Skill tracking

maD_maX_maD_maX_ Join Date: 2013-04-07 Member: 184678Members
edited September 2013 in NS2 General Discussion
I understand that the new skill tracking is not finished, however I have a few questions.

From what servers do they gather data?
-I heard someone say that moded server data is not included.
--If this is true, is there or will there be a whitelist for server mods?
---I play mostly KKG which has admin mod, spectator lights, and ns2 stats
----do these for example disqualify gameplay?

Will we be told how skill is decided?
-I understand that this might encourage people to skill whore.
--maybe skill whoring isn't bad if it encourages proper play...
---- not just K/D

Will servers be able to set their skill indicator based on desired skill level, or will it simply be based on the players in game...
- would be cool to have a progress bar where desired skill is set in yellow and current skill is in green...
--maybe if current skill>desired skill the excess is shown in red...
---thus if there are pub stompers on a rookie server it would show as red

Does UWE stats intend to replace ns2stats?
-will it have breakdowns of individual rounds.
-will it show who and how many times I killed or was killed by so and so
--or will it be simply a mechanic for the server skill bar


My idea for the server browser skill bar(drawn on iPhone)
image.jpg
free image hosting

Comments

  • MestaritonttuMestaritonttu Join Date: 2004-07-29 Member: 30229Members, Reinforced - Shadow, WC 2013 - Gold
    "including time played, score, kills, assists, deaths and match outcome."

    -in super effectives intel video.

    Why the hell would you skill w**re? To get higher epeen? In a game like ns2? Sry too busy building clog phallus in crevice. :P

    I'd prefer only kdr working as the stat. I still stand by my argument that it's the only stat people really care about. Nobody whines about how many games you've helped your team win. Just doesn't happen. But I guess it's an unpleasant truth or something.

    Speaking of assists and kdr:

    Can you please give an option to remove "assists" from score board?
    As a comm it's a HUGE bummer to check kdr to decide who gets the first personal MAC, when the assists are messing up the stats table.
  • BicsumBicsum Join Date: 2012-02-27 Member: 147596Members, Reinforced - Gold
    Taken from here: http://forums.unknownworlds.com/discussion/131978/hive-ranking-system-hacked-already/p1
    SteveRock wrote: »
    Ghosthree3 wrote: »
    The only way I see it being done is have the server send the info with a keycode unique to that server provider. That way it can be controlled which servers (or people) can send data, if you host a public server apply for a code.

    I think this would be the only sure-fire way. We'd need to verify servers though - maybe the community could help with this some how? Like each server needs to get like.. 25 NS2 players to sign an online petition for them, with their steam IDs? Basically just to say, "Yes, this is a good server with good performance that I play on a lot, and they're not running totally game-changing mods, so I'd like my stats on this server to be tracked."
  • Ghosthree3Ghosthree3 Join Date: 2010-02-13 Member: 70557Members, Reinforced - Supporter
    edited September 2013
    Gonna quote myself quoting myself,
    Ghosthree3 wrote: »
    It's not elo doe.

    EDIT: Gonna quote myself from another thread because relevance.
    Ghosthree3 wrote: »
    Did some experimenting with someone, who submitted fake games for some fake steam accounts, based on the experiments we determined the following:


    If the game was submitted (wasn't game mode modded etc)

    Commander games are dumped (or stored for later)

    Kill = 18 per 1
    Assist = 8 per 10
    Death = Skill gained by kills is divided by deaths (obviously 0 is the same as 1 like kd)
    Score = 6 per 10 divided by number of hours (eg. 30 minutes, 0.5 hours, would give 12 skill)
    Loss = Subtract 1/3 of total score

    Skill for game = Total of above

    Player skill = Average skill across games played

    The only flaw I see with it is that you gain the same skill no matter who you play against, good/bad. Not an issue if you play in the same server with the same people every night though.
  • BicsumBicsum Join Date: 2012-02-27 Member: 147596Members, Reinforced - Gold
    edited September 2013
    Ghosthree3 wrote: »
    The only flaw I see with it is that you gain the same skill no matter who you play against, good/bad. Not an issue if you play in the same server with the same people every night though.

    In a perfect scenario it should be straightened out over time. 1 round you play against players below your skill and get a higher skill level for that round, and the next you play against players above your skill, resulting in a lower skill level for that round.

    Of course you could cheat the system by playing on rookie servers only, but then you wouldn't care for matchmaking anyways.
  • RoobubbaRoobubba Who you gonna call? Join Date: 2003-01-06 Member: 11930Members, Reinforced - Shadow, WC 2013 - Shadow
    Bicsum wrote: »
    Ghosthree3 wrote: »
    The only flaw I see with it is that you gain the same skill no matter who you play against, good/bad. Not an issue if you play in the same server with the same people every night though.

