RFF's M22 XP Sliders

[Aestis]

COMING SOON

In the spirit of Tdawg, these will be entirely data-driven.

I will be running a couple dozen sims at least and visualizing the results here. I have the ability to run 25-50 sims and (mostly) auto-populate the results in a BI tool in the same time he spent manually counting 2 or 3 sims. So as a result we can get a much larger dataset and really fine-tune how we're looking at the results.

The GOAL here is to keep the 10y sims as close to the File Start levels as possible. File start league talent is what dictates win/loss %s in game, etc. For example if your OL talent increases league-wide too much, pass rush becomes useless. If QB talent increases league-wide too much, passing game is out of whack. Slider sets that worked in season 1 won't work by the time you get to season 5+. That sort of thing.

More to come in the following days.

OVR TARGETS by POSITION

Spoiler

First, a look at what we're trying to accomplish. Minimum starter OVR is just one piece of the puzzle, but it's an important one.

Minimum Starter OVR Targets:
THESE ARE NOT SLIDER SETTINGS. THESE ARE WHAT WE'RE AIMING FOR IN TERMS OF STARTER TALENT LEVELS

[Aestis]

File Start League Talent Distribution

Below is our approximate goal distribution as this is how the roster file begins. I will get more into detail where I've intentionally departed a bit from this & why.

The reason staying close to this goal is important, as mentioned in OP, is this is how gameplay is balanced. As an example, if most starting OL are between 74-80 OVR and pass rush is already questionably slow, and then a few seasons later a typical OL is 78-84 OVR.... your pass rush is going to potentially be nonexistent without you intending for that to happen.

Note the dark bars on the left represent FILE START (preseason wk 1) and the lighter bars on the right represents after a 10 year sim (preseason wk 1) on DEFAULT settings (i.e. out of the box). This is an average of a few sims, to be exact, as there is sometimes meaningful variance from sim to sim. The issues will become apparent.

Also note the dev trait distribution includes Hidden as it's from a roster file export, not from manually counting in-game. This is important, of course, because rookies would not be included otherwise.


QUARTERBACK
You can see the issues we'll need to tackle:
- Ratings across the board go way up across the 10Y sim
- The driving factor for this ratings inflation is because look at how many more Star/SS/XF players are typically created. By year 10, you've got on average 20 QBs who are SS/XF dev. That's 2/3 of league starters!



HALFBACK
- Much better than QB, but some clear rating inflation in the 80-90 range.



FULLBACK
- Much lower stakes here tbh, but you can see there's a ton more JAGs in the 70+ range, but lacking an elite guy.



TIGHT END
- Pretty close to right. 80-90 needs a tiny bump without throwing everything else out of whack, if possible.



WIDE RECEIVER
- Probably the closest yet. High-end guys aren't being created quite often enough but we also don't want to risk inflating the 80-85 range too drastically or else the average starters will be too close to the elite guys.



OFFENSIVE LINE
- This is another major problem spot, like QB. I grouped OL together to make it clearer and because this probably reflects reality better, but the issue is across the board: the average starters are getting way too good, but not nearly enough 85+ players are being created due in large part to SSs being created at half the rate of where we begin, if not lower.
- You are hopefully starting to see why for spots like OL, a simple XP slider # won't fix it by itself: either your average starters will be WAY WAY too good on a higher XP slider setting, or you will have way too few strong/elite starters. This is what we're out to solve for.



DEFENSIVE END
- Pretty close. A bit of mid-range ratings inflation to address.



DEFENSIVE TACKLE
- Similar story to DE, fairly close. 75-79 is heavily inflated but top-end looks overall pretty good.



OUTSIDE LINEBACKER
- A lot of these guys will be pass rushers, esp at the high end. Overall, not bad except for the big inflation of lower-end starters (75-79).



MIDDLE LINEBACKER
- Major mid-range inflation here due to the huge # of Star dev players created, another potential problem spot. Interesting that the high end looks rough despite SS/XF generation rate seeming pretty spot-on.



CORNERBACK
- 80+ a little low here. In the ballpark though.



SAFETY
- A bit low. Don't mind inflating 70-74 a bit as those are below starter level, but want to avoid 80-89 becoming significantly inflated. Still, the position probably needs some XP love.

[kwpit79]

I'm grateful you're doing this. I did a quick ten year sim just to look at what the league looks like, and it actually looks pretty good at default xp sliders. I bet we'll only need minor tweaks this year.

[Steely McBeam]

I did some comprehensive sliders for Madden 21 and this is what my final version look liked. I have not published these previously. The point of these sliders was to try and get the 16th and 32nd best player at any position at a level that reflected "average starter" and "poor starter". This was to make bad starters play like bad starters and average players average. The downside of these sliders is that for some positions particularly TE, RT and OLB you did not get any/many elite 90+ OVR players. These sliders were the product of three 10 year sim efforts to 2049 in addition to taking a franchise from 2020 to 2049 three times.


I think it may be possible to address the number of elite players by upping the number of X-factor, Superstar and Star players from the default but did not ever explore this as my sliders were already functioning well enough for me before it became an option.



I love Madden and across Madden 18, 19, 20 and 21 I have probably taking the Steelers from year 1 to year 30 fifteen times. And I can say this with absolute certainty that thus far there are no set and forget XP Sliders that ultimately do not alter the minimum or a maximum OVR rating in a way that doesn't break the gameplay sliders if you have the balance set for difficult play at year 1 at All-Pro level.


For me the typical breakage comes with O-line ratings. For example for the Madden 22 release the RT position:
OVR 85+ 7 players
OVR 80-84 3 players
OVR 75-79 8 players
OVR 70-74 10 players
OVR 68-69 4 players



The best RT is 89, the 16th tackle is 76 and the 32nd is rated 68. Previously it has not been possible to set sliders that as the initial rosters age and new players emerge to elevate new players to the 85+ range without raising the worst starters to the mid 70s and then suddenly what should be "bad starters" play like solid mid-level players.



