@Therius By the same reasoning(skill points = probability of win), it is doing it correctly for alien-marine disbalance. When you play the easier faction exclusively you get higher win chance(and thus more Skill Points). If you play the harder faction you get less probability of win=less SP. If you play half half you get something in between. But yes,I think it probably needs separate skill points for each side, for the purposes of the FET feature. Again average of your marine and alien skill is not good enough.
This is true. This would not be a problem if things like server sizes, faction selections etc. were randomised, since the skill rating would take it into account; one determinant of your probability of winning is the probability of you being assigned to the better team. Your skill level still converges towards the underlying value, as these variables are randomised.
However, it becomes a problem when the player themselves is given the choice. If they want to change their play, for example from a field marine to commanding or from mostly marines to mostly aliens, the skill rating suddenly becomes useless, as it has never had the opportunity to take these variables into account. Over time the player's skill, however, will start to converge towards the new value that takes his skill in this new playstyle account as well. This cannot really be helped, however, as you cannot force people to play every aspect of the game equally. But I think that separate skill ratings for marines and aliens would be so easy to implement that it certainly could not hurt.
Hmm. I am getting tired of this thread. I wonder if it would be worth while and possible to do another alternative skill and balancing system as a mod (and let CDT concentrate on possibly more important things). I have seen few of those, but they are usually bound to a single server....
I mean IMHO the current concept is sound (grade players then assign players to teams based on your grading). But it is those little oversights like this one, that makes lot of people hate it.
1) Are marines and aliens balanced, i.e. can you skew your skill by preferring one faction over the other?
2) Does the standard deviation of skill ratings affect the probability of an outcome controlling for the average skill of a team, i.e. does one pro and a bunch of rookies equal a team of mediocre players even if the average skill ratings of the team are the same?
3) How much does the player count affect gameplay, so if you switch from playing exclusively on 42-player servers to playing exclusively 20-player servers, is your initial skill rating at all comparable to people who have already played on these servers for a long period of time?
I have talked to moultano and he plans to do something for #1. He is a busy guy, with a job and a baby, so I do not know when he will get around to it. He has said in this thread and to me in other conversations that he plans on separating marine and alien hive skill. I don't think there is much to discuss on this one anymore.
#2 is a problem there seems to be some disagreement on how to handle.
Lets say your server is as follows. How would you distribute these players?
skill 3000 : 1 player
skill 1000: 8 players
skill 0 : 5 players
In my opinion rookies at best follow orders but struggle to lose most engagements, at worst a rookie can be so bad they might as well not even be on your team. So I would distribute as follows:
Team A: 3000, 1000, 1000, 1000, 0, 0, 0 = 6000 total team score
Team B: 1000, 1000, 1000, 1000,1000, 0, 0 = 5000 total team score
Essentially distributing the rookies as evenly as possible after the after every non rookie has been placed.
keep getting conflicting information about how rookies should be handled. People say the skill value of 0 for rookies is too low, but they also say that the team with more rookies almost always loses. These things can't be simultaneously true! I'm inclined to not try to special case rookies in any part of it, but just to handle them as players of low skill. The skill values of other players should gradually rise and fall against whatever value we pick for rookies since there's a continuous stream of them coming in, so I suspect there isn't too much improvement to be made there.
#3 is a difficult one. I think this is where much of disagreement over if hive works or not comes from. I play on a very small selection of servers, 1-3 servers, with usually the same group of people and little variation. Within those servers who have a bunch of regulars I see FET work extremely well producing very balanced teams. I also think that people who don't think FET works at all play on servers that have a lot of mixing of players from a wide range of servers. I have occasionally ventured out into other servers and I do see my hive skill does not seem to be comparable to other tight knit communities. This is in part because a person who plays on a 42 person server can get more skill points in a round than a person on a 16 person server.
Hmm. I am getting tired of this thread. I wonder if it would be worth while and possible to do another alternative skill and balancing system as a mod (and let CDT concentrate on possibly more important things). I have seen few of those, but they are usually bound to a single server....
I mean IMHO the current concept is sound (grade players then assign players to teams based on your grading). But it is those little oversights like this one, that makes lot of people hate it.
I don't think having an alternative balance system on a lone server would work very well. The FET already works great on lone servers where most players are regulars.
In my opinion rookies at best follow orders but struggle to lose most engagements, at worst a rookie can be so bad they might as well not even be on your team. So I would distribute as follows:
Team A: 3000, 1000, 1000, 1000, 0, 0, 0 = 6000 total team score
Team B: 1000, 1000, 1000, 1000,1000, 0, 0 = 5000 total team score
Yeah, as I said before the SP sum is a much better balancing measure than avg(of non-linear Skill Points grade). But it is not perfect. It too does have rough edges for current Skill Points. Consider case of one 3000, three 1000 and the rest 0. I think the 3000 guy will have rough time being at three places squashing those 1000 players. For 42 players server the sum starts to downplay skilled players and could assign significantly more to one team than to the other. We need experienced statistician to tell us which measure is ment to be used for our case! I don't know, maybe some geometric mean is in order - since skilled player possibly sort of a multiply skill of players around him and bad players sort of a limit even skilled players possibilities in the game. Maybe simplest solution would be for the system to only split equally skilled players to teams and not trying to average at all if possible(if it has two 3000 players available one goes to one team and the second to the other). I proposed in some previous post algorithm where one team would get best player and the second team second best player. Then the teams would switch pickings and the second team would get best player from the remaining and the first team the second best and so on untill everyone is assigned. Of course it does not work well too if the player base distribution is weird like above, but at least you dont get bunch of rookies like for the avg.
0 for rookie players is mathematically speaking complete nonsense. That would mean they are infinitely worse than even 1 SP player. Something in 200-300 range in current SPs is I think adequate. The hive could perhaps observe new players and come up with some experimentally verified number to assign to new players.
In my opinion rookies at best follow orders but struggle to lose most engagements, at worst a rookie can be so bad they might as well not even be on your team. So I would distribute as follows:
Team A: 3000, 1000, 1000, 1000, 0, 0, 0 = 6000 total team score
Team B: 1000, 1000, 1000, 1000,1000, 0, 0 = 5000 total team score
Yeah, as I said before the SP sum is a much better balancing measure than avg(of non-linear Skill Points grade). But it is not perfect. It too does have rough edges for current Skill Points. Consider case of one 3000, three 1000 and the rest 0. I think the 3000 guy will have rough time being at three places squashing those 1000 players.
