Landeye's CPU vs. CPU slider set (using editor)

[Landeye]

I've been working on a new slider set for a week now and just want to start a thread so I can have everything in one place and in case anyone would like to help test. My goal is to make as little changes as possible in order to get the CPU to play "smarter" or more realistic. I think at default gameplay sliders the game is fairly balanced as far as ratings go but it plays too arcade-like and the CPU is just plain dumb for lack of a better word.

I decided to just start from scratch since it had been a long time since I've even see how the game plays on default heisman. I've been testing a few different fields and have really liked the results so far, so I decided I might as well go ahead and post what I got so far. I expect to continue making changes to certain parts of the game and updating fairly often until they’re done.

Spoiler

I think if I post my thought process on this it might help make sense of some of the reasoning behind certain sliders or whatever. The way I think of it is there are levels to the sliders. You have sliders like OAFO/FI and PSFO/FI as well as PSSP (pass speed) that are at the top level. These sliders influence the game the most in a direct and overall way. For example, on the lower level you could lower QBA to get slower passes but that would make them more inaccurate as a side effect. Or you could lower OADB (rush defense) to make LBs react slower but that also influences pass defense as well as other things. However PSfo/fi is simply overall reaction speed and PSSP is simply overall pass speed. These settings don’t influence ratings directly to have an effect however differences in ratings will still show.

Next down would be penalty and gameplay sliders. Both of these will change what happens on the field but how they do that is different. In my opinion penalty sliders effect more overall gameplay and gameplay sliders affect ratings. For example say you think that the pass blocking is too good and looks fake. When you change holding up to around 80+ you will see the players behaving differently, all blockers will hand-fight more etc. Say you lower PBK to 45 and you might see a similar overall effect, however this would be because now all players pass blocking ratings have been reduced so rather than just keeping the defender stonewalled they are having to fight them off harder. So in this case I would rather keep pass blocking at default so matchups are not skewed towards the defense but have holding higher to get the realistic hand-fighting/block shedding and higher possibility of penalties which are few and far between. (on kind of a side note I’ve noticed the biggest impact from gameplay sliders happen around 50 in that the greatest change you will notice is going from 50 to 49 or 51.

So thinking about it this way I want to “fix” things that are catered towards human vs. cpu or that the cpu just does badly by working them out from the top level down until everything fits. While I’d rather emphasize profile/penalty sliders over gameplay sliders I still think that I’ll have to make adjustments to those as well but since it will affect blocking/passing matchups I’d rather do those last. I plan on getting a working set and then playing out a year in my dynasty and once I have some stats to look at I can make small adjustments to gameplay sliders to get those on point. Anyway sorry if this is all just me rambling like a crazy person but hopefully this makes sense.

I test by starting new dynasty and playing first game. All coaches set to have 0 skill points.

Heisman for offense/defense/user
fatigue off in Main Menu/On in Dynasty
HFA off in both
Ice kicker on in both
Threshold 50

PENALTY SLIDERS (you can just change these in game then save the profile and edit the rest)

Holding 51
Facemask 49
Offensive pass interference 51
Defensive pass interference 50
KR/PR Interference 50
Clipping 49
Intentional grounding 51


In order they appear in GOPT TAB

OACA (coverage awareness) 49
OAFA (Field goal accuracy) 47
OAPA (Punt accuracy) 70
OASA (Stiffarm) 0
OATA (Tackle) 48
OADB (Run Defense) 51
OARB (Run Blocking) 51
OAAC (QB accuracy) 47
OARC (WR catching) 48
OAAG (Agility) 50
OADK (Knock down pass) 58
OPCL (Clipping) 49
OAFL (Field goal power) 55
OPFM (Facemask) 49
OARN (Ball-carrier vision) 95
OPHO (Holding) 51
OFPO 255 (effects ball carrier, won’t see running backs “stuck in the mud” after handoff)
OAOP (Pass blocking) 51
OPOP (Offensive Pass Interference) 51
OUPP 0
OURP 0
OUSP 0
OASR (Stumble Recovery) 65
OABT (Break Tackle) 47
OAST (Strength) 44
OAFU (Fumble) 49
OAAW (Pass Coverage) 44

IN UINF TAB
PSfi 81
PSfo 62
PSSP (pass speed) 2

In UOPT Tab
OAFI 81
OAFO 64

COACH SETTINGS

When playing in a dynasty coaches that are generated by the CPU (which are really just copies of retired coaches) will have Def Strategy, Off Strategy, and Run Strategy at 0. With Off strategy it can only be 0-3 or it will corrupt the file. Def Strategy and Run strategy are 95 I think. I would recommend making sure you edit new coaches to have the default ratings. I used to change these all to 0 but I think default plays better.

[Landeye]

Since I'm bored at work I'll explain what I think these settings do.

