A look at Melee Combat. WARNING LOTS OF MATH!

[Potjeh]

You *will* get punched while you're using unarmed manoeuvres if you attempt to use them, and then it's lights out. It's not hard to see which manoeuvres the opponent is using, and anyone with even a cursory knowledge of the combat system will throw a punch when the moment he sees a shield-bearing enemy using anything but Shield.

[TeckXKnight]

You *will* get punched while you're using unarmed manoeuvres if you attempt to use them, and then it's lights out. It's not hard to see which manoeuvres the opponent is using, and anyone with even a cursory knowledge of the combat system will throw a punch when the moment he sees a shield-bearing enemy using anything but Shield.

Hmm, this tempts me to start using a shield to get my opponent to throw premature punches.

[novaalpha]

they don't just get blown out of the water like a lot of people infer they do.

Feel free to demonstrate this ingame.

And how is all my math wrong.

None of the first three formulae you present match the real numbers. Since all of the followming calclations rely on these 3, all of your math is invalid.

[Foarl]

You *will* get punched while you're using unarmed manoeuvres if you attempt to use them, and then it's lights out. It's not hard to see which manoeuvres the opponent is using, and anyone with even a cursory knowledge of the combat system will throw a punch when the moment he sees a shield-bearing enemy using anything but Shield.

Hmm, this tempts me to start using a shield to get my opponent to throw premature punches.

Or just wear a shield if you intend on having a sword out, hoping to make it look like you might be melee using anything other than block. Would be amusing if they made a mistake that way.

As for khto, I saw 2x(fist) I assumed that meant double weight of UA, didn't realize that was damage only.

The big issue here is that evil eye seems to sink the melee build by preventing any build up in IP for VS. Never used it in combat myself, but I'm assuming you could spam it enough to prevent VS will still flexing/punching. This would put a serious hole in the melee fighter's strategy since melee has no real way to counter this. The cooldown would be worth it for the UA fighter in this case I'd think.

[cloakblade]

You *will* get punched while you're using unarmed manoeuvres if you attempt to use them, and then it's lights out. It's not hard to see which manoeuvres the opponent is using, and anyone with even a cursory knowledge of the combat system will throw a punch when the moment he sees a shield-bearing enemy using anything but Shield.

Hmm, this tempts me to start using a shield to get my opponent to throw premature punches.

See even if this starts happening I've succeeded.

And yes I could see that if you switch maneuvers the opponent will immediately throw a punch (unless below like 50% but that's only 1 flex so almost immediately). But even considering that what I hear a lot is that melee people are going to give their opponent's a 2.0 delta and I mainly wanted to disprove that. The rest was just me messing with numbers.

I'm gonna start testing the build when I start playing more Haven again (too much IRL stuff right now).

@Novaalpha: Are you good at math can I post my basic entire proofs? I'll gladly show all my work.

[novaalpha]

Sure, post them.

[TeckXKnight]

If people start bluffing then you've achieved great success? That is an odd measure of success.
To disprove delta 2, which is the max delta to my understanding now, just post the delta formula then.

[cloakblade]

So the delta formula is sqrt(YourUA/TheirUA) this would allow for an infinite amount of delta as you having 100 UA and an alt jumping at you would have 1 UA resulting in a 10 delta this becomes problematic so Loftar changed it so theirs a cap. But he said the following: (on the 3 page in the announcements)

So is it 1.0 delta multiplier or 2.0?

The cap is 1.0 right now, but it will be 2.0 after the next server restart.

So right now I believe it to be capped at 2.0 sense I assume a server restart is not the same as a world restart.

Now let me share my mathematical proof from top to bottom.
The sum of continuous numbers from 1 to X can be show as: f(x) = 0.5 * X * (X+1)
The first point is free in the skills system in Haven so: s(x) = 0.5 * X (X + 1) -1
In order to find a vertex we need to be able to make there be only 1 instance of X (at least the way I know how):

Code: Select all

s(x) = 0.5 * x(x + 1) - 1 =
0.5 x^2 + 0.5x - 1 =          By distributing the 0.5 * X now we need to create a perfect square so there is one instance of X
0.5(x^2 + x - 2) =            So (X+Y)^2 = X^2 + 2YX + Y^2, so we need the constant (currently - 2) to be equal to sqrt(0.5) = .025, 0.5 being half of the  
0.5(x^2 + x + 0.25 -2.25) =   coefficient of the middle term
0.5(x^2 + x +0.25) - 1.25 =   When moving the 2.25 outside of the parenthesis you must multiply it by the 0.5
0.5(x + 0.5)^2 - 1.25

Now there is only one X so we can find the inverse, the best way I know this is to switch the input (x) and the output (s(x)) and then solver for the output (s(x)). u(x) will be the name of the function so I'll be using it and solving for it here:

Code: Select all

x = 0.5(u(x) + 0.5)^2 - 1.125
x + 1.125 =0.5(u(x) + 0.5)^2
2x +2.25 = (u(x) +0.5)^2
sqrt(2x + 2.25) = u(x) + 0.5
u(x) = sqrt(2x+2.25) - 0.5

Its not the perfect inverse as the original function had an negative half but we don't need that for what we are doing.
Melee combat has half the LP to spend on each skill. In u(x) x is the amount of LP one has so m(x) = u(.5x) = sqrt(x+2.25) - 0.5

If you want I can undo the code I don't know if it helps you read it or not.

Edit: @ TechX: Its was more of a joke to say success but I just want to try and diversify the current combat of pure UA and maybe help see why Melee is lacking.

[novaalpha]

In the first part... you are basically calculating the sum of 1..(skilllevel) but... I don't see where you are factoring in the incremental LP cost itself.

[cloakblade]

In the first part... you are basically calculating the sum of 1..(skilllevel) but... I don't see where you are factoring in the incremental LP cost itself. :? :? :?:

What do you mean? The 0.5 * X(X + 1) is basically doing that.
1+2+3+4+5+6+7+8+9+10 = 55
0.5 * 10 * (10 + 1)= 5(11) = 55

It is of course scaled by 100 which I completely forgot to mention (I'll add it ASAP).

Am I missing something?