0 for rookie players is mathematically speaking complete nonsense. That would mean they are infinitely worse than even 1 SP player. Something in 200-300 range in current SPs is I think adequate. The hive could perhaps observe new players and come up with some experimentally verified number to assign to new players.
What if I told you there were players with 50+ hours recorded in hive, that is actually time in a recorded game, and those players still had 0 points. They are so low skilled they can not win enough games to get above 0 skill points. There are many many players with tons of hours recorded in hive with skill points at or near 0.
So are you suggesting the minimum be higher than 0?
What if I told you there were players with 50+ hours recorded in hive, that is actually time in a recorded game, and those players still had 0 points. They are so low skilled they can not win enough games to get above 0 skill points. There are many many players with tons of hours recorded in hive with skill points at or near 0.
@Nordic Even AFK players should not have 0 (they still can posibly waste some skulks time). Maybe AFK guy obstructing Phase gate deserves 0? He would have to actively run in front of you to prevent you to shoot grenades and kill yourself. He would have to spam console and mic with JB songs and spam commander with med and ammo requests. He would have to play in such a way to make other players so sick they would be unable to play. He would have to run around armory to prevent you to heal. He would have to actively obstruct you from escaping and defending enemy structures with his body. He would have to have pink armor with other gross things painted on it. He would have to provide actionable inteligence to the enemy. He would have to ensure that his team of best players on hive would loose every time to worst players. It should require quite a lot of skill to get 0 SPs. Infinite skill to be precize. Because 1 SP player is still mathematically speaking infinitely more skilled than 0 SP player.
What if I told you there were players with 50+ hours recorded in hive, that is actually time in a recorded game, and those players still had 0 points. They are so low skilled they can not win enough games to get above 0 skill points. There are many many players with tons of hours recorded in hive with skill points at or near 0.
That's the saddest thing I've heard all day
The SPs actualy run into minus values. they are just rounded to 0 after. So it is not unthinkable... I think the guy you mention could still have reasonable K/D of 2/10 or something like that. Maybe he plays on touchpad, I dunno...
Not a guy, there are many with a skill of 0. Their K/D's are about .5 and their W/L is about.75. This is just the bottom 0.001% of the skill spectrum though.
I do not know how they play so poorly, and I don't understand how the very top of the skill spectrum plays so well.
My point is there are people with significant amount of playtime who are just that bad. I think 0 is fine.
I think your complaining about how the skill points are not linear but I don't think that matters either.
Cannon_FodderAUSBrisbane, AUJoin Date: 2013-06-23Member: 185664Members, Squad Five Blue, Squad Five Silver, Reinforced - Shadow
I would actually like to know if we could pull some stats out of HIVE. I would like to know the following:
All data should be from recently active players if that is possible, so we get a snap shot of the present player base:
So data should be from Active players in the last 30 to 60 days
Would like to know:
1. average skill (total skill / players) - what is considered to be the average skill rating atm
2. median skill (the person with equal players above and below them) - what is the distribution of skills compare to average, it should tell you if there are just more high skills averaging out the lower skill people playing on white listed servers.
3. distribution of the skills across the board - how many players are skill 1000-1100 etc... The slice of the band can vary, but will give you an idea of how skill is distributed atm (it is like a more detail look of median skill above.
I think knowing this may help explain why FET sometimes just doesn't balance teams the way it ought to.
@moultano, I think we have come to some good ideas here recently. They are discussed in greater detail above, but I will summarize here.
1) Have separate hive scores for individuals based on marine skill and alien skill.
2) Treat rookies differently during a force even vote.
a) Define rookies so that they can be separated. One suggested definition was 35 hours or less playtime recorded in hive.
b)Balance teams based on hive scores for everyone excluding rookies.
c)Distribute rookies as evenly as possible among marines and aliens as if they had no skill.
3) Start rookies at a higher hive skill than 0. Possibly something 1 standard deviation below the mean hive skill.
I keep getting conflicting information about how rookies should be handled. People say the skill value of 0 for rookies is too low, but they also say that the team with more rookies almost always loses. These things can't be simultaneously true! I'm inclined to not try to special case rookies in any part of it, but just to handle them as players of low skill. The skill values of other players should gradually rise and fall against whatever value we pick for rookies since there's a continuous stream of them coming in, so I suspect there isn't too much improvement to be made there.
In regards to my third suggestion quoted above, I think the problem is not rookies starting at such a low score but a convergence one. If you look at the skill distribution there is a large population gap between players under 200 skill and players at 1000 skill. Basically I think that hive scores do not grow quick enough.
I still think rookies, or people of extremely low skill should be sorted differently.
I still think rookies, or people of extremely low skill should be sorted differently.
That's discrimination :P
But yes the new players should be assigned more sane value, so they converge their Skill to their skill, more quickly. Any trick to use them differently while they coverge may help, but would be unnecessary I think, and perhaps dangerous (it changes how all players are rewarded with SPs). But I like your idea of putting equal number of rookie players(It should be weighted by number of games the system trained on them already too) to each team, so the risk, that they get scored unfairly, is equal to both teams. Of course the problem remains when FET is not used.
EDIT: I just thought of another unrelated minor problem. When people have shared Steam account(e.g. Gaming Cafés, families), their Skill is just changing as people do change on the PC. (Hmmm I wonder if cyberCafées buy this game...)
moultanoCreator of ns_shiva.Join Date: 2002-12-14Member: 10806Members, NS1 Playtester, Contributor, Constellation, NS2 Playtester, Squad Five Blue, Reinforced - Shadow, WC 2013 - Gold, NS2 Community Developer, Pistachionauts
Sorry I haven't been too active on this thread, I've been super busy lately. There are some concrete changes I'd like to make based on the feedback here as soon as I'm able. In the meantime, I'd like to leave you with a quote.
Sorry I haven't been too active on this thread, I've been super busy lately. There are some concrete changes I'd like to make based on the feedback here as soon as I'm able. In the meantime, I'd like to leave you with a quote.
