This topic contains 36 replies, has 0 voices, and was last updated by  naka 8 years, 12 months ago.

  • Author
    Posts
  • #53476

    1shybastid
    Participant

    I agree with Boy 😛

    Peanut…Your ALMOST a good player. I wish your anylitical evaluations were as good as the rest of us can SHOOT. 😛 We shoot better then your evaluations.
    As soon as I get my computer realigned……It Crashed.(Thanks for being P’d off at me Nut and Hayt while I lagged) ((Asses))..I shall work on being a better player… and better commentator.
    Im about ready to invite everyone here to Tahoe……..film at 11:00. Galvin gets 1st right of refusal.
    I LOVE this game.
    Oh What was the topic?

    peace never out

    Tahoe Bastid

    #53477

    Acid22
    Participant

    @1shybastid wrote:

    I agree with Boy 😛

    Peanut…Your ALMOST a good player. I wish your anylitical evaluations were as good as the rest of us can SHOOT. 😛 We shoot better then your evaluations.
    As soon as I get my computer realigned……It Crashed.(Thanks for being P’d off at me Nut and Hayt while I lagged) ((Asses))..I shall work on being a better player… and better commentator.
    Im about ready to invite everyone here to Tahoe……..film at 11:00. Galvin gets 1st right of refusal.
    I LOVE this game.
    Oh What was the topic?

    peace never out

    Tahoe Bastid

    The Voice of Reason Speaks 😆 Shy for President!!

    I agree with Laptops in principle. Two points for a hit & one for an assist or injure would cover it IMO. Unskillfull players could not afford to use heavy weapons for long enough to make a difference anyway. Accuracy is the measure of skill & accuracy comes with practice or time played.(It has’nt for me, but I’ll keep trying). Introducing a ranking system that requires a Masters degree in maths to understand seems to be aimed squarely at the top few players & not at noobs who are happy to get a few kills & see their ranking increase…………….we need noobs, they make me look more skilfull. 😀

    I’ll wait to be shot down in flames.

    Aceeeeeeeeeeeeeeeeeeeeed

    #53478

    1shybastid
    Participant

    If I spend all my dough at the beginning of the game,and buy 2 Hogs…I have NEVER won a game if theres good players playing.If I buy NUTHING untill 4th or 5th round THEN buy Hogs..I usually STILL dont win ..only cause I Sux..THATS not the point! [-X Shaddddup Asssid!
    You guys figure out the rank and score.I just wanna aim better.
    I like a bankroll at the end of the game.Better then being broke in rounds 9 and 10.End of story.

    #53479

    BOY
    Participant

    I would love to base all the skills on money. 🙂

    #53480

    BOY
    Participant

    @gcamp wrote:

    Yeah, thats the one.

    If a beginner kills a pro the beginner gets more points and the pro loses more points.
    If a pro kills a beginner then the pro gets less points and the beginner loses less points.
    It’s the diference in skill used. So yes if they are both the same skill an average change is affected.

    When this doesn’t work is when a new player joins who is actualy pretty good. He will take a lot of points of the others until it evens out.

    You also get more points the small the weapon used for the kill.

    Hopefully this will be the last thing I’ll say on it until I have an actual workable equation for you. (which probably will never happen :P)

    The real source of the issue for me is that the algorith assigns a specific amount of skill points for each specific kill/death/game etc….. This sort of seems impossible. There are so very many situations that call for different things.

    *A funky bomb may have to land in one specific spot to kill a guy but yet a baby missile at close range could be an easy kill.

    *Three players might be playing… the two better players manage to shoot each other and kill each other exchanging very few points, but the noob survives and gains a win and skill points though he did nothing.

    so in conclusion, there are a whole bunch of random factors that affect the outcomes, but I tend to think that trying to assign a specifiic value to a single random event makes no sense. Rather take a bunch of random events and together they show you a pattern. The ‘hard’ stats do this quite well, and thereby should be the basis for estimating skill.

