Re: Fast Break AI
Well, I just saw that specific video differently. If you'd like to put together a package of many plays happening like this I'd be more than happy to analyze them and give my thoughts. My responses to your specific points are in BOLD above, by the way.
I respectfully disagree. The PG took a bump from the defender, and didn't decide to trail. There's nothing in his actions that looks like trailing anyway. Who trails by standing still with his back to the half of the court where the play is?
It's not a "trail", it's a "release". He's standing still anticipating that Porzingis is not going to lead the break in traffic but rather pass it back to him (the PG) to setup the offense or run a secondary break with the ball in the PG's hands. The AI is not telepathic and will make decisions. In this case the PG was anticipating a pass back as he is the more skilled player to bring the ball up the court.
In this case I dumped the ball to KP because I knew by reflex that keeping the ball in the hands of the guy in traffic who had a defender right on top of him would mean, in 2K17, a very high probability of the ball getting bumped loose again. It's not sim to give the ball to the big man like that but it's necessary in 2k in that situation to dump to an open man to avoid another bump animation turnover.
Well, that was your decision and all, but I think the PG in that case should have just held the ball and moved up the right side of the floor with it.
KP is actually a pretty good passer, and pass catching animations, especially in stride, just haven't been there much for users in this game pretty much ever. The animation for the pass catch is typical of practically all user breaks. I could agree with you maybe if that were not the case, but it is. It's one of those things that slows user breaks down that doesn't really happen to the CPU.
It's not that KP isn't a good passer, it's the angle from which you threw the pass. It was awkward and not a good angle. You're expecting the receiver in this play to almost act like an NFL receiver and catch the ball over his shoulder in stride, but the angle you threw the pass from was not a good one to throw to a player with their back turned to begin with.
As for the defending SF contesting the shot, can you seriously say a player with 60-70 ratings in speed and acceleration can literally go the full length of the court from the corner, behind the foul line steal, to contest the shot under the hoop, blazing past everyone else on the court? At the same time, the 80-90+ offensive SG on the other side of the court, foul line extended even with the steal, can barely get to a lane to fill on the weak side of the break, either to the corner or toward the basket?
This player happened to sprint. In a real NBA fast break not everyone sprints and all players in general do not move at maximum speed or at the same speed. Again, the AI is going to make decisions like they would similar to real life. It's not always going to be pretty or do what you think it should (since it's reflecting decision-making that you have no control of and is ratings-based.)
I don't know what NBA games you watch where that happens, but I haven't seen it. The reason he got there is because the fast break logic assigned him to guard that (his) man, and forced his speed to put him on that man.
How do you know this? I don't agree with that assessment.
#14 in that play is a center who, if you look at the beginning of the clip, was cold with a blue circle. He was cold because he had already turned the ball over twice with his ham hands in this very early part of the first quarter. Passing to him on the move to the hoop would have been a very unwise decision.
But you're right that the shot I took was actually an awful shot and should not have gone in. I should have kicked the ball back out to KP on the wing for an open shot at 3pts.
Passing the ball to a big man who is hustling/running/filling the lane and is essentially wide open is not often a "very unwise decision". Any big playing in the NBA could catch that pass and likely finish uncontested. Of course, making a quick decision like that in the heat of the game takes skills, and I'm not judging your gaming skills or anything like that based on one clip. I'm just saying that a quick pass to the big who is filling the lane literally perfectly would have been the best option and likely resulted in an easy 2 points.
Doesn't matter. The speed nerfs and boosts and AI glitches are so typical of transition plays that my (bad) choice to take that baseline jumper doesn't negate what's actually happening on the court.
It's not a "trail", it's a "release". He's standing still anticipating that Porzingis is not going to lead the break in traffic but rather pass it back to him (the PG) to setup the offense or run a secondary break with the ball in the PG's hands. The AI is not telepathic and will make decisions. In this case the PG was anticipating a pass back as he is the more skilled player to bring the ball up the court.
In this case I dumped the ball to KP because I knew by reflex that keeping the ball in the hands of the guy in traffic who had a defender right on top of him would mean, in 2K17, a very high probability of the ball getting bumped loose again. It's not sim to give the ball to the big man like that but it's necessary in 2k in that situation to dump to an open man to avoid another bump animation turnover.
Well, that was your decision and all, but I think the PG in that case should have just held the ball and moved up the right side of the floor with it.
KP is actually a pretty good passer, and pass catching animations, especially in stride, just haven't been there much for users in this game pretty much ever. The animation for the pass catch is typical of practically all user breaks. I could agree with you maybe if that were not the case, but it is. It's one of those things that slows user breaks down that doesn't really happen to the CPU.
It's not that KP isn't a good passer, it's the angle from which you threw the pass. It was awkward and not a good angle. You're expecting the receiver in this play to almost act like an NFL receiver and catch the ball over his shoulder in stride, but the angle you threw the pass from was not a good one to throw to a player with their back turned to begin with.
As for the defending SF contesting the shot, can you seriously say a player with 60-70 ratings in speed and acceleration can literally go the full length of the court from the corner, behind the foul line steal, to contest the shot under the hoop, blazing past everyone else on the court? At the same time, the 80-90+ offensive SG on the other side of the court, foul line extended even with the steal, can barely get to a lane to fill on the weak side of the break, either to the corner or toward the basket?
This player happened to sprint. In a real NBA fast break not everyone sprints and all players in general do not move at maximum speed or at the same speed. Again, the AI is going to make decisions like they would similar to real life. It's not always going to be pretty or do what you think it should (since it's reflecting decision-making that you have no control of and is ratings-based.)
I don't know what NBA games you watch where that happens, but I haven't seen it. The reason he got there is because the fast break logic assigned him to guard that (his) man, and forced his speed to put him on that man.
How do you know this? I don't agree with that assessment.
#14 in that play is a center who, if you look at the beginning of the clip, was cold with a blue circle. He was cold because he had already turned the ball over twice with his ham hands in this very early part of the first quarter. Passing to him on the move to the hoop would have been a very unwise decision.
But you're right that the shot I took was actually an awful shot and should not have gone in. I should have kicked the ball back out to KP on the wing for an open shot at 3pts.
Passing the ball to a big man who is hustling/running/filling the lane and is essentially wide open is not often a "very unwise decision". Any big playing in the NBA could catch that pass and likely finish uncontested. Of course, making a quick decision like that in the heat of the game takes skills, and I'm not judging your gaming skills or anything like that based on one clip. I'm just saying that a quick pass to the big who is filling the lane literally perfectly would have been the best option and likely resulted in an easy 2 points.
Doesn't matter. The speed nerfs and boosts and AI glitches are so typical of transition plays that my (bad) choice to take that baseline jumper doesn't negate what's actually happening on the court.
Comment