We know you are busy with job, new child, and any other plans you may have. I hope to drive discussion far enough where we can have some really quality feedback for you. So far this has been going well. I will make sure to highlight the best feedback for you here or in slack.
I've listened to all arguments thus far, and the logical side of me knows what you are saying, but really I hope someone can just suspend their belief in the skill system for a moment just to be open to the idea that there might be something not quite right.
ShellMo is considering leaving the game because its totally demoralising to be playing in a server with an average skill level of about 1500, when she feels like she is contributing just as much as everyone else, everyone on the server believes she is, the skill board suggests she is as well, yet the skill system is saying she is barely above a complete novice.
Here is an example of a practise match we played last week
Here is her profile again, http://hive.naturalselection2.com/profile/83013717 down to 304 now. This cannot be correct, I am in no way 5 times better at the game... The skill system at the moment literally thinks it would take her combined with another skill level 3100 player to balance out against 2 players of my skill level..... I play sat right next to her, I watch her play the game while i'm waiting for a slot on the server. Something is definitely not right.
... I had a dummy account lying around. And I always assumed that rookies started off at 1000, this account started at 0. I think depending on when you key was activated there were different starting hive scores. Anyway. Took me 300 hours to climb to 1000. And that was one brutal climb... Fricking climbing in 30ish point increments.
I saw some players with like 10 hours on hive and 1000 points. I think my primary also started at 1000.
Just thought I would update everyone on my real world findings thus far.
ShellMo (who was at a skill level of 300 ish) has now got her skill level up to 1022 super fast. How did she do it? She just started playing aliens more often.. She used to play roughly 2 rounds of marines for every 1 round she played aliens. Now she is doing it the other way around, and within the space of 30 games her skill level has more than tripled.
With this knowledge I decided to see how well I could do with gaming the system, and i've realised my skill level is far more dependant on picking the right team to join than it is on my own performance. Here is how you prop your skill up.
Simply: If marines are stacked join marines, otherwise join aliens.
Using this my skill level has gone from 1600, to 2000ish.
This works because if the teams are even you have roughly a 54% of winning on aliens, and 46% of winning on marines. It also doesn't matter if the teams are stupidly stacked. You still get points from winning a stupidly stacked almost impossible game, and you still lose points from not being able to defeat a team nobody would expect you to beat.
2.78 win/loss ratio. If you play even games one would expect a win/loss ratio of 1, a bit like flipping a coin. So that player has either won a ton of unfair games, or their skill is actually significantly higher and the skill system has failed to converge on the real high value after 382 hours of recorded gameplay.
Now you would assume the system would self correct, in that if you artificially boost your skill, you will be placed with less skilled players next time and find it progressively harder to win. However due to the way the team skill is an average i've not noticed any negative consequences at all. I probably wouldn't until my skill level became hugely higher than average.
Basically you just have to join the right (stacked) team. =D>
It's true that if you play well (doing the right thing at the right moment, listen to commander, etc.) but ultimately loose; your score goes down. I always wonder why a system that tries to measure your personal level uses team results as a corner stone. There are many other ways.
I always wonder why a system that tries to measure your personal level uses team results as a corner stone. There are many other ways.
The point is to reinforce and encourage certain types of behavior, in this case team work and other things that leads to wins, as those are generally what you want to see. You could measure personal skill by something similar to K/D, but that would encourage frag hunting at the expense of the teams objectives.
Every system probably has an angle that encourages poor behavior though - i.e. stacking. But that doesn't mean that it isn't worth while to, or at least attempt to, encourage good teamwork.
The point is to reinforce and encourage certain types of behavior, in this case team work and other things that leads to wins, as those are generally what you want to see. You could measure personal skill by something similar to K/D, but that would encourage frag hunting at the expense of the teams objectives.
Every system probably has an angle that encourages poor behavior though - i.e. stacking. But that doesn't mean that it isn't worth while to, or at least attempt to, encourage good teamwork.
You still can loose with perfect teamwork. All good actions are hidden behind a single 1 or 0 which defines the outcome of the game. Tada!
Biting RTs, Building, weld/healing are part of the job non the less. It requires organization and potentially more skill sometimes.
The point is to reinforce and encourage certain types of behavior, in this case team work and other things that leads to wins, as those are generally what you want to see. You could measure personal skill by something similar to K/D, but that would encourage frag hunting at the expense of the teams objectives.
Every system probably has an angle that encourages poor behavior though - i.e. stacking. But that doesn't mean that it isn't worth while to, or at least attempt to, encourage good teamwork.
You still can loose with perfect teamwork.
Ok, I didn't drive home the point well enough, I admit my example with team work clouded it a little bit.
With a win-based skill system, like this, anything that results in a win is encouraged - whether that constitutes team work or somebody carrying the whole team. Either way, that is what you want - in contrast to people running around welding irrelevant things or frag hunting, instead of managing the resources, for points.
The skill system does not predict your W/L metric to converge towards 1. The skill system does not force you to play even teams, even if FET was used, because you can always switch teams and stack after the vote has passed. However, it doesn't prove that the skill system is flawed if someone is consistently winning more games than they are losing; those players will lose much more points from every defeat, even if they are rare. You can have two players with exactly the same skill rating, and correctly estimated to boot, who have wildly different win-lose ratios, and there's nothing fishy about that. If you like stacking, you win more games, but you are rewarded less for every victory. If you like anti-stacking, you win less often, but you are punished less for every defeat. It might seem like someone can just farm small increments of points from stacking in a way that guarantees their victory, inching up their score with minimal effort, but there's always that small chance of a defeat lurking there, and when it happens, their score will come crashing down.
Aliens and marines should have separate skill measurements, that is true.
Ok, I didn't drive home the point well enough, I admit my example with team work clouded it a little bit.
With a win-based skill system, like this, anything that results in a win is encouraged - whether that constitutes team work or somebody carrying the whole team. Either way, that is what you want - in contrast to people running around welding irrelevant things or frag hunting, instead of managing the resources, for points.
Still... Stack wins. The worst that could happen regarding the situation on the servers right now. It has come to a stupidly insane level. And when i say "stack", i don't mean Skill number. God like aim that has water between the ears is still useless. I dropped by Wooza yesterday (though i avoid it naturally) : it's just a mess. It's not the only one of course, it's just an example.
