NOTICE: This league is using the BLEEDING EDGE game engine. For more information, click here.

League Forums

Main - Beta Chat

Re: [0.4.6] Version 0d5f5ac

By raymattison21
11/24/2020 9:25 pm
setherick wrote:
raymattison21 wrote:
I am thinking acceleration is more important than speed. Especially the shorter the distance the players are traveling.


This in no way explains the play I posted.


Something is slowing down receivers throughout the route. Whether it be bump or man or both I have seen way worse in terms of slowing a receiver. Even the stutter slows them too much.

But if you look a acceleration in general the rating is more valuable. Most of the highest catch % guys are quick as well. Low in ypc but those 20 ypc players have at least a 10% lower catch rate.

The play posted might have quick guys but the sacker had 83. Near elite for 276 pounds. But watching games from the 2031 season the QBs had so much time you could see the wr on fly end his route at 30 yards. Then came the throw...

These receivers now are not making it to the end of 15 yard routes before a sack comes. That why long passes are so weak. With the dump off lowered it ends with not many options for QBs.

Re: [0.4.6] Version 0d5f5ac

By setherick
11/24/2020 9:28 pm
The DL went further with an OL blocking them than WRs did with no one in bump-and-run...that continues to be my point.

Here is yet another one: https://mfn1.myfootballnow.com/watch/12294#2292490
Last edited at 11/24/2020 9:29 pm

Re: [0.4.6] Version 0d5f5ac

By raymattison21
11/24/2020 9:46 pm
I am not against other adjustments but getting receivers through their routes quicker would help this code. If QBs threw it in those scenarios and were more accurate when pressured that would help drive hurries up and sack down.

Qb stay too tight to the pocket to often. Maybe scramble could matter more in terms of possibly pulling it down for a run. Seattle’s qb ran for 300 plus yards but I would be ok if a guy had double that. Extended plays are more frequent but QBs don’t buy time well in these scenarios.

With the shorter passes we can find an ok rhythm but our sack rate is still in the lower half of league compared to the nfl. With only a few long pass all season.

Re: [0.4.6] Version 0d5f5ac

By TheAdmiral
11/25/2020 9:34 am
I think everyone agrees the following.

A. Pass blocking needs to be improved, in particular against smaller, lighter players
B. A Pass rushers stamina for a play should see him decellerate on an extended play
C. If the O line cannot hold the pocket, the QB should be moving away and looking for open space (to set his feet and throw) or throw on the run (reduced accuracy, higher int rate or thrown away)
D. The ball should be in the air before the receiver ends his route. This may mean adjustment to the 'throwing arc' so the ball has more hang time on deeper throws. If the ball goes higher and slower there should be more catches

As an aside, could we see receivers attempting to dive for catches or is that too difficult to animate.

To Summarise,
Pass block improvement, earlier release time on throws, on scramble drills QB abandons the pocket and looks for space to make a throw, Pass rushers decellerate quickly allowing scrambling QB's to evade them and have medium to deeper ***** in the air longer to allow the receiver to reach the end of his route.


ANy chance of hearing JDB's thoughts on this?

Re: [0.4.6] Version 0d5f5ac

By raymattison21
11/25/2020 10:37 am
setherick wrote:
The DL went further with an OL blocking them than WRs did with no one in bump-and-run...that continues to be my point.

Here is yet another one: https://mfn1.myfootballnow.com/watch/12294#2292490


Again the quicker disengagement’s has blockers with less acceleration at possible fault. They couldn’t pass block either.

I wonder how much these pass keys are helping either slow up receivers or that boost to pass rush is making larger discrepancies in out comes.

Maybe guys alignments Due to the key makes a difference?

Re: [0.4.6] Version 0d5f5ac

By setherick
11/25/2020 6:20 pm
raymattison21 wrote:
setherick wrote:
The DL went further with an OL blocking them than WRs did with no one in bump-and-run...that continues to be my point.

Here is yet another one: https://mfn1.myfootballnow.com/watch/12294#2292490


Again the quicker disengagement’s has blockers with less acceleration at possible fault. They couldn’t pass block either.

I wonder how much these pass keys are helping either slow up receivers or that boost to pass rush is making larger discrepancies in out comes.

Maybe guys alignments Due to the key makes a difference?



I think you are missing the forest for the trees here.

