Home

IBM's "Deep Blue" computer beat the chess champ in '97,but we can't have adaptive AI?

This is a discussion on IBM's "Deep Blue" computer beat the chess champ in '97,but we can't have adaptive AI? within the Madden NFL Last Gen forums.

Go Back   Operation Sports Forums > Football > Madden NFL Football > Madden NFL Last Gen
Madden NFL 22 Scouting Update Pushed to Mid-October - More Details Revealed
NBA 2K22 Review: The Best 2K Game in Years
FIFA 22 First Impressions
Poll: What's more important to you, when the time comes to purchase a game? (Click to vote)
Reply
 
Thread Tools
Old 04-05-2009, 04:02 PM   #49
MVP
 
OVR: 9
Join Date: Jan 2003
Re: IBM's "Deep Blue" computer beat the chess champ in '97,but we can't have adaptive

And then we'd all complain about RoboQB because ZOMG BY THE 3RD AND 4th QUARTER THE PLAYS THAT WORKED ALL GAME AREN'T WORKING ANYMORE.

Fact is, though, that you would be very unhappy if Big Blue were your videogame partner. For one, it required a human slave to move all of the pieces on the board. Two, it printed out movements. I suppose you wouldn't mind making your first move off the line of scrimmage, pausing the game, sending messages into the computer, giving it time to compute an answer, print out a sheet, which then you take, make the motion for the defense... pause, repeat, etc., until the end of the game?
Rebel10 is offline  
Reply With Quote
Old 04-05-2009, 04:03 PM   #50
Stop The GOAT Talk
 
SageInfinite's Arena
 
OVR: 40
Join Date: Jul 2002
Location: NJ
Posts: 11,879
Re: IBM's "Deep Blue" computer beat the chess champ in '97,but we can't have adaptive

Quote:
Originally Posted by Rebel10
And then we'd all complain about RoboQB because ZOMG BY THE 3RD AND 4th QUARTER THE PLAYS THAT WORKED ALL GAME AREN'T WORKING ANYMORE.

Fact is, though, that you would be very unhappy if Big Blue were your videogame partner. For one, it required a human slave to move all of the pieces on the board. Two, it printed out movements. I suppose you wouldn't mind making your first move off the line of scrimmage, pausing the game, sending messages into the computer, giving it time to compute an answer, print out a sheet, which then you take, make the motion for the defense... pause, repeat, etc., until the end of the game?
__________________
http://twitter.com/sageinfinite

SageInfinite Bandcamp
https://sageinfinite.bandcamp.com/album/interminable
https://soundcloud.com/sageinfinite
Hidden Language Video Feat. Grandmilly
https://www.youtube.com/watch?v=hf5nW4rGTFM

#makethemaddenforumgreatagain
SageInfinite is offline  
Reply With Quote
Old 04-05-2009, 04:07 PM   #51
Pro
 
OVR: 8
Join Date: Jul 2006
Re: IBM's "Deep Blue" computer beat the chess champ in '97,but we can't have adaptive

Quote:
Originally Posted by bang911
Realize that when you put quotes around a word it takes on a different meaning. What you described is exactly what I would call "thinking".

Wouldn't it be nice if Madden ran an algorithm that placed weight on yards needed and what play types are more likely to gain those yards?

How about when one of your players gains X number of yards, the CPU then calls defenses that assign him extra coverage?

The only way any computer can "think", is by using algorithms.
You do have a point, but I term thinking as spontaneous and has the possibility of making a mistake. A computer can only do what you tell it to do. But I know what you mean.