Fast list :
-Can be compared to a blender that crush rookies.
-FET that just can't be true as it gives stacked games to a level unknown to me so far. I was right to avoid it.
-Any OK player getting on the rookie side, faces humiliation like if somebody was throwing eggs at him. Not to mention racism/nationalism, i was surprised to hear what i heard.
-The slot number acts like a magnet for these new players.
-Their throats sound like a sad flute as the blade slides.
-"Good bye fresh meat".
How would you quantify field calling or overall communication skills?
Right now: Useless, putting aside the language "obstacle". These new players are just canon fodder. They probably think the maps are designed like a tube and they always push in one direction, endlessly. They don't think the opponent can be smart and set up traps. Even after 10 tries... Giving advices and directions to them is useless. There is a dark energy that prevents any good tip to get stored in their heads.
And when you suggest (ready room) to train/teach them (and take the time to do it) : Nobody answer on the rookie side. But the best is that no Stacker do try to do the same as they want their fresh meat of the day. "It's NS2" they say. Childish to say the least. While some of them are supposed to be way beyond that.
The rookies should play Combat instead. It would at least look like what they know, then, maybe they can try NS2. It's truly leveling from the bottom.
But i see where you're going. It is still measurable. Time passed using the microphone is something that can be sampled. Can be faked but it would be a little ridiculous. It would look like a tea time full of people who make small talks... Who said "TS" ?... Girls... who knows...
You have to learn what the term "on average" means.
You should really take the time to look at the 'on average face' on the servers (concerning FET and other things). Right now, it's like a Picasso. And i like Art.
Why do I always feel like this is utter nonsense being issued by all the "stack" complainers. Quite literally NO ONE truly communicates beyond "2 marines logistics" and the like. No one gives tips, no one explains the reasons behind decisions. Obviously my argument here is logically flawed because I can't be on every server all the time to witness the lack of communication, but it happens so infrequently at least on EU servers, that it really bothers me.
Comp players don't often give advice. I do, ALL the time, I frequently point out shit plays, and frequently applaud good plays. I give explanations, I give field comm advice, I communicate with my team, and I get a reception! People DO communicate and DO learn and DO participate if you actually give a damn and try.
Pub stars often blame stack and better players but still dont give advice or field comm.
Pub regs often blame stack and dont know hot to give advice or field comm.
Rookies just trial and error to the best of their own ability to improve.
The issue most pub games face is a complete and utter inability to teach and learn and instead to just blame. I'm so tired of people complaining of stacks and unfairness because it's the complaining that ruins the games. The psychological defeat before the game starts puts teams at such a disadvantage. They give up and see themselves trapped and often suicide. Suicide is actually an adept analogy to draw parallels too, where people often feel so trapped and as if there is no way out that they give up. People who see this attitude need to man the fuck up and help them by communicating and problem solving through voice comms.
The games I enjoy most now are those when I lose because I'm being shut out the game by awesome ambushes, traps, split pressure, things where the opposite team are clearly engaged in the game instead of mind numbingly W + lmb, die, and complain.
Everyone has a duty when they join a team on a server to commit their efforts to winning, not complaining. If it's too much effort for you to communicate / trap / play to the best of your ability and the opponents don't find it too much effort, you're going to lose. Deal with it. If you don't try, you can't expect the other team to not try as much as you just to give you a good game.
FrozenNew York, NYJoin Date: 2010-07-02Member: 72228Members, Constellation
I agree with all the aliens and marines needing different skill sets and therefore different rule-sets, but I also feel that the only way to make it actually work is to set
1 match = 1 marine round + 1 alien around
and require that. Since there's not enough players to make that a requirement with punishment for game-leaves, how can we try to see it that way without forcing 2 round matches for skill-ranked games? Place a metric on rounds played per side affecting your point gains? I don't have any great ideas for it.
Why do I always feel like this is utter nonsense being issued by all the "stack" complainers. Quite literally NO ONE truly communicates beyond "2 marines logistics" and the like. No one gives tips, no one explains the reasons behind decisions. Obviously my argument here is logically flawed because I can't be on every server all the time to witness the lack of communication, but it happens so infrequently at least on EU servers, that it really bothers me.
Comp players don't often give advice. I do, ALL the time, I frequently point out shit plays, and frequently applaud good plays. I give explanations, I give field comm advice, I communicate with my team, and I get a reception! People DO communicate and DO learn and DO participate if you actually give a damn and try.
Pub stars often blame stack and better players but still dont give advice or field comm.
Pub regs often blame stack and dont know hot to give advice or field comm.
Rookies just trial and error to the best of their own ability to improve.
The issue most pub games face is a complete and utter inability to teach and learn and instead to just blame. I'm so tired of people complaining of stacks and unfairness because it's the complaining that ruins the games. The psychological defeat before the game starts puts teams at such a disadvantage. They give up and see themselves trapped and often suicide. Suicide is actually an adept analogy to draw parallels too, where people often feel so trapped and as if there is no way out that they give up. People who see this attitude need to man the fuck up and help them by communicating and problem solving through voice comms.
The games I enjoy most now are those when I lose because I'm being shut out the game by awesome ambushes, traps, split pressure, things where the opposite team are clearly engaged in the game instead of mind numbingly W + lmb, die, and complain.
Everyone has a duty when they join a team on a server to commit their efforts to winning, not complaining. If it's too much effort for you to communicate / trap / play to the best of your ability and the opponents don't find it too much effort, you're going to lose. Deal with it. If you don't try, you can't expect the other team to not try as much as you just to give you a good game.
You may have mistaken me for someone else as I probably should have used 'has become' instead of 'is'. It's been since NS2 came out that i do 'communicate'. In fact it's the same microphone since NS1. Now wonder I have a crappy voice.
I just grown 'fed up' to try to use pedagogy. And i tried different approaches. It's just not working. Each generation has it's own characteristics. Lately it seems to be "cheap" more than ever. Everything has to be easy, fast, cheap... My evaluation : dumb and pretentious. A kind of post TV generation. One would say teachers aren't helped.
