Warrior General Games

Originally posted by Ken1
Originally posted by Adderfist

Originally posted by Ken1


Originally posted by Adderfist



In light of the recent issues I suggest that HB Spy be reinstated.


HB Spy overpowered the defense against runs, as was proven in tests.

Flawed tests.

If you're talking about the "pass focus" bug, that wasn't the issue in HB Spy. It was simply "set two players to HB Spy, shut down the running game." Heck, I could set one to HB Spy and pretty much shut it down.

When you can't do that, if you overshift everyone and the other team runs left, you have trouble as you should. If you can basically shift with the HB, you can shut down any running attack far too easily. It worked plenty well-- way too well-- for me in custom plays. And you don't pay for it when they pass, as you'd normally have someone cover the HB anyway.

This may be a bit of "symantics", but it's mostly frustration. I would still like to see a realistic defense on a fixed server to be sure. Mayhaps it's just the coder in me but I have serious reservations about testing something, and realizing there's a flaw in the field that's being tested. Using the extrapolated flawed data to make a decision, goes against every fiber of my being.

Garbage in garbage out as the saying goes.


I would like to see/design to test and draw conclusions on my own :/
If however, it is to strong I would like to suggest that HB Spy be susceptible to play-action. Play-Action could be a VA that increases the strength of the play action motion. as for the -base- strength? That could be up for debate.
The other issue I have with the spy being taken out, is that if you keep your HB in to block does that not give an advantage to passing?

Originally posted by Ken1
It keeps things in the realm of realism.

I care more about game play than 'realism', but perhaps that is just the jaded DE owner in me after that fiasco... (which still needs fixing!)

My point is simply that setting a simple goal statistic/average to determine success in balancing the run game is most likely going to return flawed results. In order to create an 'exciting' (or one might say 'legitimate') choice the user has to be exposed to the possibility of failure and offered the potential of success. The run game has to offer both to all users, not just the 'typical' ones. And that requirement exists on both offense and defense. A DC must have a legitimate chance of stopping a equal level run, just like a OC must have the chance of generating forward momentum with a run. And above all this cannot be based on chance alone. Nothing disenfranchises a player quicker than removing their ability to affect meaningful change.

Ultimately if we're going to look at game stats and film to determine the health of the running game in the S9 sim, I think clear goals should be set before review as to what constitutes a reasonable success. My hope is, and what my previous post was trying to address, that those goals be framed outside an overly simplistic 'ypc'.

Ken, no offense but you have been an advocate of offense since Day 1 on the servers. Your blatant biased towards it is well known server wide. That Bort would allow you to become a tester speaks volumes on how the entire process is flawed. So pardon me if I dont buy into your theories and methods of interpretation. You havent represented yourself as someone who wants true balance.

Originally posted by Ken1

Average means average, typical. The NFL has over a 4 ypc average; the best backs get about 5 ypc in a given year. The best NCAA backs can get 7 ypc. Those happen in real life, but they don't make 3 runs an automatic first down, because a 4 ypc average can look like: 3,1,3,11,4,7,-1,2,2,8,12,2,0,4,3,5. That group is an average of 4, but there are a number of combinations of three in there that wouldn't be a first down. And those numbers probably aren't far from what a typical real life RB's carries look like, with many shorter gains raised to average around 4 by a few longer ones.

Also, if the other team expects the run and calls run-focus defense, they'll allow fewer than 4 ypc.


I would say swing and a miss but you got a piece of it. yes the NFL has that average. What the NFL doesn't have is 30+ carries week in and week out. If an NFL back ran like a GLB back does he'd make it 4 maybe 5 games into the season then have to go into physical therapy.

Originally posted by Modok
Ken, no offense but you have been an advocate of offense since Day 1 on the servers. Your blatant biased towards it is well known server wide. That Bort would allow you to become a tester speaks volumes on how the entire process is flawed. So pardon me if I dont buy into your theories and methods of interpretation. You havent represented yourself as someone who wants true balance.

This has bothered me for a while, i dont think its limited to ken, 95% of the stuff i have seen being worked on and discussed in the tester threads has to do with improving offense. From what i have observed there is nothing anyone can say to me to convince me that the current crop of testers is HEAVILY biased towards offense and that bias is evident in the way the sim has worked the last couple of seasons and evident in the work and goals and results of most testers.

While the DPC is a big plus for D, I fully expect the DPC to be heavily neutered by next midseason.

Originally posted by bluemagus
Originally posted by Ken1

Originally posted by PinTBC


Wait a minute! A 4 YPC average would mean that on average that back would get a first down every three tries. His team would be pretty much unstoppable, because since this is a On-line game, there is almost no ramifications to simply running the ball on 4th down if there are 2 or less yards to go.

