If the skill system is accurate then it should allow the creation of balanced teams.
If the skill system does not help create balanced teams, the system is flawed. Some of the zealots here seem to be re-writing history and giving the impression that the skill calculation system was created just for the sake of it. The skill system was not created for people to flaunt their giant scores.
Even if there is a 50-50 equal split in opinion, that is enough to show that not enough of the community (at least the subset represented by forum go-ers) has confidence in the system to justify its continued usage. If you go out into the wild, you will find that in the general population there's even less faith in the scoring system as is.
There's no contradiction here. The point of the skill system was to enable better team creation. If you think the current system does not help create balanced teams, then the system is flawed and needs to be corrected or scrapped.
Why should we care if people have no confidence in the system, as long as it works? I know a few people who have no confidence in evolution. I don't take them seriously.
Some of the zealots here seem to be re-writing history and giving the impression that the skill calculation system was created just for the sake of it. The skill system was not created for people to flaunt their giant scores.
I don't know who these supposed zealots are. I may be one of them because I have been very active on discussions involving the skill system. Whether or not I am one, I do not know what you are talking about with these statements. I have been very active in skill system discussions and this does not match what I remember reading in the recent past. I have not had the impression anyone is "re-writing history" at all.
I also find it contradictory that you say people are rewriting history, yet you tell us all the ways we can vote in your poll even if we can't agree with it entirely. For example, I find it very useful in balancing purposes. I in no way think it gives a perfect representation of skill. The skill system does work, but it also could be better.
Even if there is a 50-50 equal split in opinion, that is enough to show that not enough of the community (at least the subset represented by forum go-ers) has confidence in the system to justify its continued usage. If you go out into the wild, you will find that in the general population there's even less faith in the scoring system as is.
The community does have a negative view of it for a variety of reasons in my opinions, not all of which have to do with it being balanced. FET votes are hard to get passed
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
I'm curious how the people who dislike it are evaluating it. Do you see people whose score seems wrong? Does your own score seem wrong? (Please link so I can check it out) Do you vote for force even teams? When you do, do you have a bad game? Can you describe what the game was like?
it does work but only somtimes HOWEVER it is better than letting rookies and/or vets stack teams. but chance of balanced game using this system is around 20% and i would gues it might even be lower
it does work but only somtimes HOWEVER it is better than letting rookies and/or vets stack teams. but chance of balanced game using this system is around 20% and i would gues it might even be lower
I'm not sure how useful these anecdotes are, not to mention these obviously fabricated statistics.
One thing you should ask your self; How do you define a balanced game? Does it have to be 50/50?
People have biases. One could easily be inclined to consider a game where they themselves were dominating; "fair", while a game where they were only slightly disfavored; "unfair", just because of their own bias.
People also have confirmation biases, they might forget all the games that purport to their agenda and remember the ones that do.
In addition. NS2 is a game that unfortunately is designed in a way that it heavily snowballs. Snowballing games leave an illusion that you are getting stomped, because a small favor grows bigger and bigger. So a team that gains an early lead by fair means, can dominate the game and make it look really one-sided.
So anecdotes aren't very useful if you don't have some real data to back it up and some non-ambiguous definitions of what you're talking about.
This is why people like moultano repeatedly ask you to link to examples where you think the system is inaccurate.
I don't mean to single you out Das19, I'm writing this because others have raised similar anecdotes, you're simply the most recent example.
I'm curious how the people who dislike it are evaluating it. Do you see people whose score seems wrong? Does your own score seem wrong? (Please link so I can check it out) Do you vote for force even teams? When you do, do you have a bad game? Can you describe what the game was like?
By how many games it actually balances, which for me is less than 1 in 4. It doesn't do the job it is there for.
I'm curious how the people who dislike it are evaluating it. Do you see people whose score seems wrong? Does your own score seem wrong? (Please link so I can check it out) Do you vote for force even teams? When you do, do you have a bad game? Can you describe what the game was like?
By how many games it actually balances, which for me is less than 1 in 4. It doesn't do the job it is there for.
Can you describe more about what those games are like? Were there obviously excellent players that it didn't split up or was it something more subtle? Was it one race that was winning in all of the bad games or even split?
Also how often do you find games that don't use the skill system to be balanced? Is it more or less than 1 in 4?
games that don't use FET are even rarer, treated like its gospel and must be used at all times.
As for the unbalanced games, it's a mix. can be just a curb stomp on both teams from the get go with no apparent reason why(i.e. no particular special player). had games where we've had one exceptional player but he will just run off and Rambo while the rest of the team gets slaughtered and we will lose all territory. new one recently is where a rookies gets balanced to one team and I find that team will be far far stronger even with the rookie. Other games where we get an exceptional player (hive wise) in the comm chair and they tend to end up slaughter fests. The biggest one i find though is when you see comp players being put on the same side, thankfully they don't last long as playing as either side is just dull.
Personally I just think it's not a problem that can be fixed with such a small playerbase.
