New Sticky Note

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: [0.4.6] Version 0d5f5ac

By raymattison21
10/08/2020 5:03 pm
TheAdmiral wrote:
Ray, I'm not suggesting we get told what the desired results are for each play, with regards to who the ball is thrown too.

I'd just like a breakdown of what happened when I chose that play (rare for me and Smirt to agree on anything). Each play card should be able to tell me what happened when I used that

eg

WR1 thrown to 9 times, 6 catches, 45 yards, 1 TD, 1 INT
WR2 thrown to 3 times, 2 catches, 8 yards, 0 TD, 0 INT, 2 First downs
WR3 thrown to 1 time, 0 catches
RB1 thrown to 10 times, 8 catches, 40 yards, 5 First downs
TE1 thrown to 4 times, 2 catches, 12 yards, 2 First downs, 1 INT
QB scrambled 3 times for 17 yards, 2 First downs, 1 sack

Overall
play ran 30 times, 18 catches, 3 runs (QB scramble) total 122 yards, 1 TD, 2 INT, 11 first downs, 1 sack, 0 fumbles

average yards to go when play called 4.3 yards


More stats would always be great. With time you can get your own now.... I want to be told either but reads and responsibilities within certain play matchups should follow football rules. Variations from that need to be tied to skill and just plain errors so the user can adjust accordingly using football knowledge.

If we adjust to a game then that’s what it is .... making the stats fit and look good. But I want things to continue to act in a football manor and have the stats follow that. Moves like tightening up speed are good. Making man coverage and the QBs skills matter more we’re all good code moves. Most have been..

It is like my criticism of the blitz pickups. There not my favorite but something similar is going on here and in football. I know it won’t be perfect but jdb comes up with fair solutions as long as the play can match up with the code somewhere.

Dump offs , sacks / pressure, reads/progression, knockdowns/drops , and ints have always been up for heated debates. Future tweaks will be needed but it is like the suggestion of not penalizing offensive plays that face non penalized defensive plays. At least the ratings matter and not abuse.

The overuse penalties need to be fazed out. One thing is throwing off timing through predictable but another catching someone off guard. Throwing a new gameplan out there can do that but nothing natural is going on in the game. Audibles were to see around but...

The closest thing is the game play adjuster.....it s pretty good at mixing things up.

If counter weak is overused vs. olb zone flats it gets worse each time....great but what’s stopping a guy from having to use several runs ... only him making a different game plan, but if he could not stop the counter with that olb zone flats then the defense would have to pick one of the several blitz plays to stop it. Thus opening up other options offensively cause they can’t spam the flats play vs a lot of plays

Other plays act similar and making the offensive more viable in those scenarios has to be good . We’re getting past the overuse penalties. Not with the blitz yet but the offenses in general have been hampered for years now. Why penalize them at the rate they are at?

I used to use 3 plays total now on offense I try not to any play over 3 times....we come a long long way. I reserve scheming up something different for big games but thats just to catch users off guard. Still, I don’t know if it works or we just had a bad set of rolls...like I said due to out side factors .

Re: [0.4.6] Version 0d5f5ac

By TheAdmiral
10/08/2020 5:13 pm
Anyhow, Victory is now two weeks into the Regular season.

https://docs.google.com/spreadsheets/d/143k_CaYKVA0I356GFK3_JXPd7m2ZdZdPQnI_anlXRtE/edit?usp=sharing

Power rankings, league leaders, Punters and QB rankings.



I'll dig in a bit deeper over the weekend but this is how things are shaping up
Last edited at 10/08/2020 5:15 pm

Re: [0.4.6] Version 0d5f5ac

By Infinity on Trial
10/08/2020 8:37 pm
Smirt211 wrote:
I parsed something out.

which receiver gets the most throws in his direction from that play


This would be an exceptional idea. Spray ratio. You pull an offensive pass play up in your playbook and it says listen here's the % of whom the ball is distributing to. Sort of like the 1st through 4th Down stuff I always ignore on the plays.

