Breaking Down Sliders... How They Really Work

[Justin Clark]

Re: Breaking Down Sliders... How They Really Work

Pretty good stuff bill, very in depth. Question, will you have a look at Playmakers sliders and see what you think abou them and how they compare to your theory here?

[bill2451]

Re: Breaking Down Sliders... How They Really Work

first let me just say thanks for all the kind words everyone

now to g2thecore's comment, I have not screwed around with speed threshold... i had my old sliders at 65 because i thought that was a good balance for seperation for both fast and slow player (if the threshold is too low it gives fast ppl too much of an advantage while if it is too high it gives slower guys too much of an advantage because their slowness can not be taken advantage of too much)... a fast guy at 0 will kill, but a guy like Brandon Jacobs at 100 will kill also because normally if he breaks into the open field with his power he will be caught, but at 100 fast guys dont gain enough ground on him and he can take it too the house too easily (i think initially everyone assumed a 100 speed threshold reduced big plays... but imo it just shifts the type of player making the big play)

65 also seemed to mesh well with lineman speeds (guards being able to pull on outside runs, etc)

but overall, it was far less "scientific" and more "opinion" with my choice for speed threshold... so that is open to be played with, and at this point it could possibly (or probably lol) be changed to produce more realistic results

at this point though, I just figured i'd keep it something consistant while i was/am figuring out what everything does and how the sliders work together... trying to change that while doing everything else would have been far too hard to know what was actually effecting the gameplay

[g2thecore]

Re: Breaking Down Sliders... How They Really Work

Sounds good Bill...I'm using speed threshold of 30 with my current sliders, and I haven't really seen anything out of place speed wise. I'll bump it up to 65 as per your recommendation, and I'll see if I see any significant difference.

My thing is, I felt that on a higher number the fast wideouts, like Desean Jackson, weren't getting any real separation on linebackers in coverage. I want slower guys to get burned by speedy wide receivers, I hate to see a 260 lbs linebacker keep up with a faster guy. I also didn't like seeing the fast guys get caught by slow players on returns. I don't know, I'll play around with the settings and see if we can come up with a fair compromise.

When I get home tonight, I'll test the speeds with Hester. I'll try him vs a fast corner, and then against a slow LB and see what kind of separation he gets. I'll adjust from there. Thanks for your advice btw.

[bill2451]

Re: Breaking Down Sliders... How They Really Work

Pretty good stuff bill, very in depth. Question, will you have a look at Playmakers sliders and see what you think abou them and how they compare to your theory here?

some things seem pretty good, but some things seem unrealistic and simply in there to produce good "end game" stats

now that might be necessary to a degree because the cpu isnt good enough for even gameplay... but things like 100 reaction time for cpu do not make realistic gameplay

I like that he lowered QB acc and raised WR catching since I last checked (WR Catching could turn out to be a hugely important slider that we actually want up high)

both run and pass blocking also seem high

again i'm mostly working on theory here right now based purely on what the sliders actually effect... there will w/out a doubt have to be some sacrifices to realistic gameplay interactions made in order to produce more realistic results and make up for less than stellar AI... but I hope that those sacrifices will be less extreme if the "meat and potatoes" of the sliders, if set right (aka get blocking as realistic as possible, get CB/WR interactions to work better) will reduce the fudging of sliders in the AI's favor (ex: upping CPU reaction time because its too easy to pass, because the core interactions between the players just arnt right, so now the CPU DB's act unrealistically, but the stats are better)...

get the "core" right, and while I know fudging of realism will still have to occur, hopefully it can be reduced

right now playmakers sliders seem decent, but some things just seem quite a bit off, which in turn seems like he has to move other things around to make up for it

[bill2451]

Re: Breaking Down Sliders... How They Really Work

Sounds good Bill...I'm using speed threshold of 30 with my current sliders, and I haven't really seen anything out of place speed wise. I'll bump it up to 65 as per your recommendation, and I'll see if I see any significant difference.

My thing is, I felt that on a higher number the fast wideouts, like Desean Jackson, weren't getting any real separation on linebackers in coverage. I want slower guys to get burned by speedy wide receivers, I hate to see a 260 lbs linebacker keep up with a faster guy. I also didn't like seeing the fast guys get caught by slow players on returns. I don't know, I'll play around with the settings and see if we can come up with a fair compromise.

