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

The new user interface is in preview!

Want to check it out? Click here! (If you don't like it, you can still switch back)

League Forums

Main - Beta Chat

Re: Version 0.4.1 RC 1

By jdavidbakr - Site Admin
3/20/2017 9:13 pm
setherick wrote:
Are all the changes in this week's games? Your OP said Week 9 so I hadn't studied the film closely.


That last comment was not in week 9, it will be in week 10.

Re: Version 0.4.1 RC 1

By Ares
3/21/2017 3:55 am
WarEagle wrote:
Just trying to use my team as a test for this vital issue.

At one point this was mentioned as one of the items that was going to be addressed in the next version, so I'm posting updates to indicate it hasn't been yet.

I guess I should have posted in the other thread instead of this one.

Is that what your issue with my comment is, or should I remain silent on it altogether?


My point was that it's not testing anything, since this is a well documented issue and there's been no changes implemented that should impact it. If you're itching to test something, you could cut your PB down to only a handful of core plays to test how much impact the overuse penalty has. Or you could continue to use your current blitz only PB to establish a base line for when a change does get implemented. You don't need to shut up or anything, I just couldn't understand why you seemed to be expecting it to have been resolved without any adjustments to the pertinent code.

jdavidbakr wrote:

* Play experience: I noticed in your game against SF that the first half the game stayed relatively close, but started to get out of hand in the second half. Someone mentioned somewhere that the in-game play experience boost seems to benefit the team calling the play more than the opponent. So, I'm trying a new change - the team calling the play will not have the same improvement as the team facing it, and the more you call the same play the more your opponent's experience for it will increase. This may have an adverse effect too quickly if you call the same play frequently but still realistically, so let me know what you think. It's a linear increase now, but I might change it to exponential where it takes 5-6 times of seeing it before it grows faster than normal, but at that point starts to grow even faster.

* I'm also tweaking the hot read a bit more. In the last sim I added verbiage indicating that a hot read was called into the play-by-play, and noticed when looking through that the hot read was frequently unsuccessful. I'm modifying that play so that the catch is more likely to be made, and the immediate tackle probability is reduced.


I really don't like artificial band-aids, which is what this is. Especially the latter, which if I'm reading it right is suggesting a magical bonus to QB/WRs and a penalty to the initial defender? These type of solutions will just serve to mask the underlying issues that are making these plays not work. It would be far better to just figure out why plays that should be impactful aren't, rather than wave a magic wand and make a noodle armed QB throw, a stone handed WR catch, and a pro-bowl LB quiver at the thought of tackling.

So why aren't they working? To the logs!

https://mfn1.myfootballnow.com/gamecenter/view/6175#1111178

The log indicates a hot read was made, and indeed the TE is completely uncovered directly in front of the QB. Why doesn't he make that throw? There's no need to adjust anything other than making the QB throw to the wide open player right in front of him for the walk in TD. Why does he take the sack?

https://mfn1.myfootballnow.com/gamecenter/view/6175#1111299

Hilariously the exact same situation from the same game, but the sides are reversed for the same result. The TE is once more uncovered in front of the QB, who rather than actually make the uncontested throw for a sure-fire TD takes off scrambling.

https://mfn1.myfootballnow.com/gamecenter/view/6175#1111317

Executed perfectly and highlights the major risk of a heavy blitz with press man to man coverage without any help over top. The TE beats his man at the line, the QB connects to him, and it's a foot race from there.

https://mfn1.myfootballnow.com/gamecenter/view/6180#1112134

Oh my god why does this keep happening!

https://mfn1.myfootballnow.com/gamecenter/view/6179#1111921

What is even happening here? The route concept on the play card doesn't at all match-up to the final route that's run. Is this an errant throw that the WR makes an unbelievable adjustment to? Is this the intended audibled 'hot route'? The 'hot' read here is even odd to begin with. Third and 17, outside FG range, down by 17 in the 3rd... so long as you're committed to going for it, which the team is given the play call, this is a terrible time to try to make a quick throw to the "emptied space" (not even emptied, more on that below)... but then no quick throw was even made, so it's not even hot. At most it's an adjustment maybe to who the primary target is, in which case I think even that's a debatable choice.