(Edit: I'm going to update this post with new findings as I figure stuff out)

GOPT Tab

OACA (50 default) is pass coverage awareness. when it's turned up you will notice players playing zone coverage better. Players will move more in their zone and cover multiple people in the same zone better. For instance a CB will press the WR at the line and move back into his zone and if a receiver runs an out behind them they will back up to better cover both players. When it's turned up too high players will be psychic and cover players behind them without looking (you'll notice it with LB covering the middle mostly) I like 55 I think it makes zone coverage tighter without making defenders too aware.

OARN (50 default) is ball carrier vision. pretty straightforward. I noticed in my old slider set even with it turned up to say 157 you would still see players will low BCV hit the wrong hole and make bad decisions. but overall the vision for all players is better with it higher. Right now 77 seems fine but I'll probably adjust this as needed as I make other changes.

OCVW (0 or 1 default) as far as I can tell this affects WR's catching. This is one I changed from looking at other threads where people had it at 255. What i noticed is that receivers will move towards the ball slightly to make the catch. You'll see it the most on hitch routes when the receiver is waiting for the ball to get to them they will inch towards the ball (more so when the defender is close). They don't take steps but seem to kinda just slide forward barely. but it can make a difference.

UOPT Tab

OAFI/OAFO (defense 80/offense 65)
This is basically a circle or sphere of awareness surrounding the player in which once a player enters this area the other player will know they are there and react to them. However it doesn't seem to matter if they are looking at them or not. it's almost like a psychic field of awareness. Even with these at 1 players won't just run into each other like they're blind it feels like this is separate from players "seeing" the field.

Now I used to crank both of these up to 127 but what I noticed is that while it can make some things look/play better a lot of times it messes the players up more than helps them. The biggest reason I've decided to go to a lower number is for pass/run blocking. So normally when the ball is snapped you will notice the O-line get out of their stance and turn to the person they will block. Sometimes (especially vs. the blitz) you'll notice a lineman not blocking the person in front of them and instead going to block someone else which will cause another lineman to move over to block the original rusher. It always irked me that most of the time if they had just stayed in their original spot the first rusher would be blocked by the closest guy and the other lineman would either block someone near him or move to the middle waiting for someone to get beat. I used to think it was just linemen with low blocking stats "missing" assignments so I would raise the pass blocking slider to help it but it would still happen, but I'm now convinced it's because OAFO set to 65 is actually a little too high and lineman are just too aware of blitzers coming at them from everywhere. With it at a lower number you might notice CB blitzes working a little better since the Tackles won't see them blitzing at first until they get closer but by then they've started blocking someone else. Run blocking will be improved as well where you won't see so much of blockers running past people to block the person behind them.

UINF Tab

PSfi/PSfo (defense 80/ offense 65)
These I'm less sure about but someone else described them as basically making players react to things closer or farther away from them. For example if PSfi is set to a lower number the receiver will react to the long throw from farther away so you may see them abandon the route and stop to catch the underthrown ball rather than keep running downfield. I think it might also affect how quickly defensive linemen realize that the Rb doesn't have the ball on playaction.

One thing i noticed too is that it seems to affect how the QB hands off the ball to the RB. With this at 1 it's a smooth animation but with it at 255 the rb looks like he bumps into the QB and then is handed the ball. I think this is what causes the botched handoffs on some run plays where the QB just stands there afterwords until he gets hit. You'll rarely see it on default but I used to have these both at 255 and I would see that a lot.

(EDIT 11.21.16) Huge finding last night testing! So have you ever wondered why the O-line always gets such a big push at the line of scrimmage even if they're lower rated than the defender across from them? It's because PSfi and PSfo are reaction sliders. I've been testing them but have been setting them both to the same number so I never noticed. I changed PSfo to 63 and PSfi to 66 and I noticed that the D-line was slower to react to the ball being snapped then the offense. On default the PSfi is 15 points higher than PSfo and this is what causes the unnatural o-line push. I figure I'll either set these equal from now on or have PSfo 1 point lower than PSfi since the 0-line would technically know the snap count and have a little natural advantage.

Also I think I might be onto something with making qbs more inaccurate. Right now it's either completion, drop, swatted pass, or interception. Unless you have qb accuracy below 10 you won't really see missed passes (but under 10 the misses you do see are usually 15 yard overthrows or passes that go directly to a defender). I've been messing with CSLV in the GOPT Tab and I'm becoming more and more convinced it affects CPU AI (at least the QB). So I was wondering if I have this set higher and then turn qb accuracy down will the two settings work together to make qbs less accurate but not dumb as they can be with lower QBA. I put CSLV at 25 and then played around with QBA, when I had QBA set to 25 I saw some misses! Out routes being thrown too far and going out of bounds and one time a WR ran a long comeback route to the right sideline and the Qb threw it off target to his left. both QBs had accuracy ratings in the 80s and they were throwing good passes too. I didn't have time to continue messing with this last night but I think this could be huge.