When I get home tonight, I'll test the speeds with Hester. I'll try him vs a fast corner, and then against a slow LB and see what kind of separation he gets. I'll adjust from there. Thanks for your advice btw.

again, to me 65 just felt right... it helps the offensive line actually be able to block outside, speed felt pretty good between fast and slow, etc... but i'm not going to lie, it was hardly scientific, and i prolly didnt play around with it enough to truly know what setting i liked most

hell, I may just put it to 50 to keep it as standard as possible when i actually attempt to make my slider set... either way the rules of how the sliders work will remain the same, the final numbers may just have to be changed slightly based on what threshold each person is using

hopefully in time this is something i'll get a chance to play around with more... its certainly a great slider... unfortunately it is also a very hard one to predict because it changes every aspect of every other part of the game and how they interact with each other

[g2thecore]

Re: Breaking Down Sliders... How They Really Work

again, to me 65 just felt right... it helps the offensive line actually be able to block outside, speed felt pretty good between fast and slow, etc... but i'm not going to lie, it was hardly scientific, and i prolly didnt play around with it enough to truly know what setting i liked most

hell, I may just put it to 50 to keep it as standard as possible when i actually attempt to make my slider set... either way the rules of how the sliders work will remain the same, the final numbers may just have to be changed slightly based on what threshold each person is using

hopefully in time this is something i'll get a chance to play around with more... its certainly a great slider... unfortunately it is also a very hard one to predict because it changes every aspect of every other part of the game and how they interact with each other

I completely agree with that last comment. It does affect every other aspect of gameplay, which is why I'm so adamant on getting it right. Now I know, I won't get anything "perfect", but since speed is an intricate part of the game of football, I want to get something as close to realistic as possible.

I'll be testing it further to see if I can come up with fairly accurate numbers. Hopefully, I can get a "scientific" breakdown like you did with the blocking. I also hope you'll be testing that as well down the line. I can't wait to see what your sliders look like.

[bill2451]

Re: Breaking Down Sliders... How They Really Work

I think one thing we'll see is that with blocking low (in both the pass and run games), it will just look/feel more like football because in order for an animation to be triggered the conditions have to be more exact... it will be much harder for an o-lineman to randomly engage a guy 3 yards away in a block... meaning the animation that is triggered will seem more natural with what is going on in the game

most slider sets i'm seeing out there now have in the 40-80 range with blocking sliders... and i'm thinking that will have to come down more to the 10-25 range to produce initial animations that fit with the flow of the game

the thing to be careful about is that go too low and the blockers are just pathetic and can't engage anyone well

[Justin Clark]

Re: Breaking Down Sliders... How They Really Work

I completely agree with that last comment. It does affect every other aspect of gameplay, which is why I'm so adamant on getting it right. Now I know, I won't get anything "perfect", but since speed is an intricate part of the game of football, I want to get something as close to realistic as possible.

I'll be testing it further to see if I can come up with fairly accurate numbers. Hopefully, I can get a "scientific" breakdown like you did with the blocking. I also hope you'll be testing that as well down the line. I can't wait to see what your sliders look like.

G2, will you post your slider work-up that you said you created from bill's theory? would love to see em...

[pbz06]

Re: Breaking Down Sliders... How They Really Work

are you using the same sliders for human and cpu?

I think you're going the wrong way with Run Block and Block Shed... remember, run block determines the blocking animation, with it that high you're going to see a lot more suction blocking on guys that shouldnt get blocked, and with Block Shed at 5 it will be much harder to break out of those blocks...

not only do I think that is way too easy for the human, but I think while it will help the CPU run game get realistic results (which is always a good thing), I'm not sure the actual gameplay will be realistic (aka rushing yards for the cpu may be realistic, but at the cost of unrealistic blocking)

I'd put Run blocking on the lower end of the spectrum (def for human, maybe closer to the middle for the CPU cause they do need some help)... I'm not sure how low mind you, but I think both blockings should go low to create more realistic blocking animations/scenarios... block shedding i'm guessing could go more near 50...