    In a perfect scenario it should be straightened out over time. 1 round you play against players below your skill and get a higher skill level for that round, and the next you play against players above your skill, resulting in a lower skill level for that round.

    Of course you could cheat the system by playing on rookie servers only, but then you wouldn't care for matchmaking anyways.

    Whereas in the real world I've got rounds where I got over 1000 skill points and honestly I swear everyone on the server was awful, and I see others have played gathers (where I know full well what the quality of play will be like) and their score is terrible, because they don't have a constant stream of floor skulks feeding them kills constantly...

    There needs to be a modifier in the equation that scales the skill according to the mean or median skill of the server as a whole, such that comp games and gathers give a higher skill rating than green farming on UWE official servers.

    Something like:

    current skill score * (median skill score of the server/1000)

    should do it.
  • BicsumBicsum Join Date: 2012-02-27 Member: 147596Members, Reinforced - Gold
    Roobubba wrote: »
    Bicsum wrote: »
    Ghosthree3 wrote: »
    The only flaw I see with it is that you gain the same skill no matter who you play against, good/bad. Not an issue if you play in the same server with the same people every night though.

    In a perfect scenario it should be straightened out over time. 1 round you play against players below your skill and get a higher skill level for that round, and the next you play against players above your skill, resulting in a lower skill level for that round.

    Of course you could cheat the system by playing on rookie servers only, but then you wouldn't care for matchmaking anyways.

    Whereas in the real world I've got rounds where I got over 1000 skill points and honestly I swear everyone on the server was awful, and I see others have played gathers (where I know full well what the quality of play will be like) and their score is terrible, because they don't have a constant stream of floor skulks feeding them kills constantly...

    There needs to be a modifier in the equation that scales the skill according to the mean or median skill of the server as a whole, such that comp games and gathers give a higher skill rating than green farming on UWE official servers.

    Something like:

    current skill score * (median skill score of the server/1000)

    should do it.

    I would treat gathers seperately.

    Competitive matches don't need a skill system at all, because the teams are already seperated through divisions.

  • VetinariVetinari Join Date: 2013-07-23 Member: 186325Members, Squad Five Blue, Reinforced - Shadow, WC 2013 - Silver
    Roobubba wrote: »
    Bicsum wrote: »
    Ghosthree3 wrote: »
    The only flaw I see with it is that you gain the same skill no matter who you play against, good/bad. Not an issue if you play in the same server with the same people every night though.

    In a perfect scenario it should be straightened out over time. 1 round you play against players below your skill and get a higher skill level for that round, and the next you play against players above your skill, resulting in a lower skill level for that round.

    Of course you could cheat the system by playing on rookie servers only, but then you wouldn't care for matchmaking anyways.

    Whereas in the real world I've got rounds where I got over 1000 skill points and honestly I swear everyone on the server was awful, and I see others have played gathers (where I know full well what the quality of play will be like) and their score is terrible, because they don't have a constant stream of floor skulks feeding them kills constantly...

    There needs to be a modifier in the equation that scales the skill according to the mean or median skill of the server as a whole, such that comp games and gathers give a higher skill rating than green farming on UWE official servers.

    Something like:

    current skill score * (median skill score of the server/1000)

    should do it.

    Or, even better: Make kills versus good players count more than versus bad players. E.g. If I kill you, I get a dozen points, but if I kill a rookie, I get close to none.
  • MestaritonttuMestaritonttu Join Date: 2004-07-29 Member: 30229Members, Reinforced - Shadow, WC 2013 - Gold
    I'm too dumb for this, but could it be like, if you do really well on a rookie server, it just gives that rookie server a "low skill rating", for YOU personally, then if you play on another server and do poorly, it marks that server for YOU as high skill server...

    Of course then nobody could know beforehand if a server is too high or too low for them, because all the scores are personal... Unless there'd be some kinda magic saying that "hey, the guy you're playing with now did really bad on this server, and since he's kicking your ass, you'd prolly do bad on that server too, so imma put that server "high skilled" 4u".

    ...I should be a project leader or something, I'd love to see programmers try to figure out what the hell my assignments mean.
  • RoobubbaRoobubba Who you gonna call? Join Date: 2003-01-06 Member: 11930Members, Reinforced - Shadow, WC 2013 - Shadow
    Bicsum wrote: »
    Roobubba wrote: »
    Bicsum wrote: »
    Ghosthree3 wrote: »
    The only flaw I see with it is that you gain the same skill no matter who you play against, good/bad. Not an issue if you play in the same server with the same people every night though.

