MVP in a timed race game

[Malladrin]

http://twl.trenchwars.org/?module=tw...nMatchID=23236

[Jason]

gophar, the spider lacks recognition for a perfectly good reason. It's the least integral ship in a solid basing lineup. Without a good terr, you're fucked. Without a good team backing that terr, you're fucked. Without good sharks protecting that team that's protecting that terr, you're fucked. Throw in a jav or a wb as your primary offensive, and you're not as fucked. Spiders simply fill in the cracks.

Yeah, you (or someone else for that matter) are going to argue that without solid spiders the team falls apart. To that I say, "I only slightly agree." My point is you can field a solid lineup even if all your spiders are mediocre. Spiders, as a whole, are important, but on an individual basis, meh.

Here's one last thing to throw out at you. Learn another ship if you want some accolades. It's rather apparent that the only reason you're making a case for the spider is because you haven't diversified your basing portfolio.

Huzzah! (I just felt like throwing this at the end for no apparent reason.)

[Cylor]

I say we make the equation a simple inverse equation. We take the times the sharks repelled (i) times the square root of the warbird's pixels (m) plus 52 (the universal constant of terrier burst bounces) divided by the absolute value of teamkills for freq 0 (z).

[Randedl]

i think i said it before:

don't really care about twd, but come twl:

all the people who played, plus the host, vote for a player from the winning team. maybe the host's vote is a weighted a bit more to avoid bias? dunno

twd/base, i'd be for using formulas in a similar way to what i've already proposed

[Cig Smoke]

i think i said it before:

don't really care about twd, but come twl:

all the people who played, plus the host, vote for a player from the winning team. maybe the host's vote is a weighted a bit more to avoid bias? dunno

twd/base, i'd be for using formulas in a similar way to what i've already proposed

Ya thats a good idea, maybe make the host's vote as a tie breaker if its a tie between 2 ppl.

[Force of Nature]

rand or someone else give me some eqns for weasel + lanc.

[Force of Nature]

Currently this is what is implemented into the mvp for the time race after the next update

warbird: points * (1 + (2x + y)/z) * (1 + 0.1x + 0.01y)
x = terr kills
y = shark kills
z = kills

jav: points * (1 + .01x - .1y) * (1 + 0.1z)
x = kills
y = teamkills
z = terr kills

spiders: points * .7kills/deaths * (1 + .01x + .001y)
x = terr kills
y = shark kills

terr: points * 2(x/y) * (1 + 0.05z)
x = kills
y = deaths
z = terr kills

weasel: points * (x/y) * (1 + 0.2z + 0.2a)
x = kills
y = deaths
z = terr kills
a = num of times flag claimed

lanc: points * (x/y) * (1 + 0.05z + 0.001(a + b + c + d))
x = kills
y = deaths
z = terr kills
a = jav kills
b = lanc kills
c = weasel kills
d = wb kills

shark: points * (1 + .05x - .05y + .1z) * (1 + 4(x/a)) * (3 / avgRepelCount)
x = kills
y = teamkills
z = terr kills
a = deaths

[Rinse]

jav: points * (1 + .01x - .1y) * (1 + 0.1z)
x = kills
y = teamkills
z = terr kills

I think jav should be changed to:

jav: points * (1 + .01x - .15y) * (1 + 0.1z + 0.1a)
x = kills
y = teamkills
z = terr kills
a = Multi kills

I'll keep looking through it to see if I can find any other changes I think should be made. An important role of the jav is to kill the cram or stop the attack by taking out all the spiders/sharks and terr. They should be punished more for tk's since they have more control, but should also be rewarded for getting multi kills.

[BLeeN]

Situation for Jav:
100 kills and 20 teamkills. Makes him never MVP. Since the (1 + 0.01x - 0.1y) = 0, while he could had 15 terrkills in the 100 kills.

One thing is for sure, the middle part of every formulas should consist more, than just *, + and -.

I'll look at it more thoroughly when I get home.

[Force of Nature]

Situation for Jav:
100 kills and 20 teamkills. Makes him never MVP. Since the (1 + 0.01x - 0.1y) = 0, while he could had 15 terrkills in the 100 kills.

One thing is for sure, the middle part of every formulas should consist more, than just *, + and -.

I'll look at it more thoroughly when I get home.

