Is the A.I. really that good?

[LorenzoDC]

Re: Is the A.I. really that good?

I played a game last night, and it was annoying, the computer was making a game back, I had 5 or blocks because the CPU loves to cheese inside crap, so what happens? two of them I blocked go flying out of bounds, the other 3 blocks, went right back to the CPU and they scored.

It seems iike when ever I get a block in this game 90% of the time it goes out of bounds so the cpu gets the ball back or they get the lose ball from the block.

Its even more annoying, when they get 3 or 4 offense rebounds in a row, when I am using Dennis Rodman and he can't get a rebound to save his life when the CPU wants to score.

Then at the end of the game, the CPU is down by 6 they score to make it a 4 point game with 3 secs left. So I go to pass the ball in and what happens? The player he passes it to, botches the ball and it goes off his leg and to the CPU who drains a three to get it down to one.

I call a TO to get it to half court, and luckily was able to run out the clock.

But it was rediclious how the CPU can make up a 6 point lead in 2 seconds but cheesing the user to botch a inbounds pass just so they can hit a three with someone in their face.

I think the only way to minimize this stuff is with sliders.

It seems like there are three ways the CPU builds momentum: players getting hot, when the user POE is not optimized for what the CPU is running, and when the underlying game engine does its dynamic changes based on the game score or just based on how long its been since the CPU went on a run.

If you get all those same forces running at once, it will effectively raise CPU performance sliders under the hood by, if I had to guess, 20 or more points, while possibly gimping the same user sliders by the same amount.

If you adjust CPU sliders down in some targeted areas just a bit, and maybe raise user in a few targeted places, I've been able to dull the degree of craziness in this stuff so that the game still feel like I'm competing or can win, even though I can still lose. I'm talking about just a couple of points either way.

But if you leave the sliders on default simulation, the game will get silly overpowered for the CPU sometimes. I play pro custom with variations off pro simulation, but I admit I'm not the best player around, I'm sure. The key is to develop the right slider balance for you.

The game will still do things that just make me laugh, how stupid and improbable they are. But not as bad, and not as often, and not for as long a stretch.

[birthday_massacre]

Re: Is the A.I. really that good?

I think the only way to minimize this stuff is with sliders.

It seems like there are three ways the CPU builds momentum: players getting hot, when the user POE is not optimized for what the CPU is running, and when the underlying game engine does its dynamic changes based on the game score or just based on how long its been since the CPU went on a run.

If you get all those same forces running at once, it will effectively raise CPU performance sliders under the hood by, if I had to guess, 20 or more points, while possibly gimping the same user sliders by the same amount.

If you adjust CPU sliders down in some targeted areas just a bit, and maybe raise user in a few targeted places, I've been able to dull the degree of craziness in this stuff so that the game still feel like I'm competing or can win, even though I can still lose. I'm talking about just a couple of points either way.

But if you leave the sliders on default simulation, the game will get silly overpowered for the CPU sometimes. I play pro custom with variations off pro simulation, but I admit I'm not the best player around, I'm sure. The key is to develop the right slider balance for you.

The game will still do things that just make me laugh, how stupid and improbable they are. But not as bad, and not as often, and not for as long a stretch.

yeah I turned down the CPU 3PT rating because they were hitting 3PT at a 50% clip. I had two games in a row where the CPU was down and hit 3 after 3 while being doubled teamed to come back and win. So I moved the slider from 50 to 35 which seems to have evened it out. Now the CPU never shoots more than 40% which is about right.

I just think its funny I kept my 3PT rating at 50 and usually shoot around 35% tops. I can be wide open and get perfect rating and perfect shot selection with Jordan or Kerr and still miss if its near the end of the game and the CPU is down.

The best thing I ever did was turn down the CPU playing passing lanes. I turned that way down because the CPU would pick off every single pass if I tried pass from the top of the key to anyone that was near the basket even near the 3 pt line.

They still do pick them off somme of the time, but its not 90% of the time like it was when it was set to 50.

Im using the 95-96 bulls in my season mode and I am 7-3. I had to start over a few seasons because the CPU would keep cheating to win. With the defaults all at 50 i started 0-4 in one season, and 1-3 in another.

I just got tired of being up by 4 to 10 points in the 4th only to start shooting 35-40% and the CPU shooting 70% in the 4th, and cheesing to victory.

They still do it a little at my current settings but its not nearly as bad.

[23]

Re: Is the A.I. really that good?

HOF basketball is not smarter AI or whatever. All of those people claiming the game was too easy probably caused this to be formulated. I believe the CPU makes less mistakes and you have a smaller window of making them because they take advantage of it.

Anyway, just thought that should be said again because Czar has mentioned this many times before but people still complain. I personally don't care to play on it.