RFF's M21 32-man CFM USER v USER Sliders

[ranta21]

Hey RFF, on those graphs: under the net yards / dropback, you had the total snaps graph twice. Could you post the original intended graph or some hot new stuff?

[Aestis]

Apologies for the long-overdue update (OP updated as well, as always changes from last ver higlighted in yellow):


- A few weeks ago I made the tough choice to shorten games slightly. I loved being exactly on target with NFL snaps, but games were running a bit longer than I like, too close to 1:30 on a regular basis. Dropping the min play clock from 20 to 16 brings that mark back down closer to 1:10-1:15 which is my personal target. Snaps are still around 58-59 per team. Will help a bit 'cosmetically' with offense being too OP, but of course this is strictly a cosmetic benefit.

- Discovered a rain kicking glitch that seems to be a function of slider + a kicker's accuracy rating. At our setting of 30, although it's objectively better, any kicker with accuracy in the ~70s was shanking kicks completely sideways 100% of the time, even with perfect accuracy. Tested at 40 and the shanking still happens for players roughly below 70 accuracy. I don't love this change because at 40, our FG% is a bit high. I will think of a long-term solution but in the meantime it was too glitchy to have a meaningful % of starting kickers completely unable to kick FGs or XPs in the rain.


That is it for changes.

Current state of gameplay:
- Run game is maybe a hair too high still at 0 RBK, but it's pretty close to target.
- Passing game is still clearly too powerful, but a lot of that isn't a "throw by throw" problem and rather due to AI just breaking entirely on some deep coverages leading to long TDs that should not be happening, thus inflating yardage + scoring. If it weren't for those busted coverages, I think passing metrics would at least be getting close.
- 3rd Down % remains a tad higher than NFL which is one really good benchmark for how effective sliders are. We're around 41% and target is 39%. Doesn't sound like a big difference but it adds up.


The patch that supposedly boosted pass rushing doesn't seem to have moved the needle too much. It didn't actually modify win rates which is an issue on the edge especially, just made the (rare) wins happen faster. Will give that one another 1-2 weeks of data to get a better sample size, but our passing efficiency looks pretty unfazed after the first ~12 games.

Ultimately to really get gameplay in a more realistic place, we will need a patch that cleans up some of the ZCV issues and gives edge rush a desperately-needed boost.

[xthecloudx]

I appreciate everything yall do. Our league is looking to adopt your sliders. Really appreciate your hard work.

[Aestis]

Run game still high after several weeks at 0 RBK & 60 TKL. TKL getting a bump.

Meanwhile pass game is still way too high, as always. Accuracy coming down a bit more.




Ranta, this is the Net Yds per Dropback metric that is a great per-snap measurement of the overall passing game (takes into account sacks & negative yardage from sacks).

[Aestis]

Another update in the persistent battle against offense in M21:

- Completion % has actually crept back UP despite Avg Depth of Completion downfield being as high as ever (and despite sack rate remaining below NFL avg). So I have gradually been dropping QB Acc. When depth of completion is so high, completion % should be lower than NFL not at/higher. Plenty of room to drop Accuracy here given the 64-65% completion rates we're seeing in the middle of our league, and >70% at the top end.

- Similarly the run game remains a tad too high. On its own this isn't gigantic, but combined with passing being too good, our 3rd down conversion remains higher than NFL. Too easy to sustain drives. Tackle getting one more small bump.

- Some penalty adjustments. Last year each team averaged 2.5 penalties just off OFF Holding + False Start alone, the two most common penalties by a wide margin. We are running about 10% fewer snaps, so call that around 2.2 would be the NFL target here. Instead, our league even at our higher penalty sliders is only seeing 1.2-1.5 penalties per team TOTAL. So, these two are getting a meaningful bump with the secondary goal being just a few extra penalties sprinkled around to stall additional drives here & there.

- Meanwhile, Facemask penalties only occur once every 7 games per team, on avg. I'm not entirely sure but feels a bit more frequent here. Going off feel so only making a tiny adjustment, again with the secondary benefit of removing the occasional facemask that advances an otherwise stalled offensive drive.

OP updated, changes in yellow:

[Carr4Oak]

Quote: Originally Posted by Aestis

Another update in the persistent battle against offense in M21: - Completion % has actually crept back UP despite Avg Depth of Completion downfield being as high as ever (and despite sack rate remaining below NFL avg). So I have gradually been dropping QB Acc. When depth of completion is so high, completion % should be lower than NFL not at/higher. Plenty of room to drop Accuracy here given the 64-65% completion rates we're seeing in the middle of our league, and >70% at the top end. - Similarly the run game remains a tad too high. On its own this isn't gigantic, but combined with passing being too good, our 3rd down conversion remains higher than NFL. Too easy to sustain drives. Tackle getting one more small bump. - Some penalty adjustments. Last year each team averaged 2.5 penalties just off OFF Holding + False Start alone, the two most common penalties by a wide margin. We are running about 10% fewer snaps, so call that around 2.2 would be the NFL target here. Instead, our league even at our higher penalty sliders is only seeing 1.2-1.5 penalties per team TOTAL. So, these two are getting a meaningful bump with the secondary goal being just a few extra penalties sprinkled around to stall additional drives here & there. - Meanwhile, Facemask penalties only occur once every 7 games per team, on avg. I'm not entirely sure but feels a bit more frequent here. Going off feel so only making a tiny adjustment, again with the secondary benefit of removing the occasional facemask that advances an otherwise stalled offensive drive. OP updated, changes in yellow:

So would these work on All-Pro difficulty in a 32 man league? Or would I need to make adjustments.

[Aestis]

Quote: Originally Posted by Carr4Oak

So would these work on All-Pro difficulty in a 32 man league? Or would I need to make adjustments.

Depends what you mean by "work." Obviously AP is going to lead to easier offense than AM.

[Lord_Belding]

I've been loosely following your league and your slider posts for the past 3 years. Although I haven't had the time to play in a CFM myself since Madden 12, your dedication to achieving realism in an H2H CFM seems unparalleled and I find it very interesting.

But one thing that jumps out at me is that the interception numbers seem very inflated to the point that the interception slider should be lowered significantly. Possibly to 5 or even 0.

-In 2019, the median number of team interceptions thrown was 12. 30/32 teams threw 17 or more in your most recent season.
-In 2019, the 2nd highest number of team interceptions thrown was 21. James Winston threw 30 for Tampa Bay which was among the most of all time. 28/32 teams in your league threw more than 21 and 22/32 threw 30 or more, something that only 12 NFL quarterbacks have ever done since the NFL went to a 16 game season in 1978.

I understand wanting the players in your league to make good reads, but at some point maybe it's unreasonable to expect Madden players to be as capable of this as an NFL QB is if nearly the entire league is outpacing the NFL average by a very wide margin. The number of interceptions just seems wildly, wildly unsim. Linebackers especially intercept very few passes in real life compared to in Madden.

Additionally, according to Pro Football Focus, defenses drop about 45% of their interception chances. I'm guessing this doesn't even include instances where the defender fully extends and it bounces off his fingertips or where he turns around at the very last second and the ball hits him in the hands before he has any real chance to track it. Madden does a poor job at recreating these kind of plays where the defender is there to make the play but doesn't quite see the ball in time. Instead, defenders seem to have way too large of a radius that results in either an INT or dropped INT animation. The only way to compensate for this might be to drop the INT slider and deal with seeing a lot dropped INTs even though the optics are bad.

I still think what you do is awesome, I just wanted to get that out there.