    In a perfect scenario it should be straightened out over time. 1 round you play against players below your skill and get a higher skill level for that round, and the next you play against players above your skill, resulting in a lower skill level for that round.

    Of course you could cheat the system by playing on rookie servers only, but then you wouldn't care for matchmaking anyways.

    Whereas in the real world I've got rounds where I got over 1000 skill points and honestly I swear everyone on the server was awful, and I see others have played gathers (where I know full well what the quality of play will be like) and their score is terrible, because they don't have a constant stream of floor skulks feeding them kills constantly...

    There needs to be a modifier in the equation that scales the skill according to the mean or median skill of the server as a whole, such that comp games and gathers give a higher skill rating than green farming on UWE official servers.

    Something like:

    current skill score * (median skill score of the server/1000)

    should do it.

    I would treat gathers seperately.

    Competitive matches don't need a skill system at all, because the teams are already seperated through divisions.

    Fair point. At the minute, gathers are definitely included in the stats, as are games on the NSL match servers.
    They shouldn't be, I agree!
  • maD_maX_maD_maX_ Join Date: 2013-04-07 Member: 184678Members
    edited September 2013
    I'd expect, them to make it a function of skill... Such that points for a kill = (their skill)/(my skill)*(numerical value)... This way a noob who kills a pro gets a lot of points but a pro who kills a noob gets few...

    Examples
    Pro kills noob. 150/1000*8=1.2
    Noob kills pro. 1000/150*8=53
    Noob kills noob. 150/150*8=8
    Pro kills pro. 1000/1000*8=8

    This would encourage people to play against better players as killing a better player yields more, this could cause people to grow and get better as players

  • sotanahtsotanaht Join Date: 2013-01-12 Member: 179215Members
    Roobubba wrote: »
    Bicsum wrote: »
    Roobubba wrote: »
    Bicsum wrote: »
    Ghosthree3 wrote: »
    The only flaw I see with it is that you gain the same skill no matter who you play against, good/bad. Not an issue if you play in the same server with the same people every night though.

    In a perfect scenario it should be straightened out over time. 1 round you play against players below your skill and get a higher skill level for that round, and the next you play against players above your skill, resulting in a lower skill level for that round.

    Of course you could cheat the system by playing on rookie servers only, but then you wouldn't care for matchmaking anyways.

    Whereas in the real world I've got rounds where I got over 1000 skill points and honestly I swear everyone on the server was awful, and I see others have played gathers (where I know full well what the quality of play will be like) and their score is terrible, because they don't have a constant stream of floor skulks feeding them kills constantly...

    There needs to be a modifier in the equation that scales the skill according to the mean or median skill of the server as a whole, such that comp games and gathers give a higher skill rating than green farming on UWE official servers.

    Something like:

    current skill score * (median skill score of the server/1000)

    should do it.

    I would treat gathers seperately.

    Competitive matches don't need a skill system at all, because the teams are already seperated through divisions.

    Fair point. At the minute, gathers are definitely included in the stats, as are games on the NSL match servers.
    They shouldn't be, I agree!

    OTOH comp players in pub games definitely need to be shown as high skill one way or another, though theoretically they should end up with high skill anyway if they play pubs reasonably often.
  • RabidWeaselRabidWeasel Join Date: 2002-11-02 Member: 5337Members
    The fact that you don't get any adjustment for the relative skill of the team you're playing against (and your team, for that matter) kind of makes the whole thing pointless. I'm going to optimistically assume that they're still working on it and the finetuning will come in later because right now the best way to get rated as high skill is to join the crappiest server you can find and stomp the hell out of it.
  • DecoDeco Join Date: 2010-04-10 Member: 71288Members, WC 2013 - Shadow
    Deco wrote: »
    skill = (kills*18 + assists*0.8)/math.max(deaths, 1) + score*0.6/hours
    if loss then skill = skill *(2/3)
    if abs(score) > 1500 or abs(kills) > 250 or abs(assists) > 250 then ignore
    if time as commander > 75% of round time then ignore
    if play time < 90% of round time AND play time < 5 minutes then ignore
    if round time < 1 minute then ignore
    if "GetNumActiveMods" > 0 then ignore ("GetNumActiveMods" appears to be bugged, as some servers with mods are submitting)

    Player skill = average skill over all recorded rounds

  • dePARAdePARA Join Date: 2011-04-29 Member: 96321Members, Squad Five Blue
    edited September 2013
    The problem atm is: UWEs system didnt know who is a pro, a noob or whatever.

    This is cause its new an the needed data is missing and it seems that its based on an relative simple formular.
    Cause this fact a weekend rookie can have the same score like an DIV1 veteran.

