Even, if you don't win the game, or have a Kill/Death-ratio lower than 1, you get rewarded for some sublte things like healing a squad member or assisting a kill by spotting an enemy.
After a game you think yourself: "Damn, I've given everything, I didn't even played this bad. But He was just a bit better than me and now he has 50 points more than me."
The league just punishes everyone who looses, it doesn't matter how good he actually played. Some API for melees or so would be nice, like DoLeaguePoints(), which gives for some actions points. Some admins are testing the comitted scenario (don't say, you don't have the time to do something like this, there are always some thrustworthy persons) and they could say: "Hey, you get for this or this too much points" or "You should do this in another way." (Weird idea) But sven2 wants a league that is some kind of automatic and now, nothing will change at all...
> After a game you think yourself: "Damn, I've given everything, I didn't even played this bad. But He was just a bit better than me and now he has 50 points more than me."
The main consideration against that was that giving points for each game would emphasize total playing time instead of skill. If the league runs for a long time it could get more and more time-consuming to get to the top, not matter how the skill levels are.
Just because it's "fair" it's not equal to "fun". And I think, it should make fun.
>if you play 5 games with one person and you loose all the time.
the games should be written into a matrix who won how many times against someone in which scen.
With this data, you can calculate even everything. We should just collect this data and make a statistic from it (maybe by normalizing some values, calcilating determinants, calculating an asymptote, ....)
fact is - we have to trust something. And that's the certain scenario.
the more a scenario fits into the scheme of "who's the best those will win" (can be calculated bei Eigenvalues) the more worth it does have.
A Scen with a value <0 will be =0 and will be marked as a fun scen or a cheat scen.
> Just because it's "fair" it's not equal to "fun". And I think, it should make fun.
The problem is that it won't be fun if it's not fair, because some people will milk it.
Believe me: you have a list where certain people are above certain other people, you generate a lot of competition, and a part of that is - and always will be - hatred. That's just a part of our age-old programming concerning how we approach competition. And it should really not be news that teenagers are especially prone to resort to it quickly.
We might help the situation by making it our culture to handle each other with respect. Like not looking down on people playing "Etagenkampf". He who is without sin cast the first stone? Just saying.
But now another important questions appears: For which audience do we want to make Clonk? Sure, all the clonk titles were rated 6 by the USK and sure, we all were infants back then. But playing Clonk online is something, where you need to have some social skills, because you aren't this anonymous, like playing some other MMORPG. You see, Minecraft is a game thats as cuddly as our clonks (in this special way, you know :o) ). But the community is way bigger than ours (~200-300 Players online at once) and it seems, that the developer of the game did propaganda, which mostly attracted adolescent or mature persons.
And now theres my true question: If younger players need a system that's more "fun" than "fair" (you don't loose points/get more) and if oldbies prefer a "fair" system, what do you choose?
PS: Oh god... you should put this post somewhere else. :S
>You gain more and more points... whats unfair about that?
Lets say I start playing online today and you started playing online two years ago. How could I ever reach your point level even if I were twice as good (what I surely not am, considering the lack of playing praxis lately :<)?
I like a system, as the current one, that really shows how "good" a player is (in what he is good is another question - aka "i play only my own scenario!!").
Though, I would like that you gain and loose less points if your points differ a lot from the enemie's. So that games that count would have to be among players of equal strength
>And now theres my true question: If younger players need a system that's more "fun" than "fair" (you don't loose points/get more) and if oldbies prefer a "fair" system, what do you choose?
So whats your opinion?
STRENGTH = POINTS / PLAYING_TIME
Also we could check how did player get the points - if second player did not move and attack he was probably meant to be easy points for player, so we disable getting points from killing/beating him. Quite well solution would be, that player gets the more points the more points have got his enemy, while player does lose points if he does not try to fight/help others hard enough.
However my statement is still - leauge is not needed, so why?
we should integrate this into our calculations of the league points.
> we should integrate this into our calculations of the league points.
We spent a lot of time considering this. But with an open league system it's really hard to do as you can't tell if a scenario is different or just a renamed/slightly modified version of another one...
Player is good at a scenario/type of scenarios(wins most of the time), and sees a list "You may want to play with [NAME] as he have got simmiliar win/loss ratio in melees."
We could divide players into ranks in each category, example categories would be:
Magic melee
Weapon melee
Parkour
Crazy parkour
Settlement
Etc.
