Tempo-Freelance-RunPlays

[slimm44]

Quote: Originally Posted by Serra11

I have made the coach profiles shot locations test. However,I have not said that the shot locations valors affect which play is going to be called but which PLAYER is going to have more plays called. Take bigtreydawg's exemple. He has raised the "coach profile take 3" for th Jazz and more plays were called for Hornacek.Because Jeff has an high 3pt tendency and a touches valor between 50/70 (am I right bigtreydawg???). The cpu still called plays for Malone because his high touches valor but Jeff became more involved in the offense. Also,have you tried to test a game or a quarter with all coach profiles shot locations at 0??? I did it with the Thunder,Westbrook and KD both with 100 touches:many plays started with Westbrook that went to the corner and gave the ball to Perkins in the low post.With of course a shot of the Thunder center... Ah wait,one time with call plays at 0 and after with valors above 50.

Which play types were assigned to KD, Russ, and Perk? If Russ was running a lot of PNR, it's conceivable that the play could end with the ball in Perk's hands. It shouldn't, but it's conceivable. That's why I feel it's important to track who the play is called for and not where the play finishes. If the plays being called were low post plays for Perk, then that gives us a piece of information. I would have no idea what to do with that information offhand, but it would still be information.

[STLRams]

Quote: Originally Posted by slimm44

I am comfortable that the following statements are conclusive: 1. Drive the Lane affects off-ball moves ONLY and influences the player to cut to the rim in FL situations, FB situations, and after setting a pick in a PNR. If this tendency is at 100, the player will roll/cut to the rim nearly every possession WITHOUT the ball. This tendency does not effect players with the ball. Off the ball (not in PNR) usually occurs from the top of the key or one of the elbows. I believe that the Backdoor Cut slider effects cuts on the baseline, not this tendency. 2. Attack the Basket effects players with the ball ONLY. The higher the tendency, the greater the likelihood the player with the ball will try to drive to the basket in FL, FB, and PNR situations where they are the ball handler. When this is set at 100, even with the Coaching Profile Slider and In-Game Slider set at 0, a player will nearly always drive toward the rim and will usually only stop if they are stood up/lose dribble (by players with high strength and/or on-ball defense ratings) or if they have a high Pass Out Tendency. This tendency does not effect players without the ball. I WOULD NOT RECOMMEND THIS TENDENCY TO BE SET HIGHER THAN 50.

Interesting, so would you say that we should adjust all players drive the lane tendency to zero for better spacing, but also set all players attack the basket tendency to 50 and below?

[slimm44]

Quote: Originally Posted by STLRams

Interesting, so would you say that we should adjust all players drive the lane tendency to zero for better spacing, but also set all players attack the basket tendency to 50 and below?

I think ATB should be 50 or below, but I've only tested in increments of 25. Something in between 50 and 75 may be OK, but I haven't tested for that. I also don't think DTL should only be at 0. I've actually see acceptable results with that tendency being at 100, but the higher you go the more likely a player is to cut off the ball without being involved in a PNR. I don't think I've given any players more than a 75, though, because nobody rolls EVERY TIME in a PNR.

That being said, here are some more findings:

I am watching a CPU game with OKC/OKC on All Star. All FL Tendencies, DTL, ATB, Pass Out Tendencies are at 0. All rules are OFF except for shot clock so I only have to pay attention to the starters. I've given the following play types:

Westbrook PNR
Thabo 3PT
Durant MID
Serge Low Post
Perkins Low Post

1. Whenever the CPU gets somewhere between 6-8 seconds left on the shot clock, they will drive as far as they can to the bucket and use as many dribble moves as they can to get there. Keep in mind, all players have all dribble moves 0'd out except Simple Drive. They should not be doing these moves at all, which means the AI is hardcoded to make the player with the ball behave this way when this much time is left on the shot clock. I don't think there is ANYTHING we can do to effect this behavior once the shot clock gets into that range. My hope is that it is not quite as drastic on PRO. If it is based on difficulty, though, it may do it more on higher levels. This is not good.

2. If the CPU calls a play, attempts to run the play, and the play is executed but there is no shot taken at the desired location, the CPU either goes into crazy drive to the bucket mode or stupid offense that makes no sense with players running all over the court/into each other until someone is either open or chucks a shot at the end of the shot clock. Again, this is not good.

Both of these findings are not good. I will post results that reflect how the Coaching Profiles effect play calling another day, possibly tomorrow.

[bigtreydawg]

This game is amazing,

I'm going to take this time and stress the absolute importance of floor spacing. It is absolutely imparitive that hot spots and freelance tendency are in harmony with each other. Once you do that you can really see what this game is capable of. It's pain staking trying to visualize what spots on the court Haslem is going to take when he is on the court with birdman, it's annoying imagining the best locations Wade should receive a pick from Chris and it's tedious implementing every edit with a wary eye on what other player it might effect but it is totally worth it. This game flows like whatever clever analogy you can come up with for something that flows really well.

That being said, drive the lane is an on ball tendency. Lol

It's totally an off ball tendency but it is for sure an on ball tendency too.