Pass Defense reaction time and WR catching have to have a bigger difference as well... after testing more last night, both work togther to create realistic seperation... With a lower reaction time the corners wont be able to run the WR's route as effectively as the WR (which is good), unless they are playing up on the receiver and the WR has low catching... as it turns out, the WR Catching slider effects more than the drop rate... it plays a huge part in figuring out how effective the CB are at jamming the WR at the line...

the problem with the Madden design is that generally speaking 4 things happen when the CB is playing up at the line... 1) the CB completely jams the WR... 2) the CB gets a hand up into the WR and the WR is slightly slowed down but he knocks the hand away and goes upfield... 3) Very similar to #2 except the animation is a little different. The CB never really gets his hand up and kind of backs off, which may not seem important but it is... 4) The CB tries to jam the WR and gets blown by

the problem is that in Madden momentum is a fickle thing, and a lot of routes are similar in this regard...

in scenario #1 both players are slowed down equally. This sucks as the qb because it messes up your timing and you have to wait for the WR, but it does not result in the CB "jumping" the route, because both players are going the same speed, and therefore when it comes time to make their cut, both players either cut unrealistically, or cut using animations that resemble the player having momentum and having to shift their weight

scenario #2 is the worst scenario... this original jam animation results in the WR going faster than the CB. This would seem like a good thing, except that when it is time for the WR to make his cut in his route, he's going fast enough for momentum to kick in. If he's doing a simple square route in, he has to take a couple of small steps to slow down, then plant his right foot and turn 90 degrees (a very good and realistic animation)... the problem is that the CB is now going slower than the WR, therefore he doesnt reach the threshold to trigger that same type of animation, and he does a "super turn" and is suddenly cutting off the WR and seemingly "knows the route"...

scenario #3 is the same as scenario #1... this time though, instead of both guys being slowed down the same amount, both guys dont really get slowed down at all... the point is that both the WR and CB are going the same speed (slow in this case) and both do the super turn... the turn is unrealistic, but the results are acceptable because both players did it

scenario #4 the CB gets burned... except if the WR is doing any kind of square route, the WR will again have to shift using momentum, while the CB wont, meaning one second the CB will be burned, the next he will seem "super human" and suddenly be in front of your guy... if the WR is running a post or something, his momentum change isnt as big because he's not turning 90 degrees... meaning the non momentum of the CB doesnt allow him to gain as much ground back or jump the route because he's too far behind

this lack of momentum is something that is absolutely horrid on Maddens part, and I think contributes HUGELY to the gameplay flaws... scenario #2 is the most common, which means that WR are using momentum most of the time while CB's are not, resulting in crazy secondary play... its a shame too, because it seems like it would be an easy fix, just lower the threshold that players have to use at least some kind of momentum when they turn, and viola, no more CB's turning on a dime when WR's have to shift momentum.

going back to my original point now, the WR Catching slider seems to have a big effect on how successful the WR is at not being jammed (aka it seems to control what types of animation is going to happen when the CB presses the WR)... it wont always be the same animation, but more times than not it will be animation #2... the key is to get the settings so that animation #3 is the most common, with the other 3 being sprinkled in... I havent fully tested what the lowest threshold to see this happen is, but I did test 80 briefly last night (and a pretty low Pass Reaction time, which also effects the jam success) and that seemed to be working

now this will adversely effect drops, but if it can produce more realistic secondary play on every down, i'll sacrifice a few dropped passes

and if we can figure out the point where momentum is being used for the CB's and WR's, you can then move reaction time up/down to get the type of seperation you are looking for

a low reaction time will mean the CB first sees the WR shifting momentum, the CB then "reacts" and starts to [hopefully] shift his momentum. The lower the setting, the longer it takes him to "react" while at 100 he's basically "reacting" at the exact same time as the WR... however if the CB isnt using momentum and just turning on a dime, his reaction time doesnt matter because he's still going to beat the WR to the spot.

Thanks for your input!

Yes, I like to keep the User/CPU sliders as close to possible. I usually focus on CPU first to get them to play right, and then adjust User if I have to.

Reason I had Run Block so high was because the CPU seems to need a lot of help with running. It seems too easy to stuff them on just about every play.....but the suction blocks would suck. I'll drop it down then and see how it plays out.

//edited the post with the sliders

[bill2451]

Re: Breaking Down Sliders... How They Really Work

I completely agree with that last comment. It does affect every other aspect of gameplay, which is why I'm so adamant on getting it right. Now I know, I won't get anything "perfect", but since speed is an intricate part of the game of football, I want to get something as close to realistic as possible.