The biomechanics of the game are just off. At any AC, a WR should be able to move FARTHER in the same amount of time as an OL who is BEING BLOCKED for at least part of that time.

What this means is that the WR is moving SO SLOWLY off the line that the DL can disengage the block and get to the QB faster than the WR. Apparently, all DL should just play WR.

To understand just how absurd this is, check out the 20 Yard Shuttle times from the NFL combine: https://www.pro-football-reference.com/draft/2020-combine.htm

The fastest DL are >0.3 seconds SLOWER than the fastest skill players without a player in front of them.

Re: [0.4.6] Version 0d5f5ac

By raymattison21
11/25/2020 6:59 pm
Something is slowing them beyond speed acceleration weight algorithms. The dline is acting normal...there a lot of blow bys and QBs are slow looking but my faster dline make it there faster.

For receivers, I need better bump (possibly with strength) , route and the speed acceleration weight algorithm to get down field for those 7 to 12 yard routes . As if not thrown by then pressure is on it way. Forget the long passes that don’t end until 30 yards. You will rarely make it that far. The constant gainers are 5 -15 yards. It’s probably tied to the one and three second drops QBs are consistently getting.

Lower what slowing the receivers down. A fix for this...as some guys don’t even make it 5 yards in 3 seconds. It’s as if they slipped and are crawling through their route.

Re: [0.4.6] Version 0d5f5ac

By setherick
11/26/2020 12:10 pm
raymattison21 wrote:
Something is slowing them beyond speed acceleration weight algorithms. The dline is acting normal...there a lot of blow bys and QBs are slow looking but my faster dline make it there faster.


For the last time, the problem IS that the WRs are NOT accelerating. Can you just let JDB respond to my comment?

The bonus problem is that QBs are too slow in their reads, but when WRs aren't getting off the line before the QB gets sacked, it doesn't matter.

---

Updating this with an another observation that has gotten buried that I want to pull back up.

The reason that the YPA is low, and will be much lower in veteran leagues, is because M2M defense is still overpowered and zone defense is still largely useless. (This is probably what is throwing off any automated testing too since AI game plans are all RNG.)

I have been using the same game plan I put together to slow down all of the 4.5 exploit plays, and it is utterly dominant in beta.

---

Final observation I want to unbury. Most of the low YPA is because of the high school level depth on routes. This has been a problem for a long time, but it's more noticeable now that other things on the game has improved.

Let's take the drag route. In the game, the WR turns immediately and runs down the LOS at a depth of 1-2 yards. An NFL drag route is going to curve toward the 5 yard mark from the LOS.

The hook routes are also off depth and timing. The 2-3 yard quick hitch route in MFN is the appropriate depth. But the 4-5 yard curl ins should be extended out to 7-8 yards, and the WR should not drop back more than a yard from the depth because the ball either needs to be in the air or the WR needs to (1) post/stop if there is a hole, (2) run a go, or (3) run parallel to their spot. WRs should never "come back to the ball" since that pulls the defender back to the QB who is likely to scramble.

The other issue with the hook routes is that none of them are back shoulder which means you can't get the defender out of position, but that's a different story.

The dig and cross routes on some of the medium plays (see 113 FL Post) are also too shallow. That dig should happen at the 10 yard from the LOS mark between the LB and the safety since that is the gap in the field. The deep cross on that play should happen at the 20 yard from the LOS mark and force the safety to make a decision who to cover. If the safety comes up, the QB should go to the deep cross.

The flag and post plays aren't worth discussing since QBs don't anticipate their throws.
Last edited at 11/29/2020 9:57 am

Re: [0.4.6] Version 0d5f5ac

By setherick
12/01/2020 7:23 am
The Conf Championship game winners are 4.6 in a nutshell.

Pittsburgh - Rushing 6.47 YPC, Passing 4.14 YPA

Detroit - Rushing 8.03 YPC, Passing 6.28 YPA

Just like the NFL prior to 1992, there is no reason to even watch the tape.

Re: [0.4.6] Version 0d5f5ac

By raymattison21
12/01/2020 8:07 am
No run keys out of you two either. (Or zone calls) There’s really no point if you can get the right defensive plays in there (Ones without assignment or alignment flaws). It’s almost if the deeper alignment allows for better blitzing lanes to track ball carriers. You don’t need -4 to -7 yard losses ever heavy blitz. A 1 yards gain or loss is just as good to stifle momentum.