Coach Run/Pass Tendencies

[martywebb]

jello for president.

[ncaa13]

I appreciate your (jello) effort in solving this problem. I agree that sliding the slider toward what you want cpu to do correctly works in the sim engine. I thought the EA developer would provide an AH HA moment but nope as I tried his method last night and still way to few rushes.
I personally just want this to work IN GAME as much as in the SIM engine. If my opponent has a good Rb II don't want him getting 7 carries. The stats you showed on the first page of IN GAME attempts by Georgia Tech and Duke are what I want my CPU to replicate.
SO assuming the run/pass is final (I set cpu run/pass Defense near 50/50 regardless) do you believe the agg/con slider works in the same manner, move slider toward which you want?
Some findings i've found:"Defense" 50/50 makes CPU too aggressive with blitzes. 85/15 makes them too conservative sitting in 2 deep coverage most of game. 75/25 seems good balance of coverage and disguise blitzes.
"Offense" oh boy who knows. 50/50 seems to make cpu qb do those quick throw aways (which could be result of awarness or Grounding slider also) BUT looking at your pics most cpu team have 65/35 or 70/30 "which is conservative" something around there and this seems to work pretty well IN GAME vs cpu. Otherwise still working on this one. The EA guy from 2012 said it effects 4th down decisions, red zone decisions, throws down field so yeah... THOUGHTS? "I'M LOOKING FOR IN GAME RESULTS MORE THAN ANYTHING"

[jello1717]

Quote: Originally Posted by ncaa13

do you believe the agg/con slider works in the same manner, move slider toward which you want?

I would certainly guess that it works the same way as the run/pass but I have no clue. I haven't tested it as I wouldn't know what to look for. Run/pass is easy (look for rushing or passing) but w/ conservative/aggressive I wouldn't know what tangible things to look for.

[Art01]

Perhaps there is a difference between what happens in a SIM situation and in a PLAY NOW COACH game situation. My original post and subsequent analyses are all based upon PLAY NOW COACH games (CPU-CPU) and PLAY NOW COACH games (USER-CPU).

I conducted experiments with WSU, Air Force, Army, SMU, Alabama & Florida State. WSU and SMU are heavy pass, while Army and Air Force are heavy run.

For WSU versus Air Force (CPU-CPU), I had WSU's HC and OC PASS/RUN Slider set to minimum White. WSU passed the ball 33 times and ran 11 times. For Air Force's HC and OC, I had the PASS/RUN Slider set to maximum White. Air Force ran 39 times and passed 8 times.

For CPU-CPU WSU versus SMU, HC and OC PASS/RUN sliders were all set to minimum White. WSU passed 32 times and ran 8 times; SMU passed 52 times and ran 14 times.

The apparently opposite findings between Jello's data and my data are certainly interesting. If there is a difference in performance between PLAY NOW type games and SIM games, then the question might be - what other differences might there be?

[jello1717]

Quote: Originally Posted by Art01

Perhaps there is a difference between what happens in a SIM situation and in a PLAY NOW COACH game situation. My original post and subsequent analyses are all based upon PLAY NOW COACH games (CPU-CPU) and PLAY NOW COACH games (USER-CPU). I conducted experiments with WSU, Air Force, Army, SMU, Alabama & Florida State. WSU and SMU are heavy pass, while Army and Air Force are heavy run. For WSU versus Air Force (CPU-CPU), I had WSU's HC and OC PASS/RUN Slider set to minimum White. WSU passed the ball 33 times and ran 11 times. For Air Force's HC and OC, I had the PASS/RUN Slider set to maximum White. Air Force ran 39 times and passed 8 times. For CPU-CPU WSU versus SMU, HC and OC PASS/RUN sliders were all set to minimum White. WSU passed 32 times and ran 8 times; SMU passed 52 times and ran 14 times. The apparently opposite findings between Jello's data and my data are certainly interesting. If there is a difference in performance between PLAY NOW type games and SIM games, then the question might be - what other differences might there be?

My results in an earlier post included results from played games that support my claims too. They weren't "play now" games, but they were dynasty games that I did play.