25% - WR #1
25% - WR #2
12% - TE #1
13% - Incomplete
25% - RB #1

That would be completions. Ok, throws in the direction:

25% - WR #1
25% - WR #2
25% - RB #1
25% - TE #1

....Great idea. I'd remove the 1st through 4th down nonsense and insert this in its place.


Here's a snapshot from the DB:


Re: [0.4.6] Version 0d5f5ac

By Smirt211
10/09/2020 5:05 am
It's hard to focus on too much, however, I'll isolate it. I've noticed there's plays where it'll force it away from the WR #2 on the FL Hitch. (defensive plays)

Looking here, Crash Right (probably a reaction to the blitz) funnels to the WR #3 predominantly.

WR #3 would be the RB #1 in the case of the 2-1-2 set?

That means a trigger of a hot read type reaction which keeps it from waiting to nail the FL Hitch route.

Wait, I was in the wrong columns...
Last edited at 10/09/2020 5:07 am

Re: [0.4.6] Version 0d5f5ac

By raymattison21
10/09/2020 6:56 am
https://victory.myfootballnow.com/watch/424#78224

Isolated too much try to game the game and we will still have plays like this. Not flanker hitch here but a deeper slant vs a pretty poor CB and a C+ grade FS.

IRL you might get a 16-20 yard gain as the cb is beat but that safety gambling and doing a poor job at it is what turns this play into a 96 yard td. Monster RAC yards achieved where matching a decent defensive play just doesn’t matter cause that gamble.

Matching a play cause is comes up in Seth’s database is great but it only shows the inconsistencies with play. Trying to force a quick pass or hope your safety doesn’t think he’s Ed Reed everytime a ball comes his way.

Talent or no talent if this play did not happen this way we’d thrown for only 200 yards. On a side note the 74 speed CB actually was catching up to a 94 speed 63 carry WR.

Re: [0.4.6] Version 0d5f5ac

By TheAdmiral
10/09/2020 8:20 am
raymattison21 wrote:
https://victory.myfootballnow.com/watch/424#78224

Isolated too much try to game the game and we will still have plays like this. Not flanker hitch here but a deeper slant vs a pretty poor CB and a C+ grade FS.

IRL you might get a 16-20 yard gain as the cb is beat but that safety gambling and doing a poor job at it is what turns this play into a 96 yard td. Monster RAC yards achieved where matching a decent defensive play just doesn’t matter cause that gamble.

Matching a play cause is comes up in Seth’s database is great but it only shows the inconsistencies with play. Trying to force a quick pass or hope your safety doesn’t think he’s Ed Reed everytime a ball comes his way.

Talent or no talent if this play did not happen this way we’d thrown for only 200 yards. On a side note the 74 speed CB actually was catching up to a 94 speed 63 carry WR.



Yes, it's poor play but, they're poor players. Remember it's a young league and all teams have exploitable holes.

Without checking, I would guess the Safety has poor disciple, perhaps high volatility (although most seem convinced volatility only affects training camp) and likely low intelligence. Hmmm, sounds like Seattle's Tre Flowers ;0)

Re: [0.4.6] Version 0d5f5ac

By raymattison21
10/09/2020 12:13 pm
TheAdmiral wrote:
raymattison21 wrote:
https://victory.myfootballnow.com/watch/424#78224

Isolated too much try to game the game and we will still have plays like this. Not flanker hitch here but a deeper slant vs a pretty poor CB and a C+ grade FS.

IRL you might get a 16-20 yard gain as the cb is beat but that safety gambling and doing a poor job at it is what turns this play into a 96 yard td. Monster RAC yards achieved where matching a decent defensive play just doesn’t matter cause that gamble.

Matching a play cause is comes up in Seth’s database is great but it only shows the inconsistencies with play. Trying to force a quick pass or hope your safety doesn’t think he’s Ed Reed everytime a ball comes his way.