The way it can be weighted is debatable. 1st and 10 at your own 35... what play should I run, if I have not played them before and there is no calculations yet. This seems like I would take the bill walsh approach and script my first 5 plays (this is actually done in chess as well, the first 5 moves are kind of random then it begins to start using the long algorithms.) then make your calculating moves after that.
lint is offline  
Reply With Quote
Advertisements - Register to remove
Old 04-05-2009, 04:18 PM   #52
Rookie
 
OVR: 7
Join Date: Mar 2009
Re: IBM's "Deep Blue" computer beat the chess champ in '97,but we can't have adaptive

Quote:
Originally Posted by rhombic21
At the very least, it would be nice if the computer could, over the course of several games and not just one game, learn which plays are successful in certain situations and which plays are not. Every year what almost always happens is that the computer keeps calling terrible plays that a human player would, after a couple poor experiences, start avoiding. For instance, this year on NCAA, the computer will often call all out man to man blitzes against spread formations. These lead to easy touchdowns every time.
The game isn't any smarter than the programmers. I said earlier, you can really see how bad a game is made in superstar mode when you can't control everything. Multiple time in superstar mode they do what you said and it is an easy TD. My favor is the HB screen on 3rd and long.
ChampN252 is offline  
Reply With Quote
Old 04-05-2009, 04:40 PM   #53
MVP
 
OVR: 52
Join Date: May 2008
Blog Entries: 7
Re: IBM's "Deep Blue" computer beat the chess champ in '97,but we can't have adaptive

Quote:
Originally Posted by bang911
Wouldn't it be nice if Madden ran an algorithm that placed weight on yards needed and what play types are more likely to gain those yards?
It seems pretty clear that you folks are underestimating our current AI (that plus we obviously have bugs too). I smell a future blog to go into more detail wit AI.
__________________

Ian_Cummings is offline  
Reply With Quote
Old 04-05-2009, 04:48 PM   #54
Pro
 
rhombic21's Arena
 
OVR: 9
Join Date: Jul 2002
Re: IBM's "Deep Blue" computer beat the chess champ in '97,but we can't have adaptive

Quote:
Originally Posted by lint
You do have a point, but I term thinking as spontaneous and has the possibility of making a mistake. A computer can only do what you tell it to do. But I know what you mean.

The way it can be weighted is debatable. 1st and 10 at your own 35... what play should I run, if I have not played them before and there is no calculations yet. This seems like I would take the bill walsh approach and script my first 5 plays (this is actually done in chess as well, the first 5 moves are kind of random then it begins to start using the long algorithms.) then make your calculating moves after that.
One approach would be to create coach profiles. These profiles would then serve as a sort of "baseline" that would drive decisions when the AI lacks information about the user, and would alter the way that the computer interprets feedback/results.
rhombic21 is offline  
Reply With Quote
Old 04-05-2009, 04:57 PM   #55
Proctor, MN. Go Rails!
 
Vikes1's Arena
 
OVR: 23
Join Date: May 2008
Posts: 4,115
Re: IBM's "Deep Blue" computer beat the chess champ in '97,but we can't have adaptive

Quote:
Originally Posted by Ian_Cummings_EA
It seems pretty clear that you folks are underestimating our current AI (that plus we obviously have bugs too).
"Current" AI, Ian. As in Madden '10' AI?

Just wondering.
Vikes1 is offline  
Reply With Quote
Advertisements - Register to remove
Old 04-05-2009, 05:00 PM   #56
Pro
 
OVR: 10
Join Date: Jul 2006
Location: RI
Re: IBM's "Deep Blue" computer beat the chess champ in '97,but we can't have adaptive

Quote:
Originally Posted by Ian_Cummings_EA
It seems pretty clear that you folks are underestimating our current AI (that plus we obviously have bugs too). I smell a future blog to go into more detail wit AI.
Ian would love to read that blog. Cannot wait for that 1 to come.
__________________
Madden25 next gen cant come soon enough!!!
kmart2180 is offline  
Reply With Quote
Reply


« Previous Thread | Next Thread »

« Operation Sports Forums > Football > Madden NFL Football > Madden NFL Last Gen »


Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is On
Trackbacks are Off
Pingbacks are Off
Refbacks are Off



All times are GMT -4. The time now is 09:46 AM.

Top -