    I can tell that you are fond of the system as it is, if my logic doesn’t hold water in your view i doubt you’d think it worth while to change anything. If that is the case it isn’t worth it to think of new ways to calculate. 😉

    Cheers,

    -Boy on vacation :mrgreen:

    #53482

    imported_gcamp
    Participant

    The problem with the people not losing skill when dying is that they reach a point where they are uncatchable skillwize and don’t seen to play anymore.

    #53483

    PeanutsRevenge
    Participant

    I do like what Laptop suggested, but I cannot see it being workable.
    I thought along those lines for a while, but how can u measure ‘accuracy’?
    Sometimes you don’t actually want to touch the player you’re trying to kill, you actually wanna get a riot bomb under them, drop em ready for balls or some scorching, but how do you tell the game that’s what you wanna do?

    I thought about having it so that it’s closest to the ‘A’ marker for more points, however I found that some people adjust for wind using ‘A’, hence the stronger the wind, the further from the marker point they will hit, even if they actually hit exactly where they were actually aiming to.

    It’s possible to come up with hundreds of great ideas and ways to measure skill, but they have to be practical, easy (ish) to impliment and understandable to all (meaning not tooo long to explain to newbs while in game).

    Although asssid (love it shy, you the man for names) was very selfish in his point, his point was valid none the less. We need to keep the newbs coming in and staying, hence their rank must increase at a reasonable rate until they’re ‘hooked’ when a more skill based system can come into effect.
    Therefore I still say that these ideas would probably be best implimented after x No. kills or x % of total kills etc… and then kept simple.

    #53481

    Deathbal
    Participant

    Lappy has the right idea. But there should be a combination of stats that determine skill.

    We will never be able to account for every situation. A Sandhog or Funky bomb will always have a better chance at killing an opponent.

    A player in a better position has a better chance at winning the battle.

    A shielded player has an advantage. And Fuel, Teleport, batts, all come into play.

    A normally good player may waste 8 shots trying to hit a tough shot or someone under dirt, lowering their shots per kill %.

    My point is that you cannot attempt to account for all of this. All you can do is set a standard and eventually it will all balance out. This is why I’ve said to use a combination of stats to determine skill. I don’t even know if that’s possible.

    We have the stats already. Shots per kill, kills per round, round wins, kills/death ratio ect…..package them up for 1 skill rating. See where every player places in each skill, add up the numbers and viola, the lowest number is #1 rank.

    Then use kills to determine rank until the player(s) reaches top 25. Then skill determines rank for them.

    #53484

    BOY
    Participant

    @gcamp wrote:

    The problem with the people not losing skill when dying is that they reach a point where they are uncatchable skillwize and don’t seen to play anymore.

    I’m not sure I quite understand… they die, but they’re not playing?

    #53485

    imported_gcamp
    Participant

    @boy wrote:

    @gcamp wrote:

    The problem with the people not losing skill when dying is that they reach a point where they are uncatchable skillwize and don’t seen to play anymore.

    I’m not sure I quite understand… they die, but they’re not playing?

    You suggested that people don’t lose skill points for dying.
    In that case the skill points only go up, even if they get beaten soundly in every game.

    It sounds like what you are looking for is a skill system that ranks people’s skill regardless of who else is playing. i.e. a measure of peoples skill.

    What we have at the moment is more of a ladder system that ranks people against each other to achieve a ranking.

    I’ll change the skill ranking column to also print out the actual skill for kills (not counting round wins etc..) and see what that looks like.

    #53486

    imported_gcamp
    Participant

    Ok, I’ve changed the stats index to have the kill skill in a darker color. This skill is only calculated using who kills who and what weapon was used. No rounds/wins/shots etc are used in the calculation.

    Kill skill starts at 1000.
    You gain points when you kill someone, lose points when someone kills you.
    You gain more points for killing someone ranked higher.
    You gain more points for using a lesser weapon.

    If we use this only then it turns into a ladder system.

    #53487

    Thrax
    Participant

    @gcamp wrote:

    Ok, I’ve changed the stats index to have the kill skill in a darker color. This skill is only calculated using who kills who and what weapon was used. No rounds/wins/shots etc are used in the calculation.

