Fantastic Work, the best work on the test server so far
Forum > Test Server Discussion > Test Server: Testing Pass Rush Effectiveness
TheInfinity
offline
offline
Hazy, thanks a ton for doing this. Its excellent work and will hopefully lead to this problem finally being addressed fully as I assume this totally ends the debate as to whether or not DEs are being screwed over.
One thing though. You said that the average of 14.23 sacks over a 750 play season is about right. I agree there, but looking at the testing parameters, you have the offenses set to 100% pass. Meaning, that the 750 plays are all pass plays. That's never going to happen in a season, most teams get up to about 400-500 unless they are extremely pass heavy.
Other than that, awesome work in proving that there is indeed an issue here.
One thing though. You said that the average of 14.23 sacks over a 750 play season is about right. I agree there, but looking at the testing parameters, you have the offenses set to 100% pass. Meaning, that the 750 plays are all pass plays. That's never going to happen in a season, most teams get up to about 400-500 unless they are extremely pass heavy.
Other than that, awesome work in proving that there is indeed an issue here.
tautology
offline
offline
Originally posted by Time Trial
DEs shouldn't be beating them in one on one coverage very often if they have similar builds. You've got five men on the line and you are rushing four. The DEs are getting boxed out (not beaten or taken down, just given the run around), which should leave a hole for a blitzing back to run through. The defensive player is having to swim against a big man whose job is just to keep you out. A well built tackle should be able to stand up to a well built DE 9 times out of ten in a one on one situation. Try spreading your formation out to stack one side (DE wide left, DT wide left, NT Center, DE Right) and see if the guards and center pick up the force to that side.
I just don't see your point on this one DE owners. Rushing four shouldn't put up the kind of pressure that you think it should.
The issue here is not what happens with equivalent builds, but rather what happens when the DE build dominates the OT.
Which actually may not be the best perspective from which to view the problem, now that I think about it...
DEs shouldn't be beating them in one on one coverage very often if they have similar builds. You've got five men on the line and you are rushing four. The DEs are getting boxed out (not beaten or taken down, just given the run around), which should leave a hole for a blitzing back to run through. The defensive player is having to swim against a big man whose job is just to keep you out. A well built tackle should be able to stand up to a well built DE 9 times out of ten in a one on one situation. Try spreading your formation out to stack one side (DE wide left, DT wide left, NT Center, DE Right) and see if the guards and center pick up the force to that side.
I just don't see your point on this one DE owners. Rushing four shouldn't put up the kind of pressure that you think it should.
The issue here is not what happens with equivalent builds, but rather what happens when the DE build dominates the OT.
Which actually may not be the best perspective from which to view the problem, now that I think about it...
23yrwej
offline
offline
Originally posted by Time Trial
DEs shouldn't be beating them in one on one coverage very often if they have similar builds. You've got five men on the line and you are rushing four. The DEs are getting boxed out (not beaten or taken down, just given the run around), which should leave a hole for a blitzing back to run through. The defensive player is having to swim against a big man whose job is just to keep you out. A well built tackle should be able to stand up to a well built DE 9 times out of ten in a one on one situation. Try spreading your formation out to stack one side (DE wide left, DT wide left, NT Center, DE Right) and see if the guards and center pick up the force to that side.
I just don't see your point on this one DE owners. Rushing four shouldn't put up the kind of pressure that you think it should.
It should put up a sack every once in awhile especially vs. inferior OTs.
DEs shouldn't be beating them in one on one coverage very often if they have similar builds. You've got five men on the line and you are rushing four. The DEs are getting boxed out (not beaten or taken down, just given the run around), which should leave a hole for a blitzing back to run through. The defensive player is having to swim against a big man whose job is just to keep you out. A well built tackle should be able to stand up to a well built DE 9 times out of ten in a one on one situation. Try spreading your formation out to stack one side (DE wide left, DT wide left, NT Center, DE Right) and see if the guards and center pick up the force to that side.
I just don't see your point on this one DE owners. Rushing four shouldn't put up the kind of pressure that you think it should.
It should put up a sack every once in awhile especially vs. inferior OTs.
tautology
offline
offline
Originally posted by Gyuri1
thehazyone 4 Bort!