Quote: Originally Posted by jello1717

As for in game ratios, I only have the games I played myself from this season. Here's the Dook game w/ their 65/35 pass heavy ratio: Spoiler where they ran 20 times and passed 42 times. Now, here is the GT game with their 30/70 ratio. Spoiler Here they ran the ball 35 times and only passed 15. Both of these games are examples of where the gameplay engine also uses the correct run/pass ratios.

Sent from my iPad using Tapatalk

[Art01]

Apparently there is no way to resolve this issue to everyone's satisfaction:

Per Jello's findings: Increasing the width of the White Slider to the Right increases the Pass Tendency.

Per my findings: Increasing the width of the White Slider to the Right increases the Run Tendency.

Both of us have game play data to support our findings.

I cannot offer any additional insights to these conflicting findings. Another possible explanation might be differences in Rosters or Profiles being used, but I don't think that either Roster or Profile differences should affect this issue.

If anyone has other thoughts or data-backed findings, please express them.

As for me, I am getting the results I expect so I don't plan to change my approach. I presume that Jello feels the same way. So, I'll leave it at that.

[jbrew2411]

I am sure I will get told that I don't know what I am talking about, but I think there are more factors that go into the amount a team runs or throws the ball. I have said it before and I will say it again here, I have been playing the NCAA series since NCAA 98. I have wasted more of my life playing this series every year since 1997 then I should have. I have played thousands of dynasty games so I feel like with that background I know this game in and out.


I have found that with the shorter "white bar" the CPU will try to run the ball to start the game. I think several factors go into if the CPU will continue running the ball throughout the game.


1) The overall flow of the game impacts how much the CPU runs the ball. If the CPU has success running the ball early then they will keep trying to run. If you are stopping the run then they will try passing a little more.


2) Kind of goes with #1 but the score board impacts the CPU run game. I have played against my fare share of option based teams. The one thing I have noticed is once they go down a few scores they turn into a "Air Raid" Mike Leach team. In real life teams like Georgia Tech will stay true to their offensive identity despite the score, but I think the game has always had this built-in catch up that triggers more throws.


3) Sliders can impact the amount the CPU runs. They can also impact #1 & #2 as well. I have tried just about every slider set posted in the slider section. Some favor the CPU run game while other don't. A older "Community Set" is the only set I used for multiple years. It had RBA at 45 and RBK at 40 or 50 I think. In the 6 seasons of using that slider set I never allowed a 100 yard rusher and every game played the same; I would go up big early and the CPU would throw the ball. I was playing as UGA, used the sliders from dynasty years 2019-2024, so I got to play Tech (still had Paul Johnson as HC with default coaching settings). In the 6 games against Tech they had only one game, 2019 season, where the QB didn't throw 25-40 times. The one season, 2019, was the only time Tech still tried running the ball in the 2nd half. That was the first game that season I allowed 100+ team rushing yards. That game was a close game allowing the CPU to stay true to the Tech offense.


If you want some data then go check out my dynasty thread. It is following the same dynasty I referenced above but I didn't start posting until after the 2025 season. At that time I switched my sliders around some more, but the real data is at the end of the 2026 season. That is when I decided to turn fatigue off. It seemed to give the CPU a jolt. I had games against Tech, A&M with Troy Calhoun as HC, and Nebraska with a 30/70 setting to end the 2026 season. The CPU stayed true to their settings by running the ball the entire game. In the 2027 season I played A&M again, this time I won, but they still stayed true to their Air Force playbook.


My conclusion is that there are a number of factors that can impact the setting to make the setting not work. I am not even sure how much difficulty level may further impact the coach settings. I also have never once used "Coaching Mode," I can only speak for human vs. CPU dynasty games as that is all I have ever played in any of the 17 editions of the game I have played.

[return.specialist]

It is worth noting that Playbooks contain fields ostensibly pertaining to situational calls (perhaps, 3rd down, 3rd & goal, long yardage, etc.) but also to frequency. Conceivably, fields could pertain to sim, game-play, or both, but likely influence AI logic.

See this link and this one too.