Stats first:

With drive the lane at 0

Both of the Heat teams were in the range of 46 to 58 points per game. Which I believe is the NBA average (gameplay suffered a little). Both team put up low FGA which is what you want from the Heat but it was ugly with not enough freethrow attempts and the shot clock getting real low. (This is for the team with tempo, fastbreak, and attack at 0). I was wrong when I stated that drive the lane at 0 led to more passing. That is just a result of good floor spacing.

With drive the lane at 100 there is still great passing but where you would see jab steps that led to nothing of consequence, the ball handler is on a mission to get to the rim. Rephrase. . .

He is on a mission to get past his defender. But with attack at 0 (tendencies) there is a barrier around the paint. This only seems to apply to driving to the hoop which coincides with the description of the slider.

Stats were what I expected with drive the lane at 100, more freethrow attempts and high points in the paint. But fastbreak stayed within the range of 14 to 18.

SLIMM, to me DTL being an on ball tendency is undeniable but I need you to see it too.

Something you can do to ensure that what I'm saying is valid is turn some of the more redicilous dribble moves up past 0.

With drive the lane at 0 you will rarely see Dwayne spin even if he has a tendency of 50 for spin but with DTL at 100 he is a out of control.

Something else you'll notice is three point shooter like Allen and Miller running off the line to try to get inside instead of just "spotting up"

Ill start working on the second test to break down the effect of coach profiles on play calling. If you still insist that DTL is only an off ball tendency than we might have to agree to disagree about that. But if you want to do more test on that,I'm game.

[slimm44]

Quote: Originally Posted by bigtreydawg

This game is amazing, I'm going to take this time and stress the absolute importance of floor spacing. It is absolutely imparitive that hot spots and freelance tendency are in harmony with each other. Once you do that you can really see what this game is capable of. It's pain staking trying to visualize what spots on the court Haslem is going to take when he is on the court with birdman, it's annoying imagining the best locations Wade should receive a pick from Chris and it's tedious implementing every edit with a wary eye on what other player it might effect but it is totally worth it. This game flows like whatever clever analogy you can come up with for something that flows really well. That being said, drive the lane is an on ball tendency. Lol It's totally an off ball tendency but it is for sure an on ball tendency too. Stats first: With drive the lane at 0 Both of the Heat teams were in the range of 46 to 58 points per game. Which I believe is the NBA average (gameplay suffered a little). Both team put up low FGA which is what you want from the Heat but it was ugly with not enough freethrow attempts and the shot clock getting real low. (This is for the team with tempo, fastbreak, and attack at 0). I was wrong when I stated that drive the lane at 0 led to more passing. That is just a result of good floor spacing. With drive the lane at 100 there is still great passing but where you would see jab steps that led to nothing of consequence, the ball handler is on a mission to get to the rim. Rephrase. . . He is on a mission to get past his defender. But with attack at 0 (tendencies) there is a barrier around the paint. This only seems to apply to driving to the hoop which coincides with the description of the slider. Stats were what I expected with drive the lane at 100, more freethrow attempts and high points in the paint. But fastbreak stayed within the range of 14 to 18. SLIMM, to me DTL being an on ball tendency is undeniable but I need you to see it too. Something you can do to ensure that what I'm saying is valid is turn some of the more redicilous dribble moves up past 0. With drive the lane at 0 you will rarely see Dwayne spin even if he has a tendency of 50 for spin but with DTL at 100 he is a out of control. Something else you'll notice is three point shooter like Allen and Miller running off the line to try to get inside instead of just "spotting up" Ill start working on the second test to break down the effect of coach profiles on play calling. If you still insist that DTL is only an off ball tendency than we might have to agree to disagree about that. But if you want to do more test on that,I'm game.

I'll do more testing with the DTL Tendency. However, to see what DTL does, I think it is better for testing purposes to have any other individual tendencies that could possibly effect driving/cutting to the rim at 0 because it isolates the tendency we want to see the effects of. I'll do this: I'll take the Heat and do the same test as I am doing with the Thunder, which is:

All Star, CPU/CPU, One Play Per Player, One Play Per Play Style in the Playbook, All Dribbling/Freelance/ATB/DTL Tendencies at 0, and I'll chart what I see on-ball and off. With all tendencies I mentioned at 0, when in a FL situation, 4 players stand still (2 at the 3PT elbows, 2 at mid-range baseline) and the Center goes from High Post to High Post as the two guards pass the ball back to each other. This is based off OKC's Hot Spots. Other teams may be different. With DTL at 100, I'll be able to see how each player behaves differently.

I am also in the process of testing the Coaching Profile Shot Location Tendencies using the same method. I'll begin posting results in about 20 minutes.

EDIT: Also, were you able to notice when in the shot clock or what situations the players were driving the lane? With the tendencies at 0 (and the ATB slider at 0), I ONLY saw this occur at the end of the shot clock.

[slimm44]

Here are results from two tests I ran with Miami.