    So clanplayers MUST go pubstomping or the system will never know who is a pro.
  • tarquinbbtarquinbb Join Date: 2012-11-03 Member: 166314Members
    edited September 2013
    i'm (also) sad it isn't an Elo ranking system...

    some people get near the top because they kill millions of noobs on newbie servers. that only skews the entire point of the skill rank.

    you should get few points from noobs, loads of points from higher ranking players and invert that pattern for deaths. at the start it's skewed because noobs and veterans are equally placed, but higher points = higher stakes so it gets harder the higher you get.
  • NarfwakNarfwak Join Date: 2002-11-02 Member: 5258Members, Super Administrators, Forum Admins, NS1 Playtester, Playtest Lead, Forum Moderators, Constellation, NS2 Playtester, Squad Five Blue, Reinforced - Supporter, Reinforced - Silver, Reinforced - Gold, Reinforced - Diamond, Reinforced - Shadow, Subnautica PT Lead, NS2 Community Developer
    edited September 2013
    It's a very early prototype so don't fret too much over it. The intended design as I understand is to simply create a warning signal on the server browser of the overall skill level in a given game so that new players wouldn't stumble into games full of hardened veterans (although it doesn't necessarily work well in this way either as some have pointed out). I'm not sure that making everyone's stats available to the public is very responsible though as there's no way to opt out of either the data tracking or sharing right now. The fact that it links to the user's steam profile is also a little shady, although that doesn't seem to even work in all cases.

    Anyway, the tl;dr is that this is a work in progress and the data isn't actually used to do anything yet. Of course feedback is always very welcome especially because there's lots of room for improvement. :)
  • Omega_K2Omega_K2 Join Date: 2011-12-25 Member: 139013Members, Reinforced - Shadow
    On op:
    Yeah, server mods prevent the stats from showing up. Shine Admin does for example. Plus it seems the site doesn't recognize the server port the info is coming from.


    ---

    Just K/D doesn't make a good player. People suiciding on harvesters can be as useful to the team as someone with a good K/D, even more if the good K/D guy is just camping base doing nothing else.

    Generally, measuring indviduals in a team game is hard. Seriously, you can go from anywhere like 50/0 to 20/10 in another match just because of the circumstances of the game. In one you just sit there and baserape newbies, and in the other you actually have to play, like try to save phasegates and potentially dieing a few times in the process, but at least you ensured that your team holds a key position.

    Relative points is a start, but, there is also the same issue. Game is very situational. It technically also needs to take these into account. What about classes/equipment? It's so much easier to get kills as onos, fade or lerk compared to skulk. Or what about Dual-Exo vs JP/Shotgun vs Rifle Marine?
    Or what about a well-executed attack, gorge player goes suicide on a base with a skulk or two, killing the power and maybe some buildings and essentially winning the game or doing a great deal for your team.

    What about coms? So a bad com, that goes and builds turrets in every room, and doesn't support this troops nor communicates them is better because his team still manages to win vs that alien com that actually tries to give/orders support and has a smart build order as well tries to make tactics, but fails because the team is full of bad players that refuse to listen?


    It's not simple. Simple will be a flawed system.
  • MestaritonttuMestaritonttu Join Date: 2004-07-29 Member: 30229Members, Reinforced - Shadow, WC 2013 - Gold
    edited September 2013
    Had this discussion a million times. KDR would be a flawed system, yes. But so would anything else. While KDR is THE MOST RELIABLE stat to tell you're worth something. And it's the stat people care most about.

    About ranking comms, that's easy. You win the game or you don't. Of course you could have something fancy, like counting the odds you got based on the skill of your teammembers.....yeah we're gonna be waiting on that one for a while.
  • tarquinbbtarquinbb Join Date: 2012-11-03 Member: 166314Members
    Had this discussion a million times. KDR would be a flawed system, yes. But so would anything else. While KDR is THE MOST RELIABLE stat to tell you're worth something. And it's the stat people care most about.

    About ranking comms, that's easy. You win the game or you don't. Of course you could have something fancy, like counting the odds you got based on the skill of your teammembers.....yeah we're gonna be waiting on that one for a while.

    what about commander gets 0 point influence while in the comm chair..

    everyone actively playing gets skill ranking points based on position on the scoreboard (MVP's are almost always the better players) / time spent playing.

    i know it can still be abused/farmed - but this system imo is much better than KDR which is a meaningless stat... the best marine in the world isn't going to get a reasonable kdr when he has 0 upgrades and a crap team against a team of fades.
  • maD_maX_maD_maX_ Join Date: 2013-04-07 Member: 184678Members
    My Idea for skill progression bar
    .

    Forgive the drawing, can only do so much with an iphone..
Sign In or Register to comment.