Trying to give advice to commanders is the last thing that isn't bothering me. They ask and try. It's better than the usual 'binary people'. Unfortunately doing it "in-game" is rather difficult. Talking with/to the guy, rip off opponents and looking at the map at the same time is kind of difficult.
Ready rooms aren't the best place for teaching commanding. We even get vote kick for being there too long... We truly need an in-game space (ex : mod like pre-game) to be able to provide proper training. We don't have it and we burn tremendous amount of energy for poor results I'm afraid. Especially as people come and go.
People DO communicate and DO learn and DO participate if you actually give a damn and try.
When you say TEN times "don't go there", "Follow this WP", and they DO exactly TEN times the same thing, I'm afraid to say; they do not even listen. Can they understand ? I don't know.
No wonder people stack. I do as i already said. In fact everybody does. Everybody wants to win (generally speaking), isn't it. For me it's witnessing teamwork. When you call and people get there; it's nice. And i'm happy to answer calls from others as well. Teaching that should only take 2 or 3 sentences. I'm afraid to say it requires far more than that.
I communicate with my team, and I get a reception!
Play lottery. You probably are the luckiest man in the world.
Pub regs often blame stack and dont know hot to give advice or field comm.
It's rather easy to buy a microphone (cheap compared to a video card) and configure the 'use mic' key. or even use chat. I believe they do know but don't care. Cheap sticker on it.
Cannon_FodderAUSBrisbane, AUJoin Date: 2013-06-23Member: 185664Members, Squad Five Blue, Squad Five Silver, Reinforced - Shadow
@UncleCrunch agree with you 110%. If I am on a losing but working together team, with an encouraging comm the loss isn't so bad (even if the opposing team is obviously more superior in skill). It is the fact that you have team mates that work together that makes this game so much fun for me. The longish form of the game helps (I can't imagine something like CS:GO with tactics, as the rounds end so quick). To me that is fast food, whilst NS2 is the slow cooked, expertly prepared gourmet meal. So, in short, good comms that talk to the team is hard to come by, and one that gives you props trying to follow orders but fail is rarer still.
I haven't read the entire thread and nor do I plan to, but here's my two cents.
The skill system should not be based on a win/lose system. This type of system is only good for a game where one player can determine the outcome of them game - and this is not the case in NS.* NS is a team game, and requires the work of 2+ players to win. Whether that's the comm med spamming a marine or tag-teaming, it still requires a minimum of 2 players. As a result, a player's rating shouldn't be determined by the outcome of the game because a player with skill of 3000 paired with all rookies will never win a game against mediocre players - despite the average skill being equal. This does not mean the person with 3000 skill wasn't trying hard enough, or simply wasn't up-to-par with his or her skill level. This just means: a single player does not determine the outcome of the game.
I read earlier that it's based on wins/losses in order to encourage teamwork, but that's certainly not the common outcome. Instead, it encourages stacked games because if a good player joins the low-skill team and looses, he or she will always loose points regardless of how close the game was(n't) and how much he or she contributed. It is much more safe to continue stacking a team in order to spare any lost points.
Instead of the current system, I propose we increment a skill system relying on the points system within the game. Players that end up at the top of the scoreboard (aka, had the most game points) are typically the ones who had the largest impact in the game. This would spare the players that joined the "loosing" team, because they would no longer loose points due to the rest of his or hers team being incompetent.
Granted, if we adapt the game-point system it wouldn't be perfect; it would encourage players to build useless structures or go Rambo. This can be remedied by two ways:
1. Adjust the scoring system. It is pretty solid as-is, however; it needs to be more refined if it becomes more important than simply displaying who's at the top of the leader-board. Certain adjustments include, but aren't limited to:
- The gorge: important life-form, but not always the "game changer." Currently an average gorge is normally above all/most players because they get so many points for structure damage. I would adjust the points so an average gorge is mid-way through the leader-board.
- Recycled Structures: Currently, if a structure is recycled no points are given to the player. This means if it was literally one bite away, the player wasted all that time biting it - 30-40 seconds is a lot in this game! This probably isn't the most important, but is something I have noticed and been bothered by.
I'm sure there are other adjustments that would need to be observed and adjusted but my mind runs blank currently. If this system had intention of being utilized, then it wouldn't be too hard to spectate games and take note of which players are making the game-changing plays. After the game is over, if those players aren't at the top of the scoreboard then the score system needs more adjustment.
2. Weigh it with who won/lost. The winning team SHOULD have most of its players at the top of the scoreboard, so if a player that was on the loosing team is at the top of the scoreboard, they should earn a relatively high amount of points because the odds were against them. The leader-board should be cut in half, and adjust the player's skill according to being above or below the cut. * I recommend maybe a small negative weight on the winning team, because as earlier said, the winning team SHOULD be at the top of the board because normally by the end of the game it is easy to get a lot of kills. This would need a considerable amount of testing though and I'm not sure how reliable it is (as with most things in theory).
Lastly, with this alternate rating system the commander is left out. For this, I would judge the commander on win/loss ratio because a commander typically IS the difference between a winning and loosing game. The commander is the only role that SHOULD heavily determine if a team wins or losses due to the tech path, orders, response time, etc. Granted, if the team doesn't listen or don't match up with skill it's not the commander's fault, but in an average scenario the commander regularly determines the team's game-play.
Regardless if people think this is a good idea or not, the current skill system is NOT cutting it. I play with a lot of regular players and can normally tell when a game will be a slaughter; that being said, I am baffled at what the team-balance/total average skill thinks to be fair! Not only are slaughter-fests deemed "equal", these bullshit games destroy any hivescore every loosing player had by an amount of 30-60 points. Then, when winning fair games players only get 15-20 points. Hell, we've won against stacked teams and only received 3-4 points before. Man, that hurt. Maybe it would even out if alien/marine skill was separate, but that doesn't remedy what the win/loose system stipulates: If you won, then you were the cause. If you lost, well, clearly you weren't trying hard enough.
* Yes, I know there are some players that are basically god and CAN sway an entire game, but just remember those are the outliers.
You're wasting a lot of your time by entering a thread, deliberately not reading the earlier posts and then posting. Almost everything you said has been discussed at length; see the thread.
In short, you would be introducing more problems than you would solve (if you solve any), and you would deviate from what the skill system is trying to accomplish.