The CB3 and both interior LBs are blitzing. This leaves the go route in front of the QB (by far his easiest throw other than the outlet TE) in M2M with no help over top. This should be the QB's dream scenario. Instead he targets the Y WR, which isn't a bad call either, considering he's running an inside breaking post route right to where the field just got emptied by the blitzing trio. This leaves single coverage against a safety giving zero cushion, which should be cake, but for some unknown reason the WR cuts back and outside in direct opposition to his called route. This play also highlights the problem with DBs making insta-reads on throws, as the outside CB also converges on the pass, despite himself being in single coverage on a deep post.

https://mfn1.myfootballnow.com/gamecenter/view/6183#1112541
https://mfn1.myfootballnow.com/gamecenter/view/6183#1112575
https://mfn1.myfootballnow.com/gamecenter/view/6183#1112639

Look at #87 running free. That should have been the hot route and a walk-in TD. Three times!

https://mfn1.myfootballnow.com/gamecenter/view/6183#1112643

He did it! He did it! FINALLY.

https://mfn1.myfootballnow.com/gamecenter/view/6183#1112706

Oh come on, I thought we'd figured this out. YOU EVEN STARTED OFF LOOKING AT HIM!

https://mfn1.myfootballnow.com/gamecenter/view/6183#1112618

Why are you hot reading on press coverage on 3rd and 10 with four deep routes? That's four guys all in single coverage M2M in the end zone. You take that shot.

https://mfn1.myfootballnow.com/gamecenter/view/6183#1112626

Hit the TE! Hit the TE! No!! What are you dooooing?!

https://mfn1.myfootballnow.com/gamecenter/view/6183#1112689

Well are you going to hot read to the wide open dude in front of you or what? No? Okay, I see, you're going to take a sideline sack instead. Alrighty then.

Hot read executed:
https://mfn1.myfootballnow.com/gamecenter/view/6172#1110728
https://mfn1.myfootballnow.com/gamecenter/view/6172#1110737
https://mfn1.myfootballnow.com/gamecenter/view/6181#1112232
https://mfn1.myfootballnow.com/gamecenter/view/6181#1112311
https://mfn1.myfootballnow.com/gamecenter/view/6184#1112741
https://mfn1.myfootballnow.com/gamecenter/view/6184#1112820
https://mfn1.myfootballnow.com/gamecenter/view/6184#1112850

Hot read not executed:
https://mfn1.myfootballnow.com/gamecenter/view/6175#1111214
https://mfn1.myfootballnow.com/gamecenter/view/6175#1111281

This game had 31 hot reads, I included only the highlights:

https://mfn1.myfootballnow.com/log/6183

Final conclusion:

So here's the issue with the hot reads as they're currently implemented:

1) Don't take down/distance/situation into consideration. 1st and 10, plenty of clock, **** yeah, take the free yards. 3rd and 10, clock is waning in the final minutes... no, you risk the sack and go for the win.
2) They frequently hot read to the wrong guy, ignoring the player often left completely uncovered at the LOS.
3) But most importantly: M2M defenses always leave zero cushion.

Hot reads (as they're being interpreted in MFN) in the NFL work because when you call a blitz you're generally emptying the short coverage, allowing the WR to make an uncontested catch in stride with room to move. In MFN however, unless there's fewer players in coverage than designated routes, every receiver gets covered M2M without any cushion at all. In this situation in the NFL, you wouldn't hot route, but rather you'd hit whatever your deep route is, or audible one of your outside WRs into a go route if none was called. Single coverage M2M you expect your primary WR to win that contest, which is why defense rarely do it. It's a huge risk, and only works when you catch the offense off guard or you're employing someone named Revis (well, 3-8 years ago, at least).

In other words, if a defense calls a blitz with ALL M2M coverage and an offense succeeds in diagnosing the blitzers (and thus identifying where its single coverage will be), they should target an existing--or audible into--a single coverage deep route. If the defense calls a zone blitz or keeps its safeties over top, hot read to the TE or Y WR to wherever just got emptied.
Last edited at 3/21/2017 3:59 am

Re: Version 0.4.1 RC 1

By jdavidbakr - Site Admin
3/21/2017 8:34 am
Great analysis, Ares. I'm going to make another tweak to try to help the QB decide to throw the ball when he is under pressure and his receiver is wide open, again for the week 9 games.

Re: Version 0.4.1 RC 1

By blackflys
3/22/2017 9:06 am
What's up with the sim ?

Re: Version 0.4.1 RC 1

By jdavidbakr - Site Admin
3/22/2017 9:18 am
Had a bug in the latest changes that crashed it. I appears to be running now, should be done shortly.