PinTBC

Average means average, typical. The NFL has over a 4 ypc average; the best backs get about 5 ypc in a given year. The best NCAA backs can get 7 ypc. Those happen in real life, but they don't make 3 runs an automatic first down, because a 4 ypc average can look like: 3,1,3,11,4,7,-1,2,2,8,12,2,0,4,3,5. That group is an average of 4, but there are a number of combinations of three in there that wouldn't be a first down. And those numbers probably aren't far from what a typical real life RB's carries look like, with many shorter gains raised to average around 4 by a few longer ones.

Also, if the other team expects the run and calls run-focus defense, they'll allow fewer than 4 ypc.

This isn't real life though. If you base 'satisfactory results' on a number (4ypc) and adjust the sim to it, you might not have balance. It depends on how large the variance is. S8 was the same ypc on the same run over and over. If every run (good average, low variance) is 4ypc then you have a failed sim.

AS he states, you are assuming many more things than I am. If your variance is not large, then a 4 yard average would be the death of this game because you simply couldn't stop a heavy rushing game. Right now, there really isn't a bunch of variance in the rushing game. Teams find a couple plays that work well and beat them to death.

PinTBC

Originally posted by bhall43
Originally posted by Modok

Ken, no offense but you have been an advocate of offense since Day 1 on the servers. Your blatant biased towards it is well known server wide. That Bort would allow you to become a tester speaks volumes on how the entire process is flawed. So pardon me if I dont buy into your theories and methods of interpretation. You havent represented yourself as someone who wants true balance.


Ken,

You are simply looking at one stat that is part of the whole, and focusing on the real game of football. This sim isn't real football, as much as some of us would love to see that. If you get your way, and the avderage yards per carry is 4, and the average yards per pass is 7, and the variances stay the way they are now, there will be no need for a defense, because it simply won't be able to stop anyone. By the way, making large variances is probably not the way to go here as the game starts getting the random feel to it, where people can't count on their players to do what they are supposed to.

I know you hated season 3 because the defenses managed to keep most offenses to reasonable season stats, but I'd say that season three was markedly more interesting to play.

PinTBC

Honestly, a 3.x ypc is great against a run focused D.

I have always believed that the game is won on the LOS. I would point back to my initial comments in this thread, it appeared as though the DL was dominate over the OL which contributed to less yards in the early sims.

As others have said, look to the run AI for the OL for answers to this balance. Stop watching the HB and start watching the C, G and OT. A little more aggressive and a little less random/mindless and you have a good run sim. Yes, the HB effects the run game heavily. But honestly, he looks fine to me.

The OL is naturally stronger than the DL, which is an obvious intended design on Bort's part. In every match-up the OL has a leg-up on leveling. Allow them to be more active in the game and not only does your run-sim work better but you have some happier owners. And it works for the D too, because if the OL is out-classed by the DL you shouldn't have much of an inside run game to begin with...

If people have some ideas of the kind of things on defense they'd like to see tested, I'm all ears. I've always been more of a defense guy myself but it's harder to test defense because there are so many variables, and that gets even further complicated when you add in the DPC. Offense is easier to test because all you have to do is run a play against various types of defenses to see their effectiveness. With defense, let's just look at the 4-3 formation for example.

5 coverage shells
2 coverage types
3 coverage distances
3 play focuses
3 LB shifts
3 line shifts
7 different number of blitzers
7 possible positions that can blitz

That's well over 10,000 possible combinations for just one position. If someone can come up with a concrete idea and there is a verifiable problem that needs testing, I'll definitely look into it and do all I can.

Any test games where a 4-4 D has been worked with?

Originally posted by joe
Any test games where a 4-4 D has been worked with?

you kind of have to fake the 4-4, it isn't one of the options.

Originally posted by thehazyone
If people have some ideas of the kind of things on defense they'd like to see tested, I'm all ears. I've always been more of a defense guy myself but it's harder to test defense because there are so many variables, and that gets even further complicated when you add in the DPC. Offense is easier to test because all you have to do is run a play against various types of defenses to see their effectiveness. With defense, let's just look at the 4-3 formation for example.

5 coverage shells
2 coverage types
3 coverage distances
3 play focuses
3 LB shifts
3 line shifts
7 different number of blitzers
7 possible positions that can blitz

That's well over 10,000 possible combinations for just one position. If someone can come up with a concrete idea and there is a verifiable problem that needs testing, I'll definitely look into it and do all I can.

Concrete ideas, eh? I'll get you started:

Test line/linebacker shifts. You can work out of cover 1 man or cover 3 zone, run focus, short coverage, no blitzers as a reasonable baseline rushing defense (read: should be clamping down on running backs pretty well). Test plays that are running away from both shifts, into both shifts, away from line/towards LBs (with LBs and line shifted opposite, obviously), and away from LBs/towards line (ditto). Hit both off-tackle plays and pitches/sweeps for all of these shifts. This should be something that's doable in 4 games with a semi-reasonable sample size.

