think the biggest reason is because of my favorite feature, player recalc. Historic players really will have surprising twists and turns in their careers, just like real life. It's so good that you really are going to have to have your own house rules of not looking at the real stats in the player profile.

I like this kind of toughness.

02-26-2007, 05:46 AM	#27
rasnell Hall Of Famer Join Date: Jan 2003 Location: Frankfort, Kentucky Posts: 3,746	I add another factor to what I consider the AI -- statistical accuracy. Historic simmers are going to note how much better OOTP 2007 is when it comes to season and career leaderboards if you use the new feature of player recalc and auto league modifiers. Because of the work on the play-by-play text, there are a lot more variables to make the outcomes more fun. A lot of stat tracking and charting have been done by several beta testers, most notably RonCo, to make sure of aging, progression and even down to very specific issues of SB, WP, Home Runs Allowed by pitchers, etc. I think roster management is much better, free agents keep reducing their demands over time until they're eventually signed so you don't have that awful list of unsigned superstars when the season begins, but there are still weaknesses that are being tested hard right now. I still think it's too easy to outwit the AI during drafts. But you have a much greater challenge from the AI in the vastly improved trade system, especially if you use the settings of Hard, Favor Prospects. Sign and release is still causing aggravations, but that is being focused on hard and there's four more weeks. For me, the easiest way to judge whether the AI is different and tougher is this. I could have an inaugural draft and immediately build a World Series champion and then dominate for a decade. I've never been able to do this in OOTP 2007. At best, it takes three years of significant moves, shuffling, free agent signing and good rookie drafts to have a shot at post-season. And it is very tough to dominate. I think the biggest reason is because of my favorite feature, player recalc. Historic players really will have surprising twists and turns in their careers, just like real life. It's so good that you really are going to have to have your own house rules of not looking at the real stats in the player profile. I like this kind of toughness. __________________ Charlie Root won more games for the Cubs than any pitcher (201), yet was remembered for one pitch to Babe Ruth. Find out more about the 1929 World Series in my book, "Root for the Cubs: Charlie Root and the 1929 Chicago Cubs." See the web site at www.rootforthecubs.com. The book is at http://www.amazon.com/Root-Cubs-Char...t+for+the+cubs. Beta tester, OOTP 2007-2023 and iOOTP 2011-2014.

02-26-2007, 12:24 AM	#21
84CubsFan Major Leagues Join Date: Mar 2003 Posts: 376	Bump this up to the top, because I care. And because other people are starting to ask about it again. How is the gameplay in these latest builds? Are the computer GMs good enough for us to suspend our disbelief when unrealistic things happen?

02-26-2007, 03:57 AM	#24
Kemp All Star Starter Join Date: Feb 2002 Posts: 1,181	Does anyone know what kind of scouting the computer teams do? I know in 2006, the computer teams weren't scouting internationally but I was wondering how they'll be scouting the feeder leagues. I'd hate to see the computer drafting horribly because they weren't scouting the feeder leagues properly. Also, do the computer teams have a good understanding of league equivalencies? For example, if you create 5 feeder leagues of varying levels of equivalencies, does the computer know to focus more on the higher equivalency leagues?