|
|
@ bigtrey, I sent you a PM with these findings regarding the Pull Up Tendency, but I feel they're pretty conclusive (aside from a few variables). At least conclusive enough to post.
I think I have the Pull Up Tendency figured out. First, the data:
I watched CPU matches of Phil/Phil. Starting lineups were Holiday, Richardson, Turner, Young, Hawes.
The tendencies were as follows for every test:
Holiday 41 ins, 35 close, 50 mid, 32 3PT
Richardson 28 ins, 22 close, 28 mid, 71 3pt
Turner 34 ins, 40 close, 47 mid, 16 3pt
Young 70 ins, 40 close, 33 mid, 1 3pt
Hawes 45 ins, 42 close, 45 mid, 14 3pt
What were their DTL/Spot, PU/Pen, ATT/PO ratings?
Sliders were all 50 (unless specified) on pro mode. Coaching sliders? What were they at during the testing?
First test: ATB 0, PU 0, PO 0
Out of 20 possessions, there was 1 pull up jumper in mid-range before the last 5 seconds of the shot clock
Second test: ATB 0, PU 100, PO 0
Out of 20 possessions, there were 2 pull up jumpers in mid-range before the last 5 seconds of the shot clock.
Third test: ATB 100, PU 100, PO 100
Out of 20 possessions, there were 5 pull up jumpers in mid-range before the last 5 seconds of the shot clock.
Fourth test: ATB 100, PU 100, PO 0, OBD slider 0
Out of 20 possessions, there were 10 pull up jumpers in mid-range and one from 3PT land before the last 5 seconds of the shot clock.
Fifth test: ATB 100, PU 100, PO 0 OBD slider 50
Out of 20 possessions, there were 13 pull up jumpers in mid-range before the last 5 seconds of the shot clock.
1. Pull Up only works when the player has an attack the basket tendency above 0. I don't know if there are any "levels" but I know that, at 0, there were no pull up jumpers until the end of the shot clock. With ATB at 100 and PU at 100, there are pull ups galore. This combination allows the AI to VASTLY override the mid-range tendency given to a player. Yes PU only works when ATB is > 0, cause ATB tells the player whether to attack (try to score = shoot = lay = dunk) or PO at the end of the 'drive'. Drive/Spot -> Pull Up/Penetrate -> Attack/Pass.
2. Pull Up also works better when the Pass Out Tendency is low. I only tested 0's and 100's, but with PO at 100 there were a select few PU's but with PO at 0, there were SEVERAL. It works better with a low PO tendency cause a low PO to ATT ratio means more ATT at the end of the drive. More ATT means more shots, lays, dunks.
3. Having a high ATB tendency, a high PU tendency, and a low PO tendency is the best combination to get players to pull up off the drive. Drive/Spot -> Pull Up/Penetrate -> Attack/Pass.
4. I saw no PU's by PF or C. Only the PG, SG, and SF attempted a PU. For that matter, I saw no drives AT ALL by the PF or C, and that's with the ATB tendency at 100 for all players. Would depend on the player Drive/Spot, PU/Penetrate ratings.
5. The on-ball defense slider does not effect how often a player pulls up. With the slider at 0 and the slider at 50 I got the same results.
6. All but one PU jumper I saw was in mid-range. The other was a 3. This animation is triggered primarily in Mid-Range. Probably has to do with how 2k programmed 'drives' and their starting places. Too close to the 3 line imo, needs to be in between Half and 3, that way you could replicate pull up Js from 3 off the drive. As it is right now, PU only happen in mids cause the initial drive begins at the 3 line. Thats why it needs to be moved back imo again.
7. The two players who took the most PU's were Turner and Richardson, both of whom have very low mid-range tendencies. Turner is at 47, JRich at 28 but with a 71 3pt, initial drive start location and other factors could play into why JRich is pulling (tend locations not being exactly as the image of the court)
8. This testing was done with no ISO plays and the ISO FL tendency at 0. I don't know if the results would apply to those situations. I would think ISO would be the best test for DTL/Spot, Pull/Pen, Att/PO and TT/TTS.
9. There were very few PU jumpers in FB action, regardless of tendencies or sliders. Most of all (if not all) came in FL or during set plays.
10. Some drives were initiated with no pick and some were with a pick. PU jumpers, with high ATB/PU and low PO Tendencies occurred during each situation.
Conclusion: On PRO mode, having a high DTL Tendency (to initiate the drive sequence), a low PO Tendency (a lot of PO's happen at the begging of the drive, this may change if OBD slider is lower or once we find threshold for ball handling/OBD rating), and a high PU Tendency (which is primarily triggered in the Mid-Range, between the FT line and 3pt line). Bolded what tells me its an On-ball rating, one cannot initiate a drive without the ball.
This animation was not triggered one time in 100 possessions by the PF or C. It solely occurred when the PG/SG/SF had the ball. PU tendencies not high enough for the bigs?
There were 34 total PU jumpers in 100 possessions and only 1 of them were 3's. More info needed to tell whats really happening. What plays being run, more ratings/tends, etc.
I don't know if the "fancy shot" animations effect this or not. This was STRICTLY testing for PU animations, not fadeaways/floaters/stepbacks, etc. PU means to pull up for the shot (not just PU jumper), so any shot off the dribble is considered a PU (even leaners, floaters, stepbacks). When PU is triggered, the CPU then looks at shot type tends (leaners, fades, etc... which all have their weighted ratios amongst each other) to find out what to do in that situation.
My next test is going to be Triple Threat/Triple Threat Shot. I probably won't be able to post any findings until tomorrow, at best. How are you going to test TT/TTS? |
|