These are the stats you have to work with:

wbKill - # of wb kills
javKill - # of jav kills
spiderKill - # of spider kills
terrKill - # of terr kills
weaselKill - # of weasel kills
lancKill - # of lanc kills
sharkKill - # of shark kills
deaths - # of total deaths
score - points
teamKills - # of teamkills
cumrep - # of repels LEFT on shark when dead
avgRepelCount - cumrep/deaths
flagClaimed - # of times flag claimed
kill/death ratio - totalkills/deaths
totalkills - wbKill + javKill + terrKill + weaselKill + lancKill + sharkKill

you are free to make up any other combinations + ratios of these stats.
Multikills are not there because I dont have the time to implement it. It is a bit more timeconsuming to implement. All the above are already implemented.

[bloodzombie]

You must love math.

[Force of Nature]

Ok bleen sat down and decided the previous formulas arent very comparable enough and I agree to an extent. So this is what has been decided upon. If you have any questions all responisibility + credit is shifted to bleen

warbird: .6Points * (.07wb + .07jav + .05spid + 0.1terr + .05x + .06lanc + .08shark - .06deaths)

jav: .5Points * (.05wb + .05jav + .05spid + 0.09terr + .07x + .05lanc + .09shark - .06deaths - .1tks)

spiders: .4points * (.06wb + .06jav + .05spid + .08terr + .05x + .05lanc + .09shark - .05deaths)

terr: .9points * (.06wb + .06jav + .08spid + .1terr + .1x + .06lanc + .09shark - .09deaths)

weasel: .8points * (sum(.09allships) - 0.05deaths)

lanc: .6Points * (.07wb + .07jav + .05spid + 0.1terr + .05x + .06lanc + .08shark - .06deaths)

shark: points * (sum(.09allships) - 0.009deaths - .08tks) * (.4 * (3 / avgRepelLeft))

[lunch3]

warbird: .6Points * (.07wb + .07jav + .05spid + 0.1terr + .05x + .06lanc + .08shark - .06deaths)

jav: .5Points * (.05wb + .05jav + .05spid + 0.09terr + .07x + .05lanc + .09shark - .06deaths - .1tks)

spiders: spiders: .4points * (.06wb + .06jav + .05spid + .08terr + .05x + .05lanc + .09shark - .05deaths)

terr: .9points * (.06wb + .06jav + .08spid + .1terr + .1x + .06lanc + .09shark - .09deaths)

weasel: .8points * (sum(.09allships) - 0.05deaths)

lanc: .6Points * (.07wb + .07jav + .05spid + 0.1terr + .05x + .06lanc + .08shark - .06deaths)

shark: points * (sum(.09allships) - 0.009deaths - .08tks) * (.4 * (3 / avgRepelLeft))

Nice Bleen and FON. Just have a few observations.

First in that last equation you have (3/avgrepelleft). If a person uses almost all their repels all the time that number approaches infinity, plus if someone actually gets a perfect repelleft perfect game that number would actually be infinite, because dividing by zero does that and at very worst will send the bot in an endless cycle of calculating answer and therefore might kill the bot and/or server that is handling the calculation unless it is restarted in due time. But most likely it will just come up with an error. To get rid of that just say (3/(repelleft+0.001)) or something.

Secondly I think tds are far more important than the coefficients I see assigned to them. I see numbers like 0.1 and lower while other ships are 0.05-0.1. A td IMO is worth maybe a 0.8 (because essentially killing terr destroys 8 ships from combat for at least 20 seconds).

Thirdly....ummm very good format, probably just needs fine tweaking after that.

[Force of Nature]

Nice Bleen and FON. Just have a few observations.

First in that last equation you have (3/avgrepelleft). If a person uses almost all their repels all the time that number approaches infinity, plus if someone actually gets a perfect repelleft perfect game that number would actually be infinite, because dividing by zero does that and at very worst will send the bot in an endless cycle of calculating answer and therefore might kill the bot and/or server that is handling the calculation unless it is restarted in due time. But most likely it will just come up with an error. To get rid of that just say (3/(repelleft+0.001)) or something.

Secondly I think tds are far more important than the coefficients I see assigned to them. I see numbers like 0.1 and lower while other ships are 0.05-0.1. A td IMO is worth maybe a 0.8 (because essentially killing terr destroys 8 ships from combat for at least 20 seconds).

Thirdly....ummm very good format, probably just needs fine tweaking after that.

You don't have to worry about the divide by zero. I already thought about it and made it so (3/(repelleft)) evaluates to 4 if avgRepelleft is zero. So you get a bigger bonus.

Secondly, I disagree on the terr. Say it was 0.8. That means 2 terkills are 1.6 thats 1.6 times the points just because you killed the terr twice. So MVPs will be solely decided on who kills the terrs most which is probably not what we want.

[Malladrin]

If possible use a stat for the total reps a shark has done in a game as a factor.