Will ESF ever be noob friendly again?

[Grega]

Who knows. For now we have so many things still going on that we decided not to bring up new ideas for a month so that our current WIP list can get shortened a bit instead of lengthend >.<

 

[underline]

Who knows. For now we have so many things still going on that we decided not to bring up new ideas for a month so that our current WIP list can get shortened a bit instead of lengthend >.<

LOL, well, all i can do is wait now


success

 

[Snowm@n]

Who knows. For now we have so many things still going on that we decided not to bring up new ideas for a month so that our current WIP list can get shortened a bit instead of lengthend >.<

That is a wise move.
Is it of me or are people geting ansious here? :smile:

 

[underline]

That is a wise move.
Is it of me or are people geting ansious here? :smile:

indeed we are,
arent you?

 

[DarktoothDKD]

haha, it seems like it
even Im getting anxious, I cant wait to see ESF near its completion. All I know is
1.3 is awesome on sooo many levels.... >.<

 

[underline]

haha, it seems like it
even Im getting anxious, I cant wait to see ESF near its completion. All I know is
1.3 is awesome on sooo many levels.... >.<

dont tease =O

 

[vegetaspride]

...
There is nothing wrong with the physics engine in HL1. Switching over to Source won't make any difference.

Since there is no full physics engine in HL1 there can't be anything wrong with it. Search for havok or half-life 2 physics differences in google as a start and you'll understand.

And the coders are integrating a physics engine called bullet physics.

That's good news, still I believe it would have been much easier to start with a new engine for the new version.

 

[Tenzo]

That's good news, still I believe it would have been much easier to start with a new engine for the new version.

You see, here's were you're wrong. It will be more trouble to switch to another engine. Do you want this thing to be released soon or not?

 

[Grega]

That's good news, still I believe it would have been much easier to start with a new engine for the new version.

And remake the entire thing from scratch. Starting over with a new engine would mean throwing out years of development out the window.

 

[Renyori]

Since there is no full physics engine in HL1 there can't be anything wrong with it. Search for havok or half-life 2 physics differences in google as a start and you'll understand.

I know the difference in physics with the engine. I'm saying it wouldn't make a diference in the learning curve since the physics engine was never the problem in the first place.

 

[hleV]

So, you can use the stamina to quickly recover from being knocked back and almost hit into the wall/ground?

If so, that's cool.

 

[DarktoothDKD]

and a lot more.

 

[Skiwan]

DT loves teasing <3

 

[vegetaspride]

Unfortunately most of the people does not understand the importance of a physics engine especially in this game. Without good physics the game between a moderate player and the pro is like a movie where the same things repeat itself again and again.

For example the "angling" concept is totally effected by the physics engine and the whole game is based on this. To be honest the way it is in 1.2.3 it "sucks". You see models very far away when they hit each other and the veteran player kicks ass because once it is mastered there is nearly no variaty in outcome and everything including the hit is predictible(the trajectory a model follows after a hit is almost always the same).

Also for a newbie there is always the feeling that one should have hit the opponent while reverse happens and this is also the result of the lack of the physics engine(you need much better collusion calculation for this). The recovery is just as problematic, there is a rolling animation(not simulation!!!) but no matter the case one can recover the same way by pressing space and moving the mouse. A veteran player knows when opponent can not recover and hit again and again since the defender always moves the same direction etc...


Very similar problems were all over the good old street fighter series where the outcomes depended on rocks papers scissors type simple algorithm. Once mastered a new player had no chance even hitting you.

In a better simulation there is always more variaty, and little differences in the direction of hit etc. make a difference in the outcome making it hard for any player to predict. I am not saying experienced lose to an inexperienced player BUT they will trade blows and the inexperienced will surely have a chance.

The ESF team is doing a great job but if you want this game to be popular again 1.3 should better bring drastic improvements other then cosmetics otherwise the game will surely die with the few advanced players...

 

[Renyori]

Yes the person who hits at the better angle will gain the hit. If you're seeing models hit at "very far away" then there's something wrong with your game, or you're playing at very high ping. Veterens don't master whatever you're talking about. They master getting around the opponent, and hitting the opponent at better angles which is what gains the hits.

Beginners always feel they should have got the hit, or it should have been a headon because they don't understand angles. Sometimes angles are so slight but they work.

Red will get the hit. A beginner would think that should have been a headon because they don't notice the angle advantage.