Comments
This is true. This would not be a problem if things like server sizes, faction selections etc. were randomised, since the skill rating would take it into account; one determinant of your probability of winning is the probability of you being assigned to the better team. Your skill level still converges towards the underlying value, as these variables are randomised.
However, it becomes a problem when the player themselves is given the choice. If they want to change their play, for example from a field marine to commanding or from mostly marines to mostly aliens, the skill rating suddenly becomes useless, as it has never had the opportunity to take these variables into account. Over time the player's skill, however, will start to converge towards the new value that takes his skill in this new playstyle account as well. This cannot really be helped, however, as you cannot force people to play every aspect of the game equally. But I think that separate skill ratings for marines and aliens would be so easy to implement that it certainly could not hurt.
I mean IMHO the current concept is sound (grade players then assign players to teams based on your grading). But it is those little oversights like this one, that makes lot of people hate it.
1) Are marines and aliens balanced, i.e. can you skew your skill by preferring one faction over the other?
2) Does the standard deviation of skill ratings affect the probability of an outcome controlling for the average skill of a team, i.e. does one pro and a bunch of rookies equal a team of mediocre players even if the average skill ratings of the team are the same?
3) How much does the player count affect gameplay, so if you switch from playing exclusively on 42-player servers to playing exclusively 20-player servers, is your initial skill rating at all comparable to people who have already played on these servers for a long period of time?
I have talked to moultano and he plans to do something for #1. He is a busy guy, with a job and a baby, so I do not know when he will get around to it. He has said in this thread and to me in other conversations that he plans on separating marine and alien hive skill. I don't think there is much to discuss on this one anymore.
#2 is a problem there seems to be some disagreement on how to handle.
Lets say your server is as follows. How would you distribute these players?
skill 3000 : 1 player
skill 1000: 8 players
skill 0 : 5 players
In my opinion rookies at best follow orders but struggle to lose most engagements, at worst a rookie can be so bad they might as well not even be on your team. So I would distribute as follows:
Team A: 3000, 1000, 1000, 1000, 0, 0, 0 = 6000 total team score
Team B: 1000, 1000, 1000, 1000,1000, 0, 0 = 5000 total team score
Essentially distributing the rookies as evenly as possible after the after every non rookie has been placed.
On this issue moultano says:
#3 is a difficult one. I think this is where much of disagreement over if hive works or not comes from. I play on a very small selection of servers, 1-3 servers, with usually the same group of people and little variation. Within those servers who have a bunch of regulars I see FET work extremely well producing very balanced teams. I also think that people who don't think FET works at all play on servers that have a lot of mixing of players from a wide range of servers. I have occasionally ventured out into other servers and I do see my hive skill does not seem to be comparable to other tight knit communities. This is in part because a person who plays on a 42 person server can get more skill points in a round than a person on a 16 person server.
I don't think having an alternative balance system on a lone server would work very well. The FET already works great on lone servers where most players are regulars.
0 for rookie players is mathematically speaking complete nonsense. That would mean they are infinitely worse than even 1 SP player. Something in 200-300 range in current SPs is I think adequate. The hive could perhaps observe new players and come up with some experimentally verified number to assign to new players.
What if I told you there were players with 50+ hours recorded in hive, that is actually time in a recorded game, and those players still had 0 points. They are so low skilled they can not win enough games to get above 0 skill points. There are many many players with tons of hours recorded in hive with skill points at or near 0.
So are you suggesting the minimum be higher than 0?
That's the saddest thing I've heard all day
The SPs actualy run into minus values. they are just rounded to 0 after. So it is not unthinkable... I think the guy you mention could still have reasonable K/D of 2/10 or something like that. Maybe he plays on touchpad, I dunno...
I do not know how they play so poorly, and I don't understand how the very top of the skill spectrum plays so well.
My point is there are people with significant amount of playtime who are just that bad. I think 0 is fine.
I think your complaining about how the skill points are not linear but I don't think that matters either.
All data should be from recently active players if that is possible, so we get a snap shot of the present player base:
So data should be from Active players in the last 30 to 60 days
Would like to know:
1. average skill (total skill / players) - what is considered to be the average skill rating atm
2. median skill (the person with equal players above and below them) - what is the distribution of skills compare to average, it should tell you if there are just more high skills averaging out the lower skill people playing on white listed servers.
3. distribution of the skills across the board - how many players are skill 1000-1100 etc... The slice of the band can vary, but will give you an idea of how skill is distributed atm (it is like a more detail look of median skill above.
I think knowing this may help explain why FET sometimes just doesn't balance teams the way it ought to.
Over to you arm chair statisticians.
http://forums.unknownworlds.com/discussion/136931/basic-statistics-of-the-hive-skill-system?new=1
In regards to my third suggestion quoted above, I think the problem is not rookies starting at such a low score but a convergence one. If you look at the skill distribution there is a large population gap between players under 200 skill and players at 1000 skill. Basically I think that hive scores do not grow quick enough.
I still think rookies, or people of extremely low skill should be sorted differently.
But yes the new players should be assigned more sane value, so they converge their Skill to their skill, more quickly. Any trick to use them differently while they coverge may help, but would be unnecessary I think, and perhaps dangerous (it changes how all players are rewarded with SPs). But I like your idea of putting equal number of rookie players(It should be weighted by number of games the system trained on them already too) to each team, so the risk, that they get scored unfairly, is equal to both teams. Of course the problem remains when FET is not used.
EDIT: I just thought of another unrelated minor problem. When people have shared Steam account(e.g. Gaming Cafés, families), their Skill is just changing as people do change on the PC. (Hmmm I wonder if cyberCafées buy this game...)
... I had a dummy account lying around. And I always assumed that rookies started off at 1000, this account started at 0. I think depending on when you key was activated there were different starting hive scores. Anyway. Took me 300 hours to climb to 1000. And that was one brutal climb... Fricking climbing in 30ish point increments.
I saw some players with like 10 hours on hive and 1000 points. I think my primary also started at 1000.
ShellMo (who was at a skill level of 300 ish) has now got her skill level up to 1022 super fast. How did she do it? She just started playing aliens more often.. She used to play roughly 2 rounds of marines for every 1 round she played aliens. Now she is doing it the other way around, and within the space of 30 games her skill level has more than tripled.