I am planning to do set-up the same style XP sliders for this Madden but it takes time.
Step 1 Is get my game play sliders working reasonably.
Step 2 For each position assume what rating an average starter should be and what the worst starter should be.

Step 3 Do the 10 year sims to the end and play with the XP sliders to get the 16th and 32nd players* close to the level I want.
Step 4 Play through manually and check every 5 years to see where I am at.


* For QB, HB, FB, TE, T, G, C, DE, FS, SS, K and P each only has one starter so 16th player average and 32nd
* For MLB and DT I assume that 16 teams are 3-4 and 16 are 4-3 meaning the league has 48 starting MLBs and 48 starting DTs.
* For WR and CB as two take the field minimum in most formations the league's worst starter is ranked 64.


NOTE THE BELOW ARE MADDEN 21 SLIDERS as a an example.

	Year	2025	2030	2035	2040
QB	95			100
HB	110	125	140	150
FB	90
WR	90		95		100
TE	110	115	120	125	130
T	90
G	90
C	95		90
DE	95
DT	105
MLB	80			85
OLB	70		75
CB	110		115
FS	100	110		115
SS	115	120	130	137
K	110	115	130		135
P	110	115	130		135

[Aestis]

Quote: Originally Posted by Steely McBeam

I did some comprehensive sliders for Madden 21 and this is what my final version look liked. I have not published these previously. The point of these sliders was to try and get the 16th and 32nd best player at any position at a level that reflected "average starter" and "poor starter". This was to make bad starters play like bad starters and average players average. The downside of these sliders is that for some positions particularly TE, RT and OLB you did not get any/many elite 90+ OVR players. These sliders were the product of three 10 year sim efforts to 2049 in addition to taking a franchise from 2020 to 2049 three times. I think it may be possible to address the number of elite players by upping the number of X-factor, Superstar and Star players from the default but did not ever explore this as my sliders were already functioning well enough for me before it became an option. I love Madden and across Madden 18, 19, 20 and 21 I have probably taking the Steelers from year 1 to year 30 fifteen times. And I can say this with absolute certainty that thus far there are no set and forget XP Sliders that ultimately do not alter the minimum or a maximum OVR rating in a way that doesn't break the gameplay sliders if you have the balance set for difficult play at year 1 at All-Pro level. For me the typical breakage comes with O-line ratings. For example for the Madden 22 release the RT position: OVR 85+ 7 players OVR 80-84 3 players OVR 75-79 8 players OVR 70-74 10 players OVR 68-69 4 players The best RT is 89, the 16th tackle is 76 and the 32nd is rated 68. Previously it has not been possible to set sliders that as the initial rosters age and new players emerge to elevate new players to the 85+ range without raising the worst starters to the mid 70s and then suddenly what should be "bad starters" play like solid mid-level players. I am planning to do set-up the same style XP sliders for this Madden but it takes time. Step 1 Is get my game play sliders working reasonably. Step 2 For each position assume what rating an average starter should be and what the worst starter should be. Step 3 Do the 10 year sims to the end and play with the XP sliders to get the 16th and 32nd players* close to the level I want. Step 4 Play through manually and check every 5 years to see where I am at. * For QB, HB, FB, TE, T, G, C, DE, FS, SS, K and P each only has one starter so 16th player average and 32nd * For MLB and DT I assume that 16 teams are 3-4 and 16 are 4-3 meaning the league has 48 starting MLBs and 48 starting DTs. * For WR and CB as two take the field minimum in most formations the league's worst starter is ranked 64. NOTE THE BELOW ARE MADDEN 21 SLIDERS as a an example. Year 2025 2030 2035 2040 QB 95 100 HB 110 125 140 150 FB 90 WR 90 95 100 TE 110 115 120 125 130 T 90 G 90 C 95 90 DE 95 DT 105 MLB 80 85 OLB 70 75 CB 110 115 FS 100 110 115 SS 115 120 130 137 K 110 115 130 135 P 110 115 130 135

Will respond at more length when I get some time but you're absolutely right that there will be no fire & forget set of sliders to solve all XP issues. Stars/Superstars/XFs are generated at very diff rates as leagues progress vs file start and at very diff rates from one position to the next.

So for people who REALLY want league talent to progress at a rate that closely matches file start, there will be some doing. I've run a couple dozen sims already and will post some detailed visualizations of the talent distribution in the coming week or so. Testing out diff levels of XF/SS/Star to affect said distribution as XP sliders alone won't do it.

[Steely McBeam]

While my similar sliders method tries to mimic the ratings of the original roster to preserve the gameplay sliders working as expected I would not be afraid to abandon them if you think average and poor starters ratings are too low/too high.


For example in my finalised Madden 21 ratings the 24th MLB was OVR 76 and the 48th was OVR 71 in the original roster.


If half the teams run a 3-4 then the worst 3-4 MLBs will have an OVR 71 and the was no good for pass coverage. I therefore work with the XP sliders to lift the rating for MLB so the worst starters typically had an OVR of 74.


I think doing similar things for DEs is also helpful in creating pass rush which making sure the O-line ratings stay reasonably low.


And while this is a new Madden I have seen these tendencies repeat for several interations. Don't get me started about how Madden often rightfully looks like a massive patch but paid patch and roster update.

[schnaidt1]

Current gen or next gen?

[Aestis]

Quote: Originally Posted by schnaidt1

Current gen or next gen?

I'll be entirely focused on nextgen. Not sure if there are any differences in progression/sim stats.