From my years of 1.2 i haven't noticed anything wrong with the "collision calculation". It's accurate, and I really don't believe it has anything to do with the learning curve. The only way to make it more beginner friendly with the collision calculation is to make it UNaccurate so that even if the skilled player gets the better angle, it becomes a headon which would be pretty lame.

The big learning curve problem for 1.2 has always been the movement problem. It was something veterans mastered, and beginners had trouble with. Make it easier and beginners won't feel like a punching bag to the veterans. 1.1 and 1.0 was simple 1 button click to swoop and they were the most popular versions. 1.3 will also be 1 button swoop. Of course there will still be "pros" but atleast the beginners won't feel completely helpless, immobile, then ragequit. Instead, they will have fun (even when dying), keep playing, and become good themselves. Which I believe will make the game super popular.

 

[vegetaspride]

First of all the current engine treats the models as single point objects and the "collusion" give the same result no matter what part of the object it is inflicted. This changes a LOT, rolling , after collusion directions etc.

The engine does not calculate any velocity(or better force inflicted) at the time of collusion. Again big big difference.

The picture you have show is the exact problem of the game. "Right angle" is NOT a right concept the way it is implemented. First of all you have to show who is touching what part of the other player (you did not draw that, clearly you find it unimportant where it is quite vital). After that if engine "decides" (which is another big problem of the game) one player hit the other one then the new trajectory of the player getting hit must be calculated.
Instead no matter what part of the player is hit and regardless of his speed or the strength at the time of collusion, the player is thrown in the direction he is hit at the time of collusion.

Of course you will have cases where both players getting hit but it is not head on.

Even if you treat the models as ragdolls you'll get a lot more accuracy and way different results for seemingly similar cases where you have the same result in the current engine.

Before commenting on this try to understand what I am saying because it is clear that you are not even reading them.

 

[Skiwan]

I dont know where you're getting these assumptions from, because they are not correct.

A player is not a single point object, they are treated as a 6 sided (box shaped) object, this is why if you hit somebody from above it'll play "blowbackdown" animation, when you hit their left side it'll play "blowbackright", what ren is describing is the way that you can only score a hit from the front plane of the hitbox, an angle hit, is a hit that avoids the front plane of the enemy and hits the left or right plane of their hitbox instead. They are not counted as limb shaped hitboxes for a reason, because back in development, that made it near impossible at the speeds involved in ESF to ever hit your enemy when model hitboxes were used, model hitboxes are created by compiling a model. You also say how the engine does not calculate the velocity of a hit, while this may be true in 1.2, it is not true in 1.3, the speed you are travelling when you connect with your enemy has a direct relationship to the distance/speed of the knockback. This has nothing to do with the physics engine itself, but the code that determines what a simple melee hit does, and how the momentum of a character will effect the next, in 1.2/1.1 it was just a value used for the distance and speed of knockback, now, like I say, it takes into account the speed you were traveling before hand.

Besides, it's already been mentioned that Bullet physics engine is currently being implemented.

 

[vegetaspride]

Your explanation actually shows that my observations are correct(which is hard not to do so).

First let me clarify one thing. When I said single point objects, I was trying to imply that the object's trajectory calculation is made as if it is formed of a single point which is the center of the Cube you are using around each model.

You are using the planes of the cube for determination of resulting predefined animations right ? As you know with a simple ragdoll model you don't need to put predefined animations, instead the rotation, rolling etc. are calculated.

As for hit determination I respect your choice of players as simple cubes and it was a good one for the previous version but by todays standarts it is oversimplification, problematic under lag...

This has nothing to do with the physics engine itself, but the code that determines what a simple melee hit does, and how the momentum of a character will effect the next,

This needs the calculation of moment of inertia etc. and this is one of the reasons physics engine is there for.


My observations are based on 1.2.3 again because I haven't experienced 1.3 yet.

Well I will not be around for the next few weeks, I hope 1.3 comes out by the time I return and we would be talking about how great the new version is.

Whatever you do I believe you are doing a great job, have fun and I wish you luck for the next version...

 

[hleV]

There are hitboxes surrounding player's origin (the single point you were taking about). The animations depends on which hitbox was hit.

About rolling, yeah, that doesn't include hitboxes. Anyway, it's possible to do so body parts wouldn't "stuck" in some wall or mountain.

 

[Synth]

There is nothing more fustrating than being angle hit spammed.

It seems that all vets do now is try to set up an angle hit momentum spree.

Sometimes they will do nothing but keep falling back till they see the opportunity to do so.

I'd love to have the ability to quickly swoop (faster than now, because it's rarely fast enough) back or around angle hits, and fling a ki ball in his/her face.