It's effectively the same as what I suggested, but your functions (from HP to points) are linear and I was trying to stay general to account for potential nonlinearities. Put in simpler terms, you have to control for a guy just shooting UMLs at a VF across the map. If there's an engy near the VF, it might never get below 90% health, but you could keep firing and keep raking up points. An elegant way to control for that is to add higher derivatives to your function. So it might be F(HP)=a(HP)^3+b(HP)^2+c(HP)+d for HP from 0 to 100 (or 130).
I think it is not required at all. Missing out on points I felt I deserved is fixed with assists. If I miss a point taking down a barracks I don't care at all so long as it is down. Likely I am assaulting the building with my squad so they will get a point which helps me anyway. There are occassionally grens who are out to grab every single last point that they can, but nobody really cares about the scoreboard. If I am desperate to get that last skill point I'll spawn as Engineer and build something.
I tried different quadratics and some of them gave negative numbers for some values in the domain. Cubic probably solves that though. A power function, one of the form f(x) = a·x[sup]b[/sup] sidesteps the negative number issue. 3 Higher values of b make the function's graph more curved than for lower values of b, and so relatively fewer points are awarded for damage when (the opponent's) health is high and relatively more points are awarded for damage when health is low. The below family of equations uses the ratio between the amount of points awarded for damage when health is at 50% compared to the amount of points awarded for damage when health is at 0% (the "killing shot"). Let 0 ≤ x ≤ 100 (the % of health taken off from 100%) and 0 ≤ y (points you get). Let f[sub]a[/sub](x) = p((1/100)·x)[sup]a[/sup]. So ∫[sub]h_1[/sub][sup]h_2[/sup] f[sub]a[/sub](x) dx will be the total amount of points you get if you increase the health removed from h[sub]1[/sub] to h[sub]2[/sub]. Define p = g(a) such that ∫[sub]0[/sub][sup]100[/sup] f[sub]a[/sub](x) dx = 100 (to make the total amount of points constant regardless of the amount of curvature). Solving for p gives p = a + 1. So f[sub]a[/sub](x) = (a + 1)((1/100)·x)[sup]a[/sup]. Now define a = h(k) such that f[sub]a[/sub](50) = k·f[sub]a[/sub](100) where 0 < k < 1 (the abovementioned ratio). Solving for a gives a = ln(k)/ln(1/2). So let f[sub]k[/sub](x) = (ln(k)/ln(1/2) + 1)((1/100)·x)^(ln(k)/ln(1/2)). The blue line is a linear function (k = 0.5) and the green and orange curves are k = 0.3 and k = 0.15 respectively.
What the fuck am I reading? I do enough integration in my math class, I don't even want to see it in Empires.
This might seem simple to you, but I'm about to ask a REALLY silly question. Please keep in mind that I haven't done calculus in a while, and that math is really not my strong suit. What exactly is meant by f[sub]a, and if it deals with partial differentiation, could you explain to me how it works/give me a link to where I can read up on some stuff? More specifically, I can only follow this to the step where you define p as g(a), and get p= a + 1. I apologize again for my ignorance on these matters.
What the fuck are you apologizing for? I wouldn't call a question about Partial Derivatives to be a silly one, considering that it's second-year college math for most. Not to mention that trying to type out mathematics with anything short of LaTeX is hell. I can't even tell if he's using actual partials as much as just abusing the notation for them. I demand an apology for your preemptive apology.
i care about the scoreboard insofar as it takes 60 points to get 4 powers and even so the extra turret/walls is broken
In this case, f[sub]a[/sub] actually isn't partial differentiation, it's just a normal function like f, g, etc. The subscript "a" is just to denote that this function depends on a as well as x. Essentially I have a family of functions {f[sub]a[/sub](x) : 0 < a < ∞}, each function corresponding to a different amount of "curvature." And for that part, I changed some stuff around before posting and it's not really clear at all anymore. Ignore the "p = g(a)" part and just solve for p (note that p is a function of a) to make the equality "∫[sub]0[/sub][sup]100[/sup] f[sub]a[/sub](x) dx = 100" hold. Then p will be a + 1.
Yeah, unfortunately this is still relatively incomprehensible to me, so I think I'll just slink away.