With this knowledge I decided to see how well I could do with gaming the system, and i've realised my skill level is far more dependant on picking the right team to join than it is on my own performance. Here is how you prop your skill up.
If ((MarineAvgSkill / AlienAvgSkill) > ~1.2) Pick Marines
Else pick Aliens.
Simply: If marines are stacked join marines, otherwise join aliens.
Using this my skill level has gone from 1600, to 2000ish.
This works because if the teams are even you have roughly a 54% of winning on aliens, and 46% of winning on marines. It also doesn't matter if the teams are stupidly stacked. You still get points from winning a stupidly stacked almost impossible game, and you still lose points from not being able to defeat a team nobody would expect you to beat.
The point that you can gain skill points from stacked games is evident when you look at the highest skill players. For example: http://hive.naturalselection2.com/profile/3551432
2.78 win/loss ratio. If you play even games one would expect a win/loss ratio of 1, a bit like flipping a coin. So that player has either won a ton of unfair games, or their skill is actually significantly higher and the skill system has failed to converge on the real high value after 382 hours of recorded gameplay.
Now you would assume the system would self correct, in that if you artificially boost your skill, you will be placed with less skilled players next time and find it progressively harder to win. However due to the way the team skill is an average i've not noticed any negative consequences at all. I probably wouldn't until my skill level became hugely higher than average.
It's true that if you play well (doing the right thing at the right moment, listen to commander, etc.) but ultimately loose; your score goes down. I always wonder why a system that tries to measure your personal level uses team results as a corner stone. There are many other ways.
Every system probably has an angle that encourages poor behavior though - i.e. stacking. But that doesn't mean that it isn't worth while to, or at least attempt to, encourage good teamwork.
You still can loose with perfect teamwork. All good actions are hidden behind a single 1 or 0 which defines the outcome of the game. Tada!
Biting RTs, Building, weld/healing are part of the job non the less. It requires organization and potentially more skill sometimes.
With a win-based skill system, like this, anything that results in a win is encouraged - whether that constitutes team work or somebody carrying the whole team. Either way, that is what you want - in contrast to people running around welding irrelevant things or frag hunting, instead of managing the resources, for points.
How would you quantify field calling or overall communication skills?
The skill system does not predict your W/L metric to converge towards 1. The skill system does not force you to play even teams, even if FET was used, because you can always switch teams and stack after the vote has passed. However, it doesn't prove that the skill system is flawed if someone is consistently winning more games than they are losing; those players will lose much more points from every defeat, even if they are rare. You can have two players with exactly the same skill rating, and correctly estimated to boot, who have wildly different win-lose ratios, and there's nothing fishy about that. If you like stacking, you win more games, but you are rewarded less for every victory. If you like anti-stacking, you win less often, but you are punished less for every defeat. It might seem like someone can just farm small increments of points from stacking in a way that guarantees their victory, inching up their score with minimal effort, but there's always that small chance of a defeat lurking there, and when it happens, their score will come crashing down.
Aliens and marines should have separate skill measurements, that is true.
@UncleCrunch
You have to learn what the term "on average" means.
Fast list :
-Can be compared to a blender that crush rookies.
-FET that just can't be true as it gives stacked games to a level unknown to me so far. I was right to avoid it.
-Any OK player getting on the rookie side, faces humiliation like if somebody was throwing eggs at him. Not to mention racism/nationalism, i was surprised to hear what i heard.
-The slot number acts like a magnet for these new players.
-Their throats sound like a sad flute as the blade slides.
-"Good bye fresh meat".
Right now: Useless, putting aside the language "obstacle". These new players are just canon fodder. They probably think the maps are designed like a tube and they always push in one direction, endlessly. They don't think the opponent can be smart and set up traps. Even after 10 tries... Giving advices and directions to them is useless. There is a dark energy that prevents any good tip to get stored in their heads.
And when you suggest (ready room) to train/teach them (and take the time to do it) : Nobody answer on the rookie side. But the best is that no Stacker do try to do the same as they want their fresh meat of the day. "It's NS2" they say. Childish to say the least. While some of them are supposed to be way beyond that.
The rookies should play Combat instead. It would at least look like what they know, then, maybe they can try NS2. It's truly leveling from the bottom.
But i see where you're going. It is still measurable. Time passed using the microphone is something that can be sampled. Can be faked but it would be a little ridiculous. It would look like a tea time full of people who make small talks... Who said "TS" ?... Girls... who knows...
You should really take the time to look at the 'on average face' on the servers (concerning FET and other things). Right now, it's like a Picasso. And i like Art.
Did I stumble into an improvisation theater rehearsal?
Why do I always feel like this is utter nonsense being issued by all the "stack" complainers. Quite literally NO ONE truly communicates beyond "2 marines logistics" and the like. No one gives tips, no one explains the reasons behind decisions. Obviously my argument here is logically flawed because I can't be on every server all the time to witness the lack of communication, but it happens so infrequently at least on EU servers, that it really bothers me.
Comp players don't often give advice. I do, ALL the time, I frequently point out shit plays, and frequently applaud good plays. I give explanations, I give field comm advice, I communicate with my team, and I get a reception! People DO communicate and DO learn and DO participate if you actually give a damn and try.
Pub stars often blame stack and better players but still dont give advice or field comm.
Pub regs often blame stack and dont know hot to give advice or field comm.
Rookies just trial and error to the best of their own ability to improve.
The issue most pub games face is a complete and utter inability to teach and learn and instead to just blame. I'm so tired of people complaining of stacks and unfairness because it's the complaining that ruins the games. The psychological defeat before the game starts puts teams at such a disadvantage. They give up and see themselves trapped and often suicide. Suicide is actually an adept analogy to draw parallels too, where people often feel so trapped and as if there is no way out that they give up. People who see this attitude need to man the fuck up and help them by communicating and problem solving through voice comms.
The games I enjoy most now are those when I lose because I'm being shut out the game by awesome ambushes, traps, split pressure, things where the opposite team are clearly engaged in the game instead of mind numbingly W + lmb, die, and complain.