I'll be testing it further to see if I can come up with fairly accurate numbers. Hopefully, I can get a "scientific" breakdown like you did with the blocking. I also hope you'll be testing that as well down the line. I can't wait to see what your sliders look like.

yeah I will... i'm thinking get base sliders down so that the blocking, etc is as good as possible. Then move the speed threshold and see how it changes, and then adjust accordingly (at that point, the adjustments shouldnt be major as you just have to figure out what to tweak for new speed differences)

[bill2451]

Re: Breaking Down Sliders... How They Really Work

Thanks for your input!

Yes, I like to keep the User/CPU sliders as close to possible. I usually focus on CPU first to get them to play right, and then adjust User if I have to.

Reason I had Run Block so high was because the CPU seems to need a lot of help with running. It seems too easy to stuff them on just about every play.....but the suction blocks would suck. I'll drop it down then and see how it plays out.

//edited the post with the sliders

there will definately have to be a little fudging of realistic blocking, etc, to make up for poor AI (like the run game)

however the amount of fudging needed when everything else functions as it should is yet to be seen... hopefully it wont be as extreme as is currently needed to produce realistic stats

[g2thecore]

Re: Breaking Down Sliders... How They Really Work

G2, will you post your slider work-up that you said you created from bill's theory? would love to see em...

Yeah, they're a work in progress...but I will definitely post them later. I'm at work right now, so I can't really post them anytime soon. Expect something around 9 PM EST.

[stmark]

Re: Breaking Down Sliders... How They Really Work

Based on your theories, would something like this be a good idea?

(bold means updated upon further discussion)

Are these for both human and cpu......because most sliders i see nothing its ever equal for both

i used them they are great......very realistic.....Redskins (me) 26 vs tampa 7

Leftwich 13-23 176yrds 4ints 0tds
ward 27 carries 40yrds
Graham 3car 7yrds 1td
askew 1carry 1yrd

Portis 20 carries 100yrds 1td
Campbell 12-33 169 0tds 1int
Campbell

i kicked 2 field goals had a pick 6

it was closer than the score indicates the pick 6 was at the end of the game so was the rushing td

[bill2451]

Re: Breaking Down Sliders... How They Really Work

ok heres an experiment I would love for someone to try... I'm currently at work, but w/out playing a game, and if I had to put Human and CPU sliders the same, I am going to post a set I would use based solely on how the sliders effect the game and theoretically it could play out to make a realistic game

note I have not tried these sliders at all, so am completely unsure of the effectiveness certain degrees of sliders will have (aka how much better is a 75 pass rush vs an 85 pass rush)... I am simply sitting at work and quickly came up with a set that makes sense given the context of the rules... whether or not I go too extreme, or not extreme enough, is unknown because there was no prior testing

I also have no idea how good you guys are at the game, but if someone could play a few games with the following set of sliders and post how it went, it would be awesome...

Passing
QB Acc - 12
Pass Block - 5
WR Catch - 80

Rushing
Broken Tackles - 50
Run Block - 20
Fumbles - 45

Pass Defense
Reaction Time - 15
INT - 5
Pass Rush - 85

Rush Defense
Reaction Time - 10
Blk Shed - 75
Tackling - 0


One rule, you can't play defense on the D-line (unless that guy is dropping into coverage on that particular play, which is rare to say the least). You must play as a LB or in the secondary.

Pay close attention to how the blocking interacts... I expect the computer run game to struggle w/ these settings, i'm really just interested on hearing how the blocking interacts (both on the line and at the 2nd level)... if you want to try and see a good cpu run game, tune down the human BLK Shed to like 20 and see what that does... if the run game is too good, its probably the Brk Tackles being a little too high.

also pay attention to how the corners cover the WR... do they still seem to know the WR route and run it before the WR does? or is it toned down a little more?

dont worry so much about results as much as realistic interactions... the results can be tuned later with changes if the interactions seem good

just keep in mind i have not tested these at all, just threw them together while sitting here at work... but it'd be awesome if someone wanted to test them out

[Dice]

Re: Breaking Down Sliders... How They Really Work

This is some great work. Would slider modifications remediate(or at least tone down) the passes to the flats issue? Thanks.