Chalmers 20 Ins, 11 Clo, 9 Mid, 39 3PT, 57 touch, PNR play type
Wade 65 Ins, 46 Clo, 54 Mid, 10 3PT, 90 touch, Mid Range play type
Battier 7 Ins, 2 Clo, 1 Mid, 67 3PT, 30 touch, 3PT play type
James 74 Ins, 31 Clo, 45 Mid, 33 3PT, 96 touch, Low Post play type
Bosh 48 Ins, 27 Clo, 55 Mid, 9 3PT, 69 touch, High Post play type

For the first test, with all Coaching Slider Shot Location Sliders at 0, the AI called:

8 plays for Lebron
6 plays for Wade
4 plays for Bosh
0 plays for Chalmers/Battier

For the second test, with the 3PT Coaching SLider Tendency at 100, the Heat called:

4 plays for Lebron
5 plays for Wade
2 plays for Bosh
0 plays for Chalmers/Battier

I stopped the 2nd test early because I feel safe in assuming the CPU wasn't going to start calling a bunch of plays for Battier. He has the highest 3PT tendency on the team and his % breakdown is 99 for 3PT shots. If the AI playcalling was influenced by changing the Coaching Profile Shot Location Tendencies, I think there would have been at least one play called for Battier. That being said, there wasn't. Again, I am confident in saying that changing the Coaching Profile Shot Location Tendencies don't affect AI playcalling.

I also ran three tests with the same teams, one game with DTL at 100, one at 0, and the third game had DTL at 0 and ATB at 100.

With DTL and ATB at 0:
Out of 14 possessions, there was 1 backdoor cut, 3 drives before the shot clock got down to the 6-8 second range (all of which involved defenders being out of position, which makes me think this is also tied in to Offensive Awareness), 6 drives within the last 6-8 seconds of the shot clock, and the rest of the possessions were TO's or FB's.

With DTL at 100 and ATB at 0:
Out of 23 possessions, there were no backdoor cuts, 4 drives before the shot clock got down to the 6-8 second range (again, due to out of position defenders), 13 drives within the last 6-8 second range, and the rest were TO's or FB's.

With DTL at 0 and ATB at 100:
I only ran 4 possessions. One was a TO, one a FB, then there were 2 drives late in the shot clock. This one is not worth anything.

Out of the first two games, one with DTL at 0 and one at 100, there was no discernible difference in player behavior. What does this mean? There are other variables at play as there were no PNR's, Post Up's, etc. The only thing the AI should have been telling the players to do were stand still, go to a different spot, dribble, drive, pass, or shoot. They drove in the first 16-18 seconds of the shot clock roughly 10% of the time, which makes me want to test PNR to see if both ATB and DTL effect that aspect of the AI more.

Also, I am literally typing a "shorthand" game log to track all the action I can involving off-ball cuts and ball handler drives. This is tedious.

It's also frustrating.

[Serra11]

Quote: Originally Posted by slimm44

Which play types were assigned to KD, Russ, and Perk? If Russ was running a lot of PNR, it's conceivable that the play could end with the ball in Perk's hands. It shouldn't, but it's conceivable. That's why I feel it's important to track who the play is called for and not where the play finishes. If the plays being called were low post plays for Perk, then that gives us a piece of information. I would have no idea what to do with that information offhand, but it would still be information.

No one play was called in the situation I mentioned.
Russell just went to the corner and passed the ball to Perk in the post without calling anything.
It seemed more a freelance situation...invented by the cpu,since I had Perkins with low post freelance at 0 and 80 to set pick.

[Serra11]

Quote: Originally Posted by slimm44

Here are results from two tests I ran with Miami. Chalmers 20 Ins, 11 Clo, 9 Mid, 39 3PT, 57 touch, PNR play type Wade 65 Ins, 46 Clo, 54 Mid, 10 3PT, 90 touch, Mid Range play type Battier 7 Ins, 2 Clo, 1 Mid, 67 3PT, 30 touch, 3PT play type James 74 Ins, 31 Clo, 45 Mid, 33 3PT, 96 touch, Low Post play type Bosh 48 Ins, 27 Clo, 55 Mid, 9 3PT, 69 touch, High Post play type For the first test, with all Coaching Slider Shot Location Sliders at 0, the AI called: 8 plays for Lebron 6 plays for Wade 4 plays for Bosh 0 plays for Chalmers/Battier For the second test, with the 3PT Coaching SLider Tendency at 100, the Heat called: 4 plays for Lebron 5 plays for Wade 2 plays for Bosh 0 plays for Chalmers/Battier I stopped the 2nd test early because I feel safe in assuming the CPU wasn't going to start calling a bunch of plays for Battier. He has the highest 3PT tendency on the team and his % breakdown is 99 for 3PT shots. If the AI playcalling was influenced by changing the Coaching Profile Shot Location Tendencies, I think there would have been at least one play called for Battier. That being said, there wasn't. Again, I am confident in saying that changing the Coaching Profile Shot Location Tendencies don't affect AI playcalling.

But the AI always follows the player touches tendencies and if you have on the court,at the same time,two mega-stars,a first offensive option of an average team and a solid third offensive option...Battier has no chance to see a play for him(like in real life).