I'd expect this test to answer the following questions:
* Is a defense properly vulnerable to opposite-side runs when it shifts both the line and the LBs to the same side?
* When a defense is shifted entirely to one side, do running backs have cutback lanes open as much as they should? Do they take proper advantage of them?
* Does a strong-shifted line/weak-shifted LB play work as expected (most effective at shutting down off-tackle runs to the strong side and pitches to the weak side, weak at shutting down pitches to the strong side and off-tackle to the weak side if the FB can get to the ROLB)?

Originally posted by OttawaShane
Originally posted by joe

Any test games where a 4-4 D has been worked with?

you kind of have to fake the 4-4, it isn't one of the options.

Yes that is what I am talking about. Using a set up of a 4-4 just with the SS or who everyplaying that 4th LB. Has anyone tried that? Would love to see replay of it.

Originally posted by PP
Originally posted by tautology

Folks,

I've said this before and it seems to fall on deaf ears...but please consider it comes from a coach who has consistently had both the best rushing game and best rushing defense in his league season after season.


The running game was barely functional in season 8, with the exception of the Strong I Slam

Every other running play could be stuffed for a loss with alarming consistency unless the HB was an absolute stud and simply broke tackles to make gains.

Functional, but barely so. And yes, this is why the slam was so prevalent...it was the only play with reliable yardage.

And even the slam could be slowed/stopped with the right defensive calls and good defensive builds...as thehazyone showed in the USAPro championship game.


I am not advocating no changes in the sim but...


The changes being made are fundamentally misdirected

This is an over-reaction that will nerf the running game globally in response to a single-play trend in season 8.


What we need is to LOOSEN UP the running game in general, and slightly nerf the specific dynamic that made the Slam a bit over-powered (and it wasn't over powered so much as it was very difficult to ever stop for a loss..every other run was FAR TOO EASILY stopped for a loss, which is why peopled stopped using them and went All -slam).

And I will say this: The D AI alone would solve the Slam problem.

From the looks of the sims I have seen, the ONLY running games that will be viable for season 9 are ones with insane powerbacks vs poor-tackling teams.

Let's not go back to season 3, please.

100% agreed. Honestly, by the end of ssn, the SI Slam wasn't much of a concern to my D at all. I'd have LOVED playing a 80% or more SI SLam team! I don't even believe it was over powered, once the Ds caught up. Frankly, the repeated play penalty that Bort is adding, whether you like or hate it, solves the SI Slam issue all on its own. Toss in the D AI creator and I don't see it being an issue at all.

I used other run plays all ssn long. The only time they worked somewhat consistently was when you could catch the other team in the wrong D. When in the right D, they were fairly consistently shut down, many times for a loss, and I've got a damn good run team.

As far as ideas tossed arond so far (from multiple sources), here are my thoughts:
OLs with much higher str (20 or more) SHOULD get movement on the D. What Bort is responding to here by limiting it is all the guys that made ultra high agility and low str DLman. Sorry, but if you have comparatively no str, you should get blown out of the hole by a strong OLman that gets his hands on you.

There should be more of an energy drain for HBs. Having 1 HB being able to run the ball 40 or more times in a game is insane.

Reduce fall forward, but not by more than 50% of what it was at. Frankly, reducing it by 20% most likely would be enough.

PBs broke too many tkls, but, again, not many more. I'd again that a 20% reduction would be sufficient (otherwise you're just rewarding players for having shitty builds on D and again penalizing those that put the effort into stopping broken tkls)

DL Shifts (along with the D play creator) will have a very negative impact on the run game, as is. I like both, but, without a counter boost of some sort to blocking, this alone will kill the run game. IF you toss in the LB's reacting better and slowing the HB through the line, the run game is completely dead right there.

Lead blocking needs to be fixed. Far too often, the FB (or Gs, OTs & TEs) make flat out foolish blocks, allowing the best positioned defender a wide open shot at the HB

On outside runs, the HB needs to break more outside of the DE. Many plays, the HB only needs to go a couple clicks more to the outside to get a good gain. Instead, he tries to cut it up right next to the DE and is tkled for a loss.

Final thoughts:
Don't over react because ppl were screaming. By far the biggest problem was that there was no penalty to running the same play 30 or more times a game and that a PB could have 40+ carries without a drop off in production. The repeated play thing sucks, and has happened every single ssn since the O AI was unveiled (it's just that so many did it this ssn that it finally got noticed). Frankly, it sucks playing PO games where the other team runs 1 play 40-70 times. The play penalty takes care of that.

60-70% of the plays in the play book in the passing game are flat out horrible and have been since they were 1st introduced (bump that up to 70-80% for the run plays). They just don't work against a decent team under all but the most ridiculous circumstances. Some of the pass plays don't even make an football sense at all. They desperately need to be deleted and new, functional plays inserted in their place.

Don't overreact by overpowering the passing game. IMO, the only thing that absolutely needed to be fixed was the QB hitting open WRs/TEs/HBs, instead of forcing the ball so much into double and triple coverage. Leading the WR should be improved slightly. However, again, don't over react by making it any more than 20% better or you just killed the D.


Think that's it


Sad thing is, both these posts will have nothing to do with the season 9 sim about to be unveiled.