I'm joking, as I think that Bort, generally, does an excellent job given the limitation that he is only 1 person, but I'd hope that Bort sees the positive feedback that has come from this round of testing (and a couple of the other, more transparent tests like Hazy's rushing test) and pushes those working on the test server to do a better job discussing and explaining the whats, whys and hows of there tests. I know it's a lot of (unpaid) work, but I think it's well worth it in that it engenders more trust and confidence in the game from the community.
Ditto.
Test Server is currently working as intended, thanks to thehazyone in this case.
Job well done!
thehazyone 4 Bort!
I'm joking, as I think that Bort, generally, does an excellent job given the limitation that he is only 1 person, but I'd hope that Bort sees the positive feedback that has come from this round of testing (and a couple of the other, more transparent tests like Hazy's rushing test) and pushes those working on the test server to do a better job discussing and explaining the whats, whys and hows of there tests. I know it's a lot of (unpaid) work, but I think it's well worth it in that it engenders more trust and confidence in the game from the community.
Ditto.
Test Server is currently working as intended, thanks to thehazyone in this case.
Job well done!
how does it go when you use realistic builds instead of 70/70/70, 85/85/85 and 100/100/100?
i'd like to see something like 90/90/70 (DE) (speed/agility/strength) against 60/90/70 (OT)
i'd like to see something like 90/90/70 (DE) (speed/agility/strength) against 60/90/70 (OT)
Originally posted by Gart888
how does it go when you use realistic builds instead of 70/70/70, 85/85/85 and 100/100/100?
i'd like to see something like 90/90/70 (DE) (speed/agility/strength) against 60/90/70 (OT)
that's the one thing that bugs me about this experiment. very few OT's have 70 speed. maybe in the twilight of their career, but i think by that point the DE usually has either speed or agility well over 70.
how does it go when you use realistic builds instead of 70/70/70, 85/85/85 and 100/100/100?
i'd like to see something like 90/90/70 (DE) (speed/agility/strength) against 60/90/70 (OT)
that's the one thing that bugs me about this experiment. very few OT's have 70 speed. maybe in the twilight of their career, but i think by that point the DE usually has either speed or agility well over 70.
czman
offline
offline
Originally posted by thehazyone
From the test server:
Originally posted by Bort
Some notes I would like to add for consideration as we work/re-look at this:
- I was actually toying with the idea a while back of having blockers have the ability to go the wrong way when trying to cut off a player, or to fall off balance on their own if they needed to make a quick move back over at more than 100 degrees or so. Nobody really seemed to like the idea so I ditched it.
- I can very easily play with the radius for avoidance on elusive pass rush to try different settings, to make the "corner" bigger or smaller.
- I have also thought at times of adding some sort of "is_backpedaling" flag to pass blockers, which would give them a penalty to blocking when backing up.
- Most people have seemed to want us to "let the builds sort it out."
Originally posted by thehazyone
normally I'd agree with letting builds sort it out, but when it takes a 30 difference in three key attributes to become a 15 sack a season pass rusher, then I think some other things need to happen.
I like the ideas you mention - something tautology mentioned too is that the QB backpedals very fast - is there a way you can slow that down?
I did not read the entire thread.
However, the problem with pass blocking is a direct result of protect. Many many moons ago DEs were killing OTs, Bort decided to make changes to slow these DEs down. This made it harder for REs to get much pressure for most teams so they moved their pass rushers to the LE position because RTs do not get protect. For some reason Bort did not want protect to work on the right side. However and first step, tunnel vision all worked for LEs. This made Bort tweak the OT starting positions and and angels.
Ultimately, this entire situation has been brought about because the SAs are not even for OT and DEs. The DE SAs are much better than the RT ones.
Lastly, there are 4 huge mistakes in your test.
1) You left SAs out and like I said pass rushing SAs are better than pass blocking ones.
2) You did not use many blitzes. Most teams blitz almost every down. If teams could get pressure with 4 rushers consistently then they would be able to double cover all of the receivers which would destroy the passing game because man-to-man is way to powerful.
3) Your builds are not representative of most players. An o-line build like that would have great pass blocking but would suck at run blocking. I bet this o-line would have a very hard time run blocking against this d-line.
4) The pass block bonus is reality is greater than the pass rush bonus.
If we look at % you will see what I am talking about.
70*.3=21 or a total of 91
when they were all at 70 they were even at 91.
after the boost the o-line had 70% of the d-line on first down. however, on second down and 10 it was 91%. This basically makes all of the plays on second and 10 not equal to 70/100. I am not going to check how many 2nd and 10s there are but it could easily be 25% of the plays and that is a large enough sample size to skew the stats.
From the test server:
Originally posted by Bort
Some notes I would like to add for consideration as we work/re-look at this:
- I was actually toying with the idea a while back of having blockers have the ability to go the wrong way when trying to cut off a player, or to fall off balance on their own if they needed to make a quick move back over at more than 100 degrees or so. Nobody really seemed to like the idea so I ditched it.
- I can very easily play with the radius for avoidance on elusive pass rush to try different settings, to make the "corner" bigger or smaller.
- I have also thought at times of adding some sort of "is_backpedaling" flag to pass blockers, which would give them a penalty to blocking when backing up.
- Most people have seemed to want us to "let the builds sort it out."
Originally posted by thehazyone
normally I'd agree with letting builds sort it out, but when it takes a 30 difference in three key attributes to become a 15 sack a season pass rusher, then I think some other things need to happen.
I like the ideas you mention - something tautology mentioned too is that the QB backpedals very fast - is there a way you can slow that down?
I did not read the entire thread.
However, the problem with pass blocking is a direct result of protect. Many many moons ago DEs were killing OTs, Bort decided to make changes to slow these DEs down. This made it harder for REs to get much pressure for most teams so they moved their pass rushers to the LE position because RTs do not get protect. For some reason Bort did not want protect to work on the right side. However and first step, tunnel vision all worked for LEs. This made Bort tweak the OT starting positions and and angels.
Ultimately, this entire situation has been brought about because the SAs are not even for OT and DEs. The DE SAs are much better than the RT ones.
Lastly, there are 4 huge mistakes in your test.
1) You left SAs out and like I said pass rushing SAs are better than pass blocking ones.
2) You did not use many blitzes. Most teams blitz almost every down. If teams could get pressure with 4 rushers consistently then they would be able to double cover all of the receivers which would destroy the passing game because man-to-man is way to powerful.
3) Your builds are not representative of most players. An o-line build like that would have great pass blocking but would suck at run blocking. I bet this o-line would have a very hard time run blocking against this d-line.
4) The pass block bonus is reality is greater than the pass rush bonus.
If we look at % you will see what I am talking about.
70*.3=21 or a total of 91
when they were all at 70 they were even at 91.
after the boost the o-line had 70% of the d-line on first down. however, on second down and 10 it was 91%. This basically makes all of the plays on second and 10 not equal to 70/100. I am not going to check how many 2nd and 10s there are but it could easily be 25% of the plays and that is a large enough sample size to skew the stats.
Octowned
offline
offline
The key to DE success is clearly not in the stats but in the SAs.
My DE has 100 agility, 75 strength, 75 speed, 50 vision, 50 tackling and no SAs, and didn't do ANYTHING last season. As a result of such a sick attribute build, his bars were 51-60-57.
One DE had 10X my sacks, 75 last season, with the same level and 42-42-42 bars. Bars are worthless, yes, but they indicate how the player was built. He clearly had 10-10-10-10-X on the bottom tree.
--
This test needs to be run with various SAs on the players, etc., as clearly a DE who can't get a jump off the line, change directions well, shed a block, or zero in on the QB shouldn't be able to get sacks.
Appreciate your work, hope Bort sees it. Frustrating creating a beast of a DE myself and never doing anything in his career.
My DE has 100 agility, 75 strength, 75 speed, 50 vision, 50 tackling and no SAs, and didn't do ANYTHING last season. As a result of such a sick attribute build, his bars were 51-60-57.
One DE had 10X my sacks, 75 last season, with the same level and 42-42-42 bars. Bars are worthless, yes, but they indicate how the player was built. He clearly had 10-10-10-10-X on the bottom tree.
--
This test needs to be run with various SAs on the players, etc., as clearly a DE who can't get a jump off the line, change directions well, shed a block, or zero in on the QB shouldn't be able to get sacks.
Appreciate your work, hope Bort sees it. Frustrating creating a beast of a DE myself and never doing anything in his career.
Last edited Apr 19, 2009 12:35:43
fbjiii
offline
offline
First: Thanks for this and all the other testing. I don't have any doubts that the testing has already led to GLB being a better game. I wish there were more testers tbh, I could think up a list of tons of things I'd like to see tested. There are a lot of things that seem to either not work, or not work well enough to be worth using SP on. Confidence and some SAs
Second: As to the DE problems, I think a lot of it has to do with how fast the qb releases the ball. Even DTs are raising their speed really really high. It's the only thing that works right now. In reality even containment type guys with 60 speed should be able to get to the QB. The sim needs more "coverage" type sacks.
Second: As to the DE problems, I think a lot of it has to do with how fast the qb releases the ball. Even DTs are raising their speed really really high. It's the only thing that works right now. In reality even containment type guys with 60 speed should be able to get to the QB. The sim needs more "coverage" type sacks.
PackMan97
offline
offline
Hazy, I would be interested to see this with say, First Step and Tunnel Vision at very high levels. ROT really has nothing to combat Tunnel Vision or First Step and LOT only has Protector.
It could be that a small edge combined with TV produces the results you want.
It could be that a small edge combined with TV produces the results you want.
Bukowski
offline
offline
Yeah.. another thing is that Protect HAS to be turned down to it's original level.
A couple of seasons ago, when Bort was nerfing the DEs, he gave the Protect SA a boost, and RDEs have been non-existent every since.
That's a definite. There is no reason for it to be worth so damn much at this point. Every LT knows how valuable it is, so they just pump it up to the max, which makes it that much more difficult for any RDE to do anything.
Another thing, like I said, is that Bort really needs to go through the DE SAs (especially Tunnel Vision) just to verify that they are working correctly. It wouldn't be the first time, that an SA was working incorrectly, when we all assumed that it did.
A couple of seasons ago, when Bort was nerfing the DEs, he gave the Protect SA a boost, and RDEs have been non-existent every since.
That's a definite. There is no reason for it to be worth so damn much at this point. Every LT knows how valuable it is, so they just pump it up to the max, which makes it that much more difficult for any RDE to do anything.
Another thing, like I said, is that Bort really needs to go through the DE SAs (especially Tunnel Vision) just to verify that they are working correctly. It wouldn't be the first time, that an SA was working incorrectly, when we all assumed that it did.
russ668
offline
offline
I've been working hard to build an effective DE over the several seasons and spent some cash on this game and I have been VERY disappointed over the last two seasons. So much so that I've considered leaving the game.
It's nice to see Hazy take the time to prove the issue. My DE spent season 7 on the right side and he barely had any tackles and sacks were all but nonexistent. Last season I choose to take a backup role to play on the left side and ended up with decent stats, but honestly, with as much work that has gone into him...he needs to be a starter.
I understand the need to make adjustments in the game to keep things fair, but honestly there is no way our team will ever out level the opponent by 10 or 15 points in all 3 attributes. So unless there is something done this season to fix things I'm afraid it's not worth my time to spend money on a game where there is never any improvement for efforts trying to build a monster.
Please fix the DE problem...we should not have to rely on a gauntlet of CPU teams to pad our stats to end a season with some admirable numbers in the sack and tackle depts.
It's nice to see Hazy take the time to prove the issue. My DE spent season 7 on the right side and he barely had any tackles and sacks were all but nonexistent. Last season I choose to take a backup role to play on the left side and ended up with decent stats, but honestly, with as much work that has gone into him...he needs to be a starter.
I understand the need to make adjustments in the game to keep things fair, but honestly there is no way our team will ever out level the opponent by 10 or 15 points in all 3 attributes. So unless there is something done this season to fix things I'm afraid it's not worth my time to spend money on a game where there is never any improvement for efforts trying to build a monster.
Please fix the DE problem...we should not have to rely on a gauntlet of CPU teams to pad our stats to end a season with some admirable numbers in the sack and tackle depts.
Last edited Apr 19, 2009 14:42:20
Going to run another series of tests with varying DE builds, OT's having less speed, and implementing SA's. We'll see if it makes a difference.
23yrwej
offline
offline
Originally posted by thehazyone
Going to run another series of tests with varying DE builds, OT's having less speed, and implementing SA's. We'll see if it makes a difference.
Make sure to try LTs with Protect. It tips the scales even more.
Going to run another series of tests with varying DE builds, OT's having less speed, and implementing SA's. We'll see if it makes a difference.
Make sure to try LTs with Protect. It tips the scales even more.
You are not logged in. Please log in if you want to post a reply.