[Landeye]

I would highly recommend anyone using the editor to try turning on CSLV and seeing what you get. It's set to 0 default and can go up to 127. If I had to guess I'd say this adjusts the differences between ratings kind of like threshold does for speed. I felt like at 127 if a defender with say 85 block shedding went up against a lineman with 84 pass blocking he would shed the block as if he was rated 15-20 points higher. Also at 127 It felt like the qbs were throwing with less throw power than normal, almost like it adjusts the scale of ratings so now their 84 throw power rating is less than it was before but the difference between 84 and 86 throw power would be a bigger gap now. It's definitely doing something I just can't put my finger on what exactly.

[return.specialist]

I will flirt adjustments to the CSLV field.

[Landeye]

Quote: Originally Posted by return.specialist

I will flirt adjustments to the CSLV field.

Thanks for the help! Seeing a good variety of accurate and inaccurate passes with CSLV at 15 and QBA at 45. I've already tried CSLV10/QBA40 but I'd like to keep QBA as close to 50 as possible for all the other effects QBA has besides pure passing. Think I might try CSLV17/QBA47 next. Also I dont know if you're using the other settings too but I changed some things in the first post and updated the second post with some new info.

[return.specialist]

Quote: Originally Posted by Landeye

Thanks for the help! Seeing a good variety of accurate and inaccurate passes with CSLV at 15 and QBA at 45. I've already tried CSLV10/QBA40 but I'd like to keep QBA as close to 50 as possible for all the other effects QBA has besides pure passing. Think I might try CSLV17/QBA47 next. Also I dont know if you're using the other settings too but I changed some things in the first post and updated the second post with some new info.

Unlike other avid users, I am unable to remain focused through an entire game in coach-mode. With that stated, I use a RTG OLB to limit human-decisional influence on situational play calling, offensive audibles, and QB behavior. The RTG is in the third season and was created using unadulterated default roster. All sliders remain at default settings and games are played on the Varsity setting. There is additional info contained within the spoiler that should be noted by the OP but may be of interest to others.

Spoiler

The OLB, his speed, accelerations, and block-shedding are in high-80s or low-90's; catching, catching in traffic, and jumping in high-70s or low-80s; all other attributes at 99 or very close. Barf Lewinsky, as he is known, stands 7' and weighs 395lbs and is the starting ROLB of Middle Tennessee State. The OLB's stature is inhuman because I have been experimenting with other fields in the GOPT tab of the RTG file itself in an attempt to elucidate increased fumbling rates and more accurately replicate suboptimal team play.

Likewise, I assigned the OLB to a Group-of-Five conference because the game recreates dominant team-play fairly accurately but, in my experience, suboptimal team-play is often recreated inaccurately. That is, poor teams are not merely overpowered by superior defenses and do not merely struggle to stop superior offenses, poor teams make ridiculously inaccurate football plays. For instance, a quick RB will sluggishly miss the fleeting hole created by his O-Line on every handoff or an adequate TE will catch a pass in the flat, wide open, turn up field and in the process of turning, step out of bounds for a gain of 1 (indeed these events are observable in vivo but not with such frequency or indifference on players' behalf).

In the GOPT I have OHTA (tackling), OHDB (shed blocks?), OHRC (run defense?), OHFS (first step?), and OHST (strength) jacked up to 199 or 255, and OAFU (AI fumble) at 127. Anyhow, I apologize for the verbose explanation. Just wanted you to know my preexisting adjustments.

I set CSLV to 127. I definitely noticed greater variability in passing errors. The QBs (Oregon St., FIU) still completed an appropriate number of passes but I noted more short and overthrown incompletions than is typical; fewer inane drops also. The QBs also seemed to make better decisions when making checkdowns.

Although our run defense has been suboptimal this season, I observed the running by opposing RBs and scrambles by opposing QBs to be more deliberate; as if they were running with a purpose or to a destination. I noted this also with WRs and, especially, with TEs on catches underneath zone coverage--as if they would fight through the whistle.

When O-linemen were in the position to make a block, say, in the open field, it appeared they at least would attempt to block instead of running obliviously and precisely linearly toward the endzone. Perhaps there were less suction blocks but I am unsure.

Lastly, I did have 18 and 16 solo tackles in the two games. Before implementing the CSLV adjustments, I was averaging 9-11 with a career-high of 13 earlier in the season. I forced 2 fumbles in one game on big hits (FIU) and had 4 sacks in the other (Oregon St.).

I hope my observations are helpful. I will play more games and provide additional reports.

[skaithius10]

i am trying these out tonight

[Landeye]

Quote: Originally Posted by skaithius10

i am trying these out tonight

I'm thinking having PSfi and PSfo so close together is making the pass rushing and run defense a little overwhelming. I've been testing PSfi at 69 and PSfo at 64 and OAFI at 68 with OAFO at 63. This way the defense has a little bigger awareness circle but reacts slower than the offense. I can see now why they had PSfi and OAFI set higher than the offensive counterparts but 15 points higher is too much.