I don't understand, why you want to choose a nonlinear function. Hitpoints are Hitpoints, regardless of a player dies or not. To your graph: Why should you get more points (y goes up), if you take off more % of Hitpoints? What's the reason? What is a UML? Artillery? I see no problem there... one player attacks, one repairs, that's the game. Otherwise you can say: Two teams (one rifleman and one engineer per team) are shooting at each other, the engineers get points for healing, the riflemen get points for hitting. Nobody dies, the keep up points. But that's the game: if you are healing or hitting you get points for your activities, regardless of an maybe infinite "point farming" situation - which will be a rare situation. If you use weights (a linear function) for every activity, you have an easy and transparent solution (for developer and player). Complicated things do not (everytime) make a game better, so keep it simple.
Yeah, this shit is still important even if we disregard wages. Luckily, I can just build shit and top the charts without actually playing, but it would be nice for the system to just work from time to time. Oh you cute little thing, that's called "subscript." Partial derivatives don't have a monopoly on subscript. I've interacted with no more than a handful of partial derivatives in the last year (or really any calculus) while doing an unhealthy amount of math with plenty of subscripts. By the way, that was really fucking elegant considering the poor kid was trying to type integrals using bbcode. I didn't even know we had superscript and subscript. His only sin was generalizing way too much for unexperienced eyes. P.S. We aren't even dealing with partial derivatives. This is Calc AB shit, brah. Kids take that in high school. I don't even want to know how you got a Harvard interview without basic calc knowledge. Top notch, dude. Seriously, good job. Hitpoints are not hitpoints. I hate to pull a Paradox, but are we even talking about the same game? When I drop a guy from 20HP to 0HP, that's a fuckload more important than from 130HP to 110HP. We must never forget that this point system is about measuring your impact to the team. Killing a guy has a much larger impact than just nicking him. If our point system values those activities equally, then it's a bad system.
Are you trying to tell me something or just insult me right off the bat? If you're trying to actually say something, I'd recommend not sounding like an arrogant douche to start your replies. I don't remember ever calling you a cute little thing, and I'd imagine you to have the same courtesy to do the same. He's talking about integrals and calculus while using the notation for partial derivatives. It's well known that the notation for a function of variables a,x should be f(a,x), and not f[sub]a[/sub](x). And if you're using multiple functions, then you have f(x), g(x), h(x) and so forth. Now if this was a sequence or physics, then it'd be something different entirely. And it is obvious to me that they're not partials. That's why I pointed out the whole abuse of notation part. The point is that it's not clear to others. For what it was, yes it was good. That still doesn't make it easy on the eyes for most people. I wasn't even insulting him, I was just pointing out how it's hell to try and show mathematics without mathematical typesetting software.
Ironically, I just had my Harvard interview about an hour and a half ago with an alumnus who is now a grad student at Stanford in mathematics, and it went extremely well, so reading this (not going to lie) really angered me. I typed up a response that was much much much ruder than the one I'm about to give, but I honestly thought better to post it. So I'll just say this much: I already apologized in advance for my ignorance on these matters, just so I wouldn't have people looking down on my lackluster math ability and writing what you just wrote. Math is honestly not my strong suit; I took AB calculus in 10th grade just to get it over with, but I'm only finishing BC Calculus this year, and I really have forgotten everything after u-substitution integrals. However, not being good at Calculus doesn't really affect my application, as I'm applying for a Classics, Philosophy, and Comparative Lit triple major. Furthermore, I have an 800 on my SAT IIs for math, and a 770 on my Math SAT I section, which is honestly well-rounded enough for probably any school. I don't know why you thought it fit to judge me over my incompetence in math, and I was honestly going to go on a tirade against you for judging me like that, but honestly, this is the Empires forums, and I have since realized that I should have expected to be insulted for this. Next time I don't understand something, I'll just keep my mouth shut.
No, I don't think so. If you drop a guy from 20HP to 0HP you have luck... and if you kill somebody, you have more luck. If you use a system, that supports luck (by giving more points if you kill somebody) I think that's a bad system. THe "Battle Grounds Mod" ( http://www.bgmod.com/ ) supports my mentioned system, and I like it really, because it gives you points for your activities, not for your luck. You should reward the abilities of a player, not their luck.
We're not measuring skill, we're measuring your effect on the team. It doesn't matter if you were lucky and your 20hp happened to kill the guy, your actions were more influential than if you just nicked him. If I'm just rolling an apc through enemy main and a scout happens to spawn in the apc and go on to stun the cv and cause us to win the game, then I did something awesome. It doesn't matter if I did it by accident or not. It happened.
Hm I think I forgot the tickets... a kill should be more rewarded, so it is better to give more reward for killing and destroying than for gathering hitpoints. You have convinced me.
To be honest, I actually forgot about tickets as well, lol. I was just thinking of the potential bad stuff that he could've done against your team if he lived.