In addition. NS2 is a game that unfortunately is designed in a way that it heavily snowballs. Snowballing games leave an illusion that you are getting stomped, because a small favor grows bigger and bigger. So a team that gains an early lead by fair means, can dominate the game and make it look really one-sided..
DING DING DING WE HAVE A WINNA!
It's why I've been focused lately on trying to find ways to lessen that snowball
In addition. NS2 is a game that unfortunately is designed in a way that it heavily snowballs. Snowballing games leave an illusion that you are getting stomped, because a small favor grows bigger and bigger. So a team that gains an early lead by fair means, can dominate the game and make it look really one-sided..
DING DING DING WE HAVE A WINNA!
It's why I've been focused lately on trying to find ways to lessen that snowball
Two things you can do:
1. Allow simple majority surrender at any time of game
2. Untie cysting and power nodes from building requirements
A hypothetical question to you: if someone were to show you a comprehensive dataset of every single NS2 game played since the implementation of the skill system, and it were to show that games created with FET do have balanced outcomes, would you still deny it and continue to protest against the system? Would you still think that your personal gut feeling invalidates objective data?
I'm just trying to get inside the mindset of people like you.
IronHorseDeveloper, QA Manager, Technical Support & contributorJoin Date: 2010-05-08Member: 71669Members, Super Administrators, Forum Admins, Forum Moderators, NS2 Developer, NS2 Playtester, Squad Five Blue, Subnautica Playtester, Subnautica PT Lead, Pistachionauts
@mattji104
Just those two options, eh?
Besides you never elaborating on your proposal regarding powernodes, despite Nordic pointing that out, I wholeheartedly disagree - there are many more methods.
Just give it some thought, it doesn't need to be as black and white as this poll.
games that don't use FET are even rarer, treated like its gospel and must be used at all times.
As for the unbalanced games, it's a mix. can be just a curb stomp on both teams from the get go with no apparent reason why(i.e. no particular special player). had games where we've had one exceptional player but he will just run off and Rambo while the rest of the team gets slaughtered and we will lose all territory. new one recently is where a rookies gets balanced to one team and I find that team will be far far stronger even with the rookie. Other games where we get an exceptional player (hive wise) in the comm chair and they tend to end up slaughter fests. The biggest one i find though is when you see comp players being put on the same side, thankfully they don't last long as playing as either side is just dull.
Personally I just think it's not a problem that can be fixed with such a small playerbase.
There's an interesting distinction there that I think is worth making. In many of the cases you describe, the games were not balanced because important players weren't doing what was most effective to win. This doesn't necessarily mean that the teams were unbalanced. If your best player decides to fool around, you will probably get stomped even if the teams are perfectly composed. There's nothing that even a perfect system could do about this.
As much i dont want to derail this thread into a reduce-the-snowball one, I still think the best idea is dimishing returns on RT resource generation.
This way the team with more RTs has an advantage, but not a huge one.
The numbers would have to be balance tested, but:
Something like each additional rt after 1 is worth 10%-20% less resourse/sec. So if one team controls the whole map but the other can hold their naturals, they would still be disadvantaged, but not as much as now.
Half resource points are already used if a team has 0 RTs, so i know fractions of a resource point are possible.
i wouldnt mind testing out a map with 2x the number of resource points give 50% of the current res per turn and costing half the current res to build..
A hypothetical question to you: if someone were to show you a comprehensive dataset of every single NS2 game played since the implementation of the skill system, and it were to show that games created with FET do have balanced outcomes, would you still deny it and continue to protest against the system? Would you still think that your personal gut feeling invalidates objective data?
I'm just trying to get inside the mindset of people like you.
I'd be fine with it, still doesn't change the issue of games being curb stomps. And this isn't gut feeling, it's observation of a fet happening and then a rookie throwing the balance out because he's a 0 value. prem div players with lower hive scores than me being assigned to one team.
A hypothetical question to you: if someone were to show you a comprehensive dataset of every single NS2 game played since the implementation of the skill system, and it were to show that games created with FET do have balanced outcomes, would you still deny it and continue to protest against the system? Would you still think that your personal gut feeling invalidates objective data?
Yes! Fight da system, maan! :P
ok, DO IT, I want those data now! :P Seriously, it would be nice to have it. How would you determine the outcome is balanced? Round duration?
Repeating my earlier post about people having too high expectations of any balancing system. "Balanced teams" does not equal even games.
even with a magic skill system that evaluates skill accurately, the player that goes commander after teams are assigned has a large impact on the team. A good/bad commander will affect the team accordingly, especially on marines. Also, unless the skill level on the alien team is even, the choice of lifeform of individual players will be significant. If the one exceptional player goes gorge/commander, the team is probably done for on "balanced teams".
@moultano
I think FET is pretty much hit and miss. There have been stomps and even games, but I don't know if it is better than if we had forced random. As I mentioned before, I think the 1300-2000 range is pretty murky. It is just my experience though.
But It is definitely better than stacking that has a 90% chance of other team struggling to hold naturals.
A hypothetical question to you: if someone were to show you a comprehensive dataset of every single NS2 game played since the implementