Warrior General Games

Originally posted by Bort
Originally posted by Deathblade

Originally posted by AngryDragon


Those dead zones for fakes should also be key zones for good tacklers vs poor tacklers. A good tackler should be able to wrapup a HB from this angle and roll him to the ground but a poor tackler should only push him off his path a bit imo.

Oh shit...

Tackling from the front requires mostly strength, with some tackling ability, while tackling from the back requires almost all tackling ability, and little strength, with a gradient scale between them.

Mind = Blown

Actually, would be very easy to implement. Just have a "tackle from front" ability formula, and a "tackle from behind" ability formula, and average between them with a coefficient based on velocity vector differences.

oh snap...toss that into production on the test server bort and see how it does...

Originally posted by jrry32
Originally posted by PackMan97

Originally posted by taurran



Makes sense, but is still not very realistic considering a player fakes once and everyone sees and reacts at the exact same time.

Why not just allow a HB to affect a number of defenders with the same fake, but slightly decrease the frequency of fake rolls?

Let's say you've got three guys, one in front, one to the left and one to the right with the HB heading to the middle. Now. The guy in the middle *should* be harder to fake, because why would he bite on a fake to the left or right, he knows he has his buds there. Ditto for the outside guys, they wouldn't bit on a fake to the outside.

Or a guy breaking one down the sideline, why would a defender bite on a fake TOWARD the sideline? The only fake he'd bite on would be one toward the open field since he'd use the sideline as another defender.

This is why it's tough to get code like this right while making a sim that can complete in a reasonable time frame.

If it were *my sim*, I'd count the side line as a defender and make it so the more defenders in your fake radius, the tougher it is to pull off a fake, but make it so you could fake more than one person a tick.

Well that's adding human intelligence into dots...that'll be hard because it's hard to make the dots aware that they have help on both sides...or else dots in cover 2 would do a better job of playing underneath the WR knowing they have safety help over the top.

They used to do that. People complained, thinking their guy was getting burned or not following their settings, when he was really just playing underneath.

Originally posted by Bort
Originally posted by Deathblade

Originally posted by AngryDragon


Those dead zones for fakes should also be key zones for good tacklers vs poor tacklers. A good tackler should be able to wrapup a HB from this angle and roll him to the ground but a poor tackler should only push him off his path a bit imo.

Oh shit...

Tackling from the front requires mostly strength, with some tackling ability, while tackling from the back requires almost all tackling ability, and little strength, with a gradient scale between them.

Mind = Blown

Actually, would be very easy to implement. Just have a "tackle from front" ability formula, and a "tackle from behind" ability formula, and average between them with a coefficient based on velocity vector differences.

Fuck.

Originally posted by Bort
People complained

Bort - how many more seasons before you quit letting "People" influence you?


The time is now!!

End the madness!

Stop listening to "people" -

Originally posted by Bort
Originally posted by jrry32

Originally posted by PackMan97


Originally posted by taurran




Makes sense, but is still not very realistic considering a player fakes once and everyone sees and reacts at the exact same time.

Why not just allow a HB to affect a number of defenders with the same fake, but slightly decrease the frequency of fake rolls?

Let's say you've got three guys, one in front, one to the left and one to the right with the HB heading to the middle. Now. The guy in the middle *should* be harder to fake, because why would he bite on a fake to the left or right, he knows he has his buds there. Ditto for the outside guys, they wouldn't bit on a fake to the outside.

Or a guy breaking one down the sideline, why would a defender bite on a fake TOWARD the sideline? The only fake he'd bite on would be one toward the open field since he'd use the sideline as another defender.

This is why it's tough to get code like this right while making a sim that can complete in a reasonable time frame.

If it were *my sim*, I'd count the side line as a defender and make it so the more defenders in your fake radius, the tougher it is to pull off a fake, but make it so you could fake more than one person a tick.

Well that's adding human intelligence into dots...that'll be hard because it's hard to make the dots aware that they have help on both sides...or else dots in cover 2 would do a better job of playing underneath the WR knowing they have safety help over the top.

They used to do that. People complained, thinking their guy was getting burned or not following their settings, when he was really just playing underneath.

Well you should have shown the CB dot running over and yelling in the face of the FS dot for not covering over the top well enough...they would have figured it out.

Originally posted by jrry32
Well that's adding human intelligence into dots...that'll be hard because it's hard to make the dots aware that they have help on both sides...or else dots in cover 2 would do a better job of playing underneath the WR knowing they have safety help over the top.

Exactly! That's why this isn't all that simple. Bort is trying to approximate intelligent behavior with fairly straightforward math.

There is a great game design article that talks about AI design and allowing the computer to "cheat". It needs to be smart enough not to do really STUPID stuff, but it also needs to be smart enough to make "intelligent" mistakes. The example given were Chess AIs in which the dumbed down version would be playing a great game and it's "handicap" was 10% of the time it would make a random move. It took all reality out of the game because even a bad player would never make that move, but the programmers didn't know a better way to make it stupid.

That's what Bort needs, a better way to make the defense stupid

Originally posted by PackMan97
Originally posted by taurran


Makes sense, but is still not very realistic considering a player fakes once and everyone sees and reacts at the exact same time.

Why not just allow a HB to affect a number of defenders with the same fake, but slightly decrease the frequency of fake rolls?

Let's say you've got three guys, one in front, one to the left and one to the right with the HB heading to the middle. Now. The guy in the middle *should* be harder to fake, because why would he bite on a fake to the left or right, he knows he has his buds there. Ditto for the outside guys, they wouldn't bit on a fake to the inside, only bite to the outside.

Or a guy breaking one down the sideline, why would a defender bite on a fake TOWARD the sideline? The only fake he'd bite on would be one toward the open field since he'd use the sideline as another defender.

This is why it's tough to get code like this right while making a sim that can complete in a reasonable time frame.

If it were *my sim*, I'd count the side line as a defender and make it so the more defenders in your fake radius, the tougher it is to pull off a fake, but make it so you could fake more than one person a tick.

Well, if you think about this realistically, a player is always aware of two things. a) where he is going, and b) what is happening downfield.

The rotation of this "cone" based on his directional vector should be variable depending on what direction he is running. If he's heading straight downfield, he should be aware of what's going on in front of him ,and what's going on in his peripheral vision. If he's heading straight laterally toward the sideline, the cone should encompass directly in front of him, but also be weighted a little more in the downfield direction.

I'm just saying, it's something that's going to have to be variable based on the direction of travel, not just a static cone that faces downfield at all times.

Maybe i'm putting too much thought into this....

Originally posted by TxSteve
Originally posted by Bort

People complained

Bort - how many more seasons before you quit letting "People" influence you?


The time is now!!

End the madness!

Stop listening to "people" -

But if he stopped listening to people, he wouldn't have listened to any of the ideas that we just had. It goes both ways but I do agree that whining should be ignored.

Originally posted by DL24
Originally posted by Bort

Originally posted by Deathblade


Originally posted by AngryDragon



Those dead zones for fakes should also be key zones for good tacklers vs poor tacklers. A good tackler should be able to wrapup a HB from this angle and roll him to the ground but a poor tackler should only push him off his path a bit imo.

Oh shit...

Tackling from the front requires mostly strength, with some tackling ability, while tackling from the back requires almost all tackling ability, and little strength, with a gradient scale between them.

Mind = Blown

Actually, would be very easy to implement. Just have a "tackle from front" ability formula, and a "tackle from behind" ability formula, and average between them with a coefficient based on velocity vector differences.

Fuck.

To quote Martin Lawrence:

sh!t just got real!

Originally posted by PackMan97
Originally posted by jrry32

Well that's adding human intelligence into dots...that'll be hard because it's hard to make the dots aware that they have help on both sides...or else dots in cover 2 would do a better job of playing underneath the WR knowing they have safety help over the top.

Exactly! That's why this isn't all that simple. Bort is trying to approximate intelligent behavior with fairly straightforward math.

There is a great game design article that talks about AI design and allowing the computer to "cheat". It needs to be smart enough not to do really STUPID stuff, but it also needs to be smart enough to make "intelligent" mistakes. The example given were Chess AIs in which the dumbed down version would be playing a great game and it's "handicap" was 10% of the time it would make a random move. It took all reality out of the game because even a bad player would never make that move, but the programmers didn't know a better way to make it stupid.

That's what Bort needs, a better way to make the defense stupid

Allow dots to get drunk and play hungover?

Originally posted by AngryDragon
Originally posted by DL24

Originally posted by Bort


Originally posted by Deathblade



Originally posted by AngryDragon




Those dead zones for fakes should also be key zones for good tacklers vs poor tacklers. A good tackler should be able to wrapup a HB from this angle and roll him to the ground but a poor tackler should only push him off his path a bit imo.

Oh shit...

Tackling from the front requires mostly strength, with some tackling ability, while tackling from the back requires almost all tackling ability, and little strength, with a gradient scale between them.

Mind = Blown

Actually, would be very easy to implement. Just have a "tackle from front" ability formula, and a "tackle from behind" ability formula, and average between them with a coefficient based on velocity vector differences.

Fuck.

To quote Martin Lawrence:

sh!t just got real!

Not yet. Bort's got test that sh!t out first

Originally posted by Deathblade
Originally posted by entropy

so basically free safeties would be fucked

Free Safeties with 20 strength, taking a Powerback head on...probably.

as they should

Originally posted by jrry32
Originally posted by TxSteve

Originally posted by Bort


People complained

Bort - how many more seasons before you quit letting "People" influence you?


The time is now!!

End the madness!

Stop listening to "people" -

But if he stopped listening to people, he wouldn't have listened to any of the ideas that we just had. It goes both ways but I do agree that whining should be ignored.

"whining" is what takes place when people are left in the dark to make assumptions based on what they perceive.

"constructive debate" is what happens when people have the ability to get factual information and provide feedback to the source.

All i'm saying is that discussions involving a developer or a second party with knowledge from the source, would put a lot of this to rest.

I agree that the chain of complaints starts with the original cry for a nerf. It just sucks it seems to carry so much weight around here.

Originally posted by taurran
Well, if you think about this realistically, a player is always aware of two things. a) where he is going, and b) what is happening downfield.

Everybody except maybe Jim Marshall, for one moment that will live in the annals of NFL history.

Originally posted by jrry32
Originally posted by PackMan97

Originally posted by jrry32


Well that's adding human intelligence into dots...that'll be hard because it's hard to make the dots aware that they have help on both sides...or else dots in cover 2 would do a better job of playing underneath the WR knowing they have safety help over the top.

Exactly! That's why this isn't all that simple. Bort is trying to approximate intelligent behavior with fairly straightforward math.

There is a great game design article that talks about AI design and allowing the computer to "cheat". It needs to be smart enough not to do really STUPID stuff, but it also needs to be smart enough to make "intelligent" mistakes. The example given were Chess AIs in which the dumbed down version would be playing a great game and it's "handicap" was 10% of the time it would make a random move. It took all reality out of the game because even a bad player would never make that move, but the programmers didn't know a better way to make it stupid.

That's what Bort needs, a better way to make the defense stupid

Allow dots to get drunk and play hungover?

ala Favre?