Everyone has a duty when they join a team on a server to commit their efforts to winning, not complaining. If it's too much effort for you to communicate / trap / play to the best of your ability and the opponents don't find it too much effort, you're going to lose. Deal with it. If you don't try, you can't expect the other team to not try as much as you just to give you a good game.
1 match = 1 marine round + 1 alien around
and require that. Since there's not enough players to make that a requirement with punishment for game-leaves, how can we try to see it that way without forcing 2 round matches for skill-ranked games? Place a metric on rounds played per side affecting your point gains? I don't have any great ideas for it.
I just grown 'fed up' to try to use pedagogy. And i tried different approaches. It's just not working. Each generation has it's own characteristics. Lately it seems to be "cheap" more than ever. Everything has to be easy, fast, cheap... My evaluation : dumb and pretentious. A kind of post TV generation. One would say teachers aren't helped.
Trying to give advice to commanders is the last thing that isn't bothering me. They ask and try. It's better than the usual 'binary people'. Unfortunately doing it "in-game" is rather difficult. Talking with/to the guy, rip off opponents and looking at the map at the same time is kind of difficult.
Ready rooms aren't the best place for teaching commanding. We even get vote kick for being there too long... We truly need an in-game space (ex : mod like pre-game) to be able to provide proper training. We don't have it and we burn tremendous amount of energy for poor results I'm afraid. Especially as people come and go.
When you say TEN times "don't go there", "Follow this WP", and they DO exactly TEN times the same thing, I'm afraid to say; they do not even listen. Can they understand ? I don't know.
No wonder people stack. I do as i already said. In fact everybody does. Everybody wants to win (generally speaking), isn't it. For me it's witnessing teamwork. When you call and people get there; it's nice. And i'm happy to answer calls from others as well. Teaching that should only take 2 or 3 sentences. I'm afraid to say it requires far more than that.
Play lottery. You probably are the luckiest man in the world.
It's rather easy to buy a microphone (cheap compared to a video card) and configure the 'use mic' key. or even use chat. I believe they do know but don't care. Cheap sticker on it.
The skill system should not be based on a win/lose system. This type of system is only good for a game where one player can determine the outcome of them game - and this is not the case in NS.* NS is a team game, and requires the work of 2+ players to win. Whether that's the comm med spamming a marine or tag-teaming, it still requires a minimum of 2 players. As a result, a player's rating shouldn't be determined by the outcome of the game because a player with skill of 3000 paired with all rookies will never win a game against mediocre players - despite the average skill being equal. This does not mean the person with 3000 skill wasn't trying hard enough, or simply wasn't up-to-par with his or her skill level. This just means: a single player does not determine the outcome of the game.
I read earlier that it's based on wins/losses in order to encourage teamwork, but that's certainly not the common outcome. Instead, it encourages stacked games because if a good player joins the low-skill team and looses, he or she will always loose points regardless of how close the game was(n't) and how much he or she contributed. It is much more safe to continue stacking a team in order to spare any lost points.
Instead of the current system, I propose we increment a skill system relying on the points system within the game. Players that end up at the top of the scoreboard (aka, had the most game points) are typically the ones who had the largest impact in the game. This would spare the players that joined the "loosing" team, because they would no longer loose points due to the rest of his or hers team being incompetent.
Granted, if we adapt the game-point system it wouldn't be perfect; it would encourage players to build useless structures or go Rambo. This can be remedied by two ways:
1. Adjust the scoring system. It is pretty solid as-is, however; it needs to be more refined if it becomes more important than simply displaying who's at the top of the leader-board. Certain adjustments include, but aren't limited to:
- The gorge: important life-form, but not always the "game changer." Currently an average gorge is normally above all/most players because they get so many points for structure damage. I would adjust the points so an average gorge is mid-way through the leader-board.
- Recycled Structures: Currently, if a structure is recycled no points are given to the player. This means if it was literally one bite away, the player wasted all that time biting it - 30-40 seconds is a lot in this game! This probably isn't the most important, but is something I have noticed and been bothered by.
I'm sure there are other adjustments that would need to be observed and adjusted but my mind runs blank currently. If this system had intention of being utilized, then it wouldn't be too hard to spectate games and take note of which players are making the game-changing plays. After the game is over, if those players aren't at the top of the scoreboard then the score system needs more adjustment.
2. Weigh it with who won/lost. The winning team SHOULD have most of its players at the top of the scoreboard, so if a player that was on the loosing team is at the top of the scoreboard, they should earn a relatively high amount of points because the odds were against them. The leader-board should be cut in half, and adjust the player's skill according to being above or below the cut. * I recommend maybe a small negative weight on the winning team, because as earlier said, the winning team SHOULD be at the top of the board because normally by the end of the game it is easy to get a lot of kills. This would need a considerable amount of testing though and I'm not sure how reliable it is (as with most things in theory).
Lastly, with this alternate rating system the commander is left out. For this, I would judge the commander on win/loss ratio because a commander typically IS the difference between a winning and loosing game. The commander is the only role that SHOULD heavily determine if a team wins or losses due to the tech path, orders, response time, etc. Granted, if the team doesn't listen or don't match up with skill it's not the commander's fault, but in an average scenario the commander regularly determines the team's game-play.
Regardless if people think this is a good idea or not, the current skill system is NOT cutting it. I play with a lot of regular players and can normally tell when a game will be a slaughter; that being said, I am baffled at what the team-balance/total average skill thinks to be fair! Not only are slaughter-fests deemed "equal", these bullshit games destroy any hivescore every loosing player had by an amount of 30-60 points. Then, when winning fair games players only get 15-20 points. Hell, we've won against stacked teams and only received 3-4 points before. Man, that hurt. Maybe it would even out if alien/marine skill was separate, but that doesn't remedy what the win/loose system stipulates: If you won, then you were the cause. If you lost, well, clearly you weren't trying hard enough.
* Yes, I know there are some players that are basically god and CAN sway an entire game, but just remember those are the outliers.
#EndLifeAsAnAnonymousViewer
You're wasting a lot of your time by entering a thread, deliberately not reading the earlier posts and then posting. Almost everything you said has been discussed at length; see the thread.
In short, you would be introducing more problems than you would solve (if you solve any), and you would deviate from what the skill system is trying to accomplish.