How is "skill" rated?

[AimMe]

dont forget I have to die a zillion times before I make knifekills. Well, thats an overstatement, I do tend to be numer one even with playing only knife, but my ratio will be 1:1 or less then. If I would make 3:1 ratio with knifeonly I would have skill way beyond 2000 me thinks

But I dont give a rats ass about stats (xlrstats anyway), but I do care about fun and breaking knife records, even if it means having only 800 skill score
this week I used !xlrstats for the first time in months I believe

That's normal game, but there is Norma or somebody like him on other team, your (and mine) score can go below 1:3, and that kills skill. But like you, and i guess all knifers, score means nothing to me

But it's still good to know how this works, at least to be able to explain as admin.

[dmmh]

dont forget I have to die a zillion times before I make knifekills. Well, thats an overstatement, I do tend to be numer one even with playing only knife, but my ratio will be 1:1 or less then. If I would make 3:1 ratio with knifeonly I would have skill way beyond 2000 me thinks

But I dont give a rats ass about stats (xlrstats anyway), but I do care about fun and breaking knife records, even if it means having only 800 skill score
this week I used !xlrstats for the first time in months I believe

That's normal game, but there is Norma or somebody like him on other team, your (and mine) score can go below 1:3, and that kills skill. But like you, and i guess all knifers, score means nothing to me

But it's still good to know how this works, at least to be able to explain as admin.

yeah, that sucks
and you are right ofcourse, it is important

[urtdemos]

some conclusions from last few days, when i decided play for "skills".

I think this system is really bad,

its promote passiveness.
places camping and hit and run tactics over rush and keeping pressure
destroys team play behaviour, ("sorry friend i wont help you, too risky")
better for your "skill" is considering your team mates as brainless meat shield than partners
so better to sacrifice, than rescue them
statistically and theoretically "skills" growing better when you end map with 8 frags and 2 deaths than 79:20


imo for faster gameplay much better system would be like:
[frags] / [hours] * [K/D ratio]
(knife frag counts *2)


so, after same playtime
rusher with stats 50 : 25
would be equal to camper with 10 : 1
or knifer 25 : 25
[K : D]

i wonder if its possible to make it like this, at least system counts played hours from last 30 days (but maybe more i can see only last 30)

[SvaRoX]

I understand your concern about the way of evaluating skills. There is no perfect system, the current one used in XLRstats is not that bad IMO...

Of course considering how much you help your team should be part of the skills but it's difficult to evaluate, for example there is no way to know when you heal someone, so as you said "no need to help you" effect.
Being a rusher too, I like the way you want to take rushing and fragging into account. It rewards players who take more risk, and avoid people who get 1:0 at the end of each map to gain lots of skills points.
However I think this system has two big drawbacks :
- at the end it doesn't really reward what you bring to your team in term of score. From your example, the camper having 10:1 is way more efficient than the knifer who have 25:25. Ok it's hard to knife, so 25:25 is quite a good performance, but the goal of TDM is to have more frags than your opponents, whatever the weapon you use and the way you play.
- it doesn't take skills points of other players into account. That's why current XLRstats is not bad and used in other games such as chess to evaluate people. If you play against total beginners, you will have a huge amount of skills point according to your formula, whereas you would have only few more points using the current system, which is more realistic at the end because raping noobs is not hard for you.

Just to be sure, your formula is equal to K^2/(t*K*F), with t = duration of the period and F the weapon factor, right ?

[draqon]

I gonna go offtopic maybe a little but, on this page that svarox gave http://urt.taggedzi.com/weapcompair.php it says for ak103 that it looses accuracy with silencer ? I mean is that mistake or ? wtf, i always play with AK and silencer in belief that i have better accuracy ??

[urtdemos]

Just to be sure, your formula is equal to K^2/(t*K*F), with t = duration of the period and F the weapon factor, right ?

(K*F/t)*ratio

anyway, just wanted to say, current system promote one type of skill (paranoid - coward )
and not include time factor, so after 10 minutes 5:1 (K:D) guy is equal with guy who mange 50:10
but in fact first one is boring camper, second is total butcher who work hard at enemy area

but maybe not equal because: the farther I go, the more probably is death from hand of less skilled quy (respawn-zoom-crouch-i play 1st day)

whatever

I gonna go offtopic maybe a little but, on this page that svarox gave http://urt.taggedzi.com/weapcompair.php it says for ak103 that it looses accuracy with silencer ? I mean is that mistake or ? wtf, i always play with AK and silencer in belief that i have better accuracy ??

silencer not affect accuracy, its nice that someone did some reseach, but tables and conclusions after few fired bullets is not enough, few thousands would be much better

btw simple test, change your croshair to dynamic (one with black parts would be best) and see how its change when you drop/pickup laser (or turn off/on)
and do same thing with silencer

[SvaRoX]

Oops I made a typo in the formula, what I thought was (K^2*F)/(t*D).

Yes I agree on the promoted behavior, no need to rush to be on the top. GG to you for being the number one, by far =)
We could give this formula a try, if someone want to code it (s)he is free to do so (or I will if I get enough motivation). In fact it reminds the awards of XLRstats 2, maybe we could add a "rusher" award base on K/rounds ratio (unless it already exists). Yours is 48 frags/round in average oO - probably one of the highest here...

I gonna go offtopic maybe a little but, on this page that svarox gave http://urt.taggedzi.com/weapcompair.php it says for ak103 that it looses accuracy with silencer ? I mean is that mistake or ? wtf, i always play with AK and silencer in belief that i have better accuracy ??

I know this link, not a bad idea to do such a study but to be honest this one is crap : they evaluated accuracy by firing the whole magazine, who does that ingame ??? Ok I do spam sometimes, but I really don't care about the spread of the weapon after 25 bullets... Sure, the result with AK+silencer is weird. But look, 60% accuracy with AK, not surprising it's so low because the spread increases quickly after some bullets, but everybody using AK knows that it's an accurate weapon after few shots. I would like to see the same study with only 2 or 3 bullets fired.
It reminds me this one : http://forums.urbanterror.info/topic/19 ... nd-spread/. Conclusion of this study : "silencer doesn't improve accuracy". So obviously one of these studies is wrong, maybe both =)
Anyway I've made thousands of frags with a silencer, I still like it, it's a matter of... feeling

[dmmh]

I gonna go offtopic maybe a little but, on this page that svarox gave http://urt.taggedzi.com/weapcompair.php it says for ak103 that it looses accuracy with silencer ? I mean is that mistake or ? wtf, i always play with AK and silencer in belief that i have better accuracy ??

afaik silencer always gives better accuracy...

[AimMe]

Anyway I've made thousands of frags with a silencer, I still like it, it's a matter of... feeling

Yes FEELING! I can play with lr300+laser+silencer and without silencer. I am better without silencer because of feeling. Without it i know what spread i have, can i hit and am i shooting at all. Yes, it is more accurate with silencer, but i hit more without it.
Same is le300/m4 difference, it's all in feeling.

[dmmh]

that's why I am better with sounds on bullet impact then without