abneo3sierra Posted February 9, 2011 Author Share Posted February 9, 2011 ... Lots of testing would need to be done to prove this one way or the other. I agree...it has also occurred to myself that even the 20 tests I have done myself, in the big scheme of things can be a small sample,as it is just as likely if I do 20 more,the results would even back out again. I have mostly tested in my own scenario as I already know the starting positions, AI plans,etc..had not really tested on another scenario yet, but good work on that one. And I am still unsure if I am reporting it as a bug anyway...I have seen many games where the AI "cheats" to help itself out against a thinking human opponent,and on some level, I would not mind it here,my post was mostly intended to state my observations, and inquire if anyone has A-seen similar behavior, and/or B-knows for sure if the AI really does have a built in advantage.. for certain in playing entire scenarios, the AI does a good job, but generally gets hammered, so if it has an advantage built into it, that without it, it would get hammered harder, I am fine with it...just would like to know.if no matter what I do as player, "AI Alan" can see me...then my tactics as a player will change to adapt to that,and maybe to my advantage..as it is ALWAYS an advantage if you can predict an opponents' move, even if that move itself may hurt you. So I will continue testing and looking at this...something that to be honest, until a couple of days ago, I had not even thought about enough to test. 0 Quote Link to comment Share on other sites More sharing options...
permanent666 Posted February 9, 2011 Share Posted February 9, 2011 ok i did a quick test. uncons (player) are hiding in the building. temperatur is cold for the termals civilian density is none - after 24 minutes in game they are still alive and undetected. did the same test with unhidden uncons with an target arc (i guess ai ambush works this way) - same result after 24 minutes in game the ai did not spot them in my opinion spotting in buildings is not modelled very well but i don't think that the ai has an advantage over the player 0 Quote Link to comment Share on other sites More sharing options...
abneo3sierra Posted February 9, 2011 Author Share Posted February 9, 2011 LOL... I'd really,REALLY hate to be them when they look out the window though...talk about a welcome party... 0 Quote Link to comment Share on other sites More sharing options...
Lanzfeld Posted February 9, 2011 Share Posted February 9, 2011 The ending of Butch Cassidy and The Sundance Kid! Thanks for the test. 0 Quote Link to comment Share on other sites More sharing options...
LongLeftFlank Posted February 10, 2011 Share Posted February 10, 2011 More like the penultimate scene from Blues Brothers. Cue elevator music.... 0 Quote Link to comment Share on other sites More sharing options...
vincere Posted February 10, 2011 Share Posted February 10, 2011 Matrix. "Need more me". Nice test- picture says enough really. 0 Quote Link to comment Share on other sites More sharing options...
Sakai007 Posted February 10, 2011 Share Posted February 10, 2011 They really need to implement my idea of having special textures when infantry is spotted by, then fired on by the main gun of a modern battle tank. These textures, when used properly, will show a growing brown stain in the pants of said infantry, to simulate what happens BESIDES being suppressed and/or killed. This test would also be a great test of the functioning of those new textures, cause that is exactly what would happen to me if I were the bad guys in that house! 0 Quote Link to comment Share on other sites More sharing options...
Erwin Posted February 10, 2011 Share Posted February 10, 2011 It would be a more convincing test if you set up a situation where the AI COULD spot the enemy, and THEN retry with human-human and see if the human-controlled unit could or could not spot. Your current test could simply be too "hard" for the AI under any condition. 0 Quote Link to comment Share on other sites More sharing options...
abneo3sierra Posted February 11, 2011 Author Share Posted February 11, 2011 It would be a more convincing test if you set up a situation where the AI COULD spot the enemy, and THEN retry with human-human and see if the human-controlled unit could or could not spot. Your current test could simply be too "hard" for the AI under any condition. This is my stance also regarding this. The test here is great, but not any more convincing to me than my 24 now that would seem to contradict it, at least in specific situations. Ron 0 Quote Link to comment Share on other sites More sharing options...
Erwin Posted February 11, 2011 Share Posted February 11, 2011 One needs to find out the max distance it takes an AI unit to spot an enemy in (say) 60 seconds. Then, do the same test with exactly the same situation/parameters, but this time the spotting unit is human-controlled. If the human-controlled unit spots within the same time limit, only then one can claim objectively that the AI has no spotting advantage. 0 Quote Link to comment Share on other sites More sharing options...
dan/california Posted February 12, 2011 Share Posted February 12, 2011 Identical teams walking towards each other in the dark perhaps, red vs red, and no thermals? 0 Quote Link to comment Share on other sites More sharing options...
Erwin Posted February 12, 2011 Share Posted February 12, 2011 Might work. 0 Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.