Talent or no talent if this play did not happen this way we’d thrown for only 200 yards. On a side note the 74 speed CB actually was catching up to a 94 speed 63 carry WR.



Yes, it's poor play but, they're poor players. Remember it's a young league and all teams have exploitable holes.

Without checking, I would guess the Safety has poor disciple, perhaps high volatility (although most seem convinced volatility only affects training camp) and likely low intelligence. Hmmm, sounds like Seattle's Tre Flowers ;0)


I have heard it time and time again. Most would like to think ratings do more than that can do but it’s just not the case. Never has been but the best info I have found tells a tale of it being random with a tie to experience.

Alignment of possibilities and low non factoring ratings is easy when you see tons of plays. But I realness check rookies stats compared to veterans speaking of knockdowns and ints.

As that’s the only difference in the code. Proximity of possibly making a play has to do something too so speed is a great leveler and intelligence plays a role with familiarity but that effect is supposed to be low. This is the first time we call this pass in a long while if any with this team. Our familiarity had to be low. Idk about theirs but it should make a difference anyway unless it “mashed”

The safety had 75 zone coverage.....and 75 speed, both are in the top 75% possible skill so I would expect something that looked familiar in term s of football. Jumping a route and ending up that far behind the receiver is the last thing you want to do.

The safeties responsibility is the deep half....no one should get behind him. You point out that this should happen but if a pro had did that he be sitting next game. If he let up a 20 yard gain he would still be starting next week.

The debate here is not about whether it’s a poor result or not ........ it’s how poor the results are .

Focusing on that alone is the bigger weakness in the code. Yes 100 zone does a bit better but 0 zone doesn’t fail enough. And the point when it does fail your screwed and in a foot race no matter the ratings.

It’s like the olb flats play they drive hard into the deep curl which is where the FS should be and don’t get angles the same as playing an actual FS position and go for less of these big whiffs and misses.

Some of it looks and acts like football but playing the deep half should be enough. The wr is running right into his zone. It’s a good call to stop it.

Count how many times tre flowers jumps ball like that. I doubt is anywhere near the rate it happens here. Count them up....when he’s in a deep half and has the ball thrown in front of him. Not when he has the flats hooks or curls when he’s in the deep half.

As I would jump more passes knowing someone had me deep but our guys do it pretty randomly...blowing up yards after the catch...thus making coverage better than it has to be in other areas....to make up for the huge numbers gained in scenarios like this.

Re: [0.4.6] Version 0d5f5ac

By TheAdmiral
10/09/2020 1:21 pm
I think it comes partly from jdb looking at stats from a macro point of view where he'll run a play 1,000+ times and see what average yards that throws out. If the average ypp looks right, he may not delve any further - so it may be that there is a quirk in the output that doesn't get picked up until it is spotted and highlighted in here.

The issue with that is people are unlikely to point out that "I run this play a lot, 95% of the time it does what it supposed to do, but once or twice a game BOOM"

Are the outliers looked at in detail?

Maybe that play (in your game) was an outlier that would only hit once in a 1000 attempts?
Last edited at 10/09/2020 1:24 pm

Re: [0.4.6] Version 0d5f5ac

By raymattison21
10/09/2020 3:13 pm
I picked thinking it would force it in to wrs running those slants. So far it worked 1/1 in games vs a poor cb. The QB probably won’t even attempt a throw there vs a better set of DBs. This guys accuracy is on the lower end of functional around 50.... alot of risk every pass.

Re: [0.4.6] Version 0d5f5ac

By TheAdmiral
10/09/2020 4:03 pm
raymattison21 wrote:
I picked thinking it would force it in to wrs running those slants. So far it worked 1/1 in games vs a poor cb. The QB probably won’t even attempt a throw there vs a better set of DBs. This guys accuracy is on the lower end of functional around 50.... alot of risk every pass.



The three guys in the vicinity (inc the Linebacker) were rookie, rookie and 2nd year player (I think - it's difficult to pick out shiirt numbers even if you zoom right in.) - they could maybe do with some veteran help.