League Forums

For the 2048 season I'm rolling back the passing game changes (namely the zone and pass decision changes) and only including the updates that will actually be in 0.4.6. One key that I just pushed thanks to Tarquin's comments is an update in the pass blocking algorithm to fix a bug when the OL exceeds the rushers, which specifically caused issues with the 4-6 Heavy 2 Deep Man Under defensive play.

As always, please report thoughts as we go through the 2048 season, hopefully I can get some time in to fine tune the blocking issues further. The scope is basically just the pass blocking and punt blocking issues.

From the other thread...

"I'm keeping the relative speed changes"

I see this as a part of the code releated to Blocking only?

Where raising the value of zero speed did or did not make it through as well....or they are some how intertwined?

It's the scale of how much slower 0 speed is vs 100 speed. I reduced the difference early on in this version because it seems that a lot of issues are coming from that span being too great.

"For the 2048 season I'm rolling back the passing game changes (namely the zone and pass decision changes) and only including the updates that will actually be in 0.4.6"

Is the decrease to long pass drop rate going foward with this 4.6 ? Or is it out as well?

I think the long passing is about right. Teams are getting 4000-6000 yards per season. Which seems accurate. Do you really want a game where 8,000-12,000 yards a season is 'normal'

Drops are bad under 4.5. Beta rates had better drop numbers, but we had other passing tweaks not making into 4.6 which will change reults again.

Say drops are twice as high as they should be under 4.5 and old 4.6 was right on target. In general when compareing the two averages both have passing yards, completion rate, QB passer rating, and pretty much pass stat is down compared to the current NFL.

Going back to 4.5 with with a few of the tweaks passing will be different than both versions. How much?

My bets are if the "long passing nerf" is left in drops will increase even more than 4.5 has produced, because of the "raising of zero speed" .Protection will be better, and running short yards will be easier, but those tight passing windows will be even tigher.

With the zone changes being left out....zone schemes on the outside will be severly hampered making the slants even more important to stop. In the NFL Micheal Thomas connected on 82% of his slants last season, but only a 12 yard average. We will be battling 20+ averages much like 4.5 for that play, but with lower completion rates. Lets see if the tweak to speeed will help here.

The backs in the flats love to drop the ball as well....the "long passing nerf" produced many chances there and if the drop rate is not lowered back down the drops will be tied to the distance of the pass and not if the play call ,QB, or reciever is any good.

Hopefully the pass block tweaks help the passing game as this bug was across all plays where there was more blocker than rushers. Thats why i never like when the TE stays in to block, because the success was tied to alignment not actual skill. But pass blocking/rushing will still be exploitable...hey maybe we can get QBs to run in this new version.

And 8000 yards might be possible for a season but two great owners in the playoffs will not be producing many 400 yard passing games....and 12000 yards seems highly unlikely even with the perfect senerio.

The drops were so bad in 4.5 you could pick a slow CB and actually hope the Wr just drops it, and with zone at a severe disadvantage most of these "stats" produced are vs. man coverage on the outside. You saw what Micheal Thomas did.....Qbs should be at 80% completion rates not 60's......its the defenders jumping routes badly and the speed gaps allowing for big yards after the catch.

This current version the speed gaps are still to big and man is overpowered compared to zone allowing for the stats we will get. Allocations will get different numbers due to lower ratings so there will be plenty to look at but those bad drops on wide open deep passes does not need to stay on this new version of 4.6 . Zone and man need to be more aligned...in balance to get yards after catch right...nerfing long passes has never been the best solution

My goal is to get 0.4.6 out as quickly as possible to address the pass blocking and punt blocking issues. The changes to the dropping logic were part of the overall passing game update, which the consensus seems to be that it is not an improvement over 0.4.5, so I have decided to remove that until 0.4.7 in order to accelerate the release of 0.4.6 by reducing its scope.

https://victory.myfootballnow.com/watch/14#3082

This was 5 oline vs. 5 rushers....Looks similar but this must be a separate isssue...If you count the LB as not being recognized as a rusher due to the late arrival thats one thing, but the three oline on the strong side have only two guys to block. That side falls in to a similar senerio of "too many blockers"

On the weakside , if you count the LB blitz, its a 4 defenders vs. 3 blockers....either way its was a good players getting beat...or they didn't recognize the blitz....but this also holds true that a smaller faster rusher lined up in a five technique could create a true "edge" rusher.

But why didn't the LT pick up the RDE off the snap?

Difficult to pick apart a play from a new League in first week of pre-season as there is so little play knowledge. The RB (who is a WR) should've picked up the block or at least attempted to slow it down. But I actually quite like that as an OOP player he got his assignment wrong.

As for the LT, he turns his back, then teleports into the middle of the LG and C. It's not noted in the log but he may have picked up or accentuated an injury here (he's Questionable for 4 games now).

The LG was also beat on the play, which explains the LT jumping across to block him - hoping the RB would pick up the blitzer. If the LT stays on the DE, the DT is in on the QB.

I'll plug that play pair into my test engine and see if it repeats there or if it looks like a play familiarity issue.