Re: Version 0.4.1 RC 1

By Mr.Krazy
3/23/2017 12:16 am
Anyone else feel that the passing game was slightly overpowered in week 9? Maybe it was just pure luck, but both QB's seemed to be on fire this game

https://mfn1.myfootballnow.com/box/view/6194

One instance saw a receiver catch the ball in between 4 defenders!

https://mfn1.myfootballnow.com/gamecenter/view/6194#1117125

Saw improvement in Linebacker play, 2 of my linebackers were the game leaders in tackles

Re: Version 0.4.1 RC 1

By blackflys
3/23/2017 9:53 am
I would think by testing you could get the numbers to match up to NFL.
It's not even close. Passing was already a problem being to easy and now the numbers are getting worse and worse.

This season passing rating over 150 attempts
8 QBs over 110 rating
11 QBs over 100 rating
22 QBs over 90

NFL
1 QB over 110
5 QBs over 100
16 over 90

LG 10 old engine
1 QB over 110 this guy is in a two QB system (also multi QB systems are too successful)
5 QBs over 110
11 QBs over 90

Clearly this proves that passing is unrealistic and has changed for the worst. Looks like the old release was **** near spot with the most important stat for the most important position.

Two and three QB systems shouldn't even be an option. Name one situation in the last 20 years where you seen it be successful. Funny thing is the QBs in these systems are on the average side and have seen garbage ones play like MVPs. If a team ends a season with three QBs over 105 rating there is a problem.

Re: Version 0.4.1 RC 1

By CCSAHARA
3/23/2017 6:33 pm
LOL despite 2 QB with high rankings and several WR of #1 receiver status. The Giants are no threat to adding to those stats.

Re: Version 0.4.1 RC 1

By setherick
3/23/2017 7:12 pm
blackflys wrote:
I would think by testing you could get the numbers to match up to NFL.
It's not even close. Passing was already a problem being to easy and now the numbers are getting worse and worse.

This season passing rating over 150 attempts
8 QBs over 110 rating
11 QBs over 100 rating
22 QBs over 90

NFL
1 QB over 110
5 QBs over 100
16 over 90

LG 10 old engine
1 QB over 110 this guy is in a two QB system (also multi QB systems are too successful)
5 QBs over 110
11 QBs over 90

Clearly this proves that passing is unrealistic and has changed for the worst. Looks like the old release was **** near spot with the most important stat for the most important position.

Two and three QB systems shouldn't even be an option. Name one situation in the last 20 years where you seen it be successful. Funny thing is the QBs in these systems are on the average side and have seen garbage ones play like MVPs. If a team ends a season with three QBs over 105 rating there is a problem.



QBR is probably the worst stat to use as a way of comparison. Here's why:

1) MFN QBR is inflated by the number of TDs that QBs can throw. The number of TDs is more from WRs making big plays than it is QB play. WRs break tackles easily, especially when you put your 100 break tackle players at WR. And they also catch a lot of contested ***** (and drop a lot of wide open ones).

2) NFL QBR is driven more off of completion rate than TDs. Good NFL QBs complete 60-65% of their passes. Good MFN QBs complete 55-60%. ****, many owners are happy with 50% (although I think that's a game planning problem).

So in MFN it's harder to complete passes, but easier to score passing TDs. That's going to artificially inflate QBR to a point where it's not a good comparison tool.

Re: Version 0.4.1 RC 1

By jdavidbakr - Site Admin
3/23/2017 7:19 pm
setherick wrote:
QBR is probably the worst stat to use as a way of comparison. Here's why:

1) MFN QBR is inflated by the number of TDs that QBs can throw. The number of TDs is more from WRs making big plays than it is QB play. WRs break tackles easily, especially when you put your 100 break tackle players at WR. And they also catch a lot of contested ***** (and drop a lot of wide open ones).

2) NFL QBR is driven more off of completion rate than TDs. Good NFL QBs complete 60-65% of their passes. Good MFN QBs complete 55-60%. ****, many owners are happy with 50% (although I think that's a game planning problem).

So in MFN it's harder to complete passes, but easier to score passing TDs. That's going to artificially inflate QBR to a point where it's not a good comparison tool.


This is my perspective as well - ideally the QBR will get into the proper range, but right now I'm focusing mostly on completion percentage, INT percentage, and YPA stats. Completion percentage has been too low and INT percentage too high previously, that's the main point of focus for this game engine version at the QB position. QBR is likely a symptom of other issues with the WR/DB play as setherick says.

The 2-QB system issue is something else I'll probably want to address in a future update, but really need to think that through. I think when I get player-to-player familiarity built (possibly in the next version, but that's a pretty big undertaking and may get pushed further back) that may be a place to apply logic that makes a multi-qb system difficult to use. Maybe some sort of an in-game adjustment period for both the WR and QB when a new QB comes on the field in the middle of the game as well. But that's really a topic for another thread.