And give them ranks in each category basing on their winning percentage:
N/A - player does not play this type of games(too much)
Complete loser - <19%
Loser - 19-23%
Terrible - 24-29%
Poor - 30-34%
Very bad - 35-38%
Bad - 39-42%
Below average - 43-47%
Average - 48-52% of victories
Above average - 53-57%
Good - 58-61%
Very good - 62-65%
Excellent - 66-70%,
Awesome - 71-75%
Master - 76-80%
Godlike - >80%
And then there would be few leagues:
n00b - Loser & Complete loser
Novice - Terrible & Poor
Begginer - Very bad & Bad
Normal - From above to below average
Advanced - Good & Very good
Expert - Excellent & Awesome
Master - Master & Godlike
[If neccesary this could be split into more/less leagues]
Player gains points individually in each of these leagues, and is only able to play in his current league, one above and one below. Victory against player from higher rank counts 2x as victory against even player, while victory against someone weaker counts as 1/2 of normal victory.
Win/Loss ratio would be calculated as follows: Victory counts as 100% and loss as 0%, then last 50 game results get averaged. Why 50? Well, it could as well be 40, 60, 100 or 200, but I picked 50 instead of all games, as it could become hard to change rank after >1000 games, which could result in a average player, that becomes master(his last 50 games were victorious in his current league), but still plays within Normal league, which is a bit pointless.
Player starts with all of the last 50 games marked as 50% (not a victory, but not a loss either), and then slowly climbs upwards/downwards moving into other leagues, where players are roughly of same skill as him. So player starts, wins every match in Normal league, then moves quickly into Advanced one, wins 75% of games, climbs upwards to Expert, where he wins only 60% of the time, then after some time he reaches Master league, where he wins only 45% of time, this way he sometimes falls back into Expert league, then climbs to Master quickly, and his overall rank is Expert/Master.
PS.:
We could as well have the leagues based on player rank instead of their level*
*For example: Master league is composed of 5% best players, n00b of 5% worst, Normal of 14% average, and so on. This way when amount of REALLY good players is low more with lower win/loss ratio are pulled into Master league to keep it not-empty.
Also we could make amount of leagues dependent on amount of regulary playing players, so when there are only 20 players they all play together, when there are 40 they are divided into two leagues. Basic principle could be: At least 20 ppl per league. (more or less for different leagues; it's sure thing that there will be more Normal players than Masters.)
> We could divide players into ranks in each category
I don't like splitting the league up. It only serves to make the individual scores less meaningful. Plus we have to abolish the open league for that.
> And give them ranks in each category basing on their winning percentage:
You are basically describing a rougher equivalent of ELO - the different player categories are equivalent to rank score levels, and the ELO formula produces exactly the result you are describing in terms of players going up and down. ELO is just much more refined than 1/2x and 2x, so players won't go up and down forever, but just stay exactly where their skill level lies.
A suggestion I made sometime (and I think Peter had already though of a similar idea back then?) was that the points you win and lose become less when the difference between your and your enemies' points is too high. For example:
Player A: 2000, Player B: 40
Currently: Player A: Win: 0, Lose :alot; Player B: Win: alot Lose: 0
In my system both would win / lose nothing since they are just too far from each other. The intention is the following: You could _always_ host a game with league enabled. Either a player of your level plays with you and you have a more or less fair game (nearly same win/lose gain/loss) or only players far above you join and you wouldn't lose/win a lot of points at all. You wouldnt have to be afraid anymore that a good player with only very few points totally defeats you and you lose all your points or even that a newbie has his lucky day, wins, and you loose like half of your points.
If a player wanted to really get a lot higher with his points he would have to fight players of his own level and not just play games against very low players which he is sure to win anyway.
In that way the system would still show the players current strength (since it is still ELO) but would reduce the danger of totally unfair games a lot, imo
>But still there is problem of best player not able to gain points
The "best" player would have to fight against the players just below him to gain more than a few points, that is correct. But since this system is about comparing player strengths (aka "you are the best player - why do you want to be even better?") it's not too bad imo.
>and bad player quickly jumping to higher rank after lucky battle against strong enemy.
That is not true. In my system not only the gain and loss of the stronger player A is significantly lowered when he plays against a very low-level player B, but player B also gains/loses nearly no points.
Games between unequal opponents would basically "not" be evaluated, since both (also the low-level player!) would have nearly no gain when he defeats the stronger player
I already explained the benefits I imagine to have in this system in the post before - and I think that would at least be better than the current solution :)
Powered by mwForum 2.29.7 © 1999-2015 Markus Wichitill