    Kill skill starts at 1000.
    You gain points when you kill someone, lose points when someone kills you.
    You gain more points for killing someone ranked higher.
    You gain more points for using a lesser weapon.

    If we use this only then it turns into a ladder system.

    So far, this existing method is working somewhat reasonably on
    the mod-servers. If you want to climb rank faster, Kill someone higher
    than you. Wasting on the rookies will gain you little, you have more to
    lose if one of them kills you. (if you ever catch Scar online.. Thump him! :D)
    Pretty much how I always targetted personally.. Kill those of high rank first.

    #53488

    Outer
    Participant

    @gcamp wrote:

    Ok, I’ve changed the stats index to have the kill skill in a darker color. This skill is only calculated using who kills who and what weapon was used. No rounds/wins/shots etc are used in the calculation.

    Kill skill starts at 1000.
    You gain points when you kill someone, lose points when someone kills you.
    You gain more points for killing someone ranked higher.
    You gain more points for using a lesser weapon.

    If we use this only then it turns into a ladder system.

    This only seems to be working for the first ten and only on the main stats page. If see rankings or extended ranking this does not show up.

    #53489

    BOY
    Participant

    @gcamp wrote:

    @boy wrote:

    @gcamp wrote:

    The problem with the people not losing skill when dying is that they reach a point where they are uncatchable skillwize and don’t seen to play anymore.

    I’m not sure I quite understand… they die, but they’re not playing?

    You suggested that people don’t lose skill points for dying.
    In that case the skill points only go up, even if they get beaten soundly in every game.
    I was only saying that they SHOULD not lose points, and no, my idea is that a performance measure would govern points such as points per round, so, in effect players are penalized for each round they play. Their points can only go as high as the possible points per round. In a standard game the max theoretical score would be 11 kills per round. ;).

    It sounds like what you are looking for is a skill system that ranks people’s skill regardless of who else is playing. i.e. a measure of peoples skill.
    Yes, I am looking for a performance measure that best judges the skill, in the same way that a sports team is ranked by some kind of statistic, ie. points per game, or points allowed per game. Players tend to have goals such as “a 10 kill game” or of course Game wins, or high kill ratios.

    #53490

    BOY
    Participant

    @gcamp wrote:

    Ok, I’ve changed the stats index to have the kill skill in a darker color. This skill is only calculated using who kills who and what weapon was used. No rounds/wins/shots etc are used in the calculation.

    Kill skill starts at 1000.
    You gain points when you kill someone, lose points when someone kills you.
    You gain more points for killing someone ranked higher.
    You gain more points for using a lesser weapon.

    If we use this only then it turns into a ladder system.

    Okay, as a sort of fun project, I’ll go a different route, start from the top. 😛 Reading further is NOT encouraged, I strongly discourage everyone from reading any futher.

    @wikipedia wrote:

    A game ladder is a series of ranking levels used to measure playing skill in competitive games. Game ladders are an integral part of the playing experience for many games. The most widely known system of ranking players is the ELO rating system, which is used for Chess and Go. Every player in the ELO rating system receives a rating based on his or her win/loss record, which establishes his or her position (or level) on the game ladder. Numerous efforts have been made to design better game ladders by analyzing the statistical correlation between relative ladder levels and a player’s expected performance.

    A unique game ladder system is the Masterpoints rating system used for contract bridge by the American Contract Bridge League. The Masterpoints system is noteworthy because, unlike the ELO rating system, it emphasizes participation (i.e., experience in terms of number of games played) over demonstration of skill.

    having read the link above concerning ELO rating system, I believe that it supports my claim that a ladder system is innapropriate for scorched3d. The basic explanation is that there is a very slim statistical correlation between player deaths and skill. I can name various examples of how and why, but the most important aspect is that a ladder system is not effective for an FFA situation. The concept behind a ladder system is one VS one competition or team VS team.

Viewing 15 posts - 16 through 30 (of 38 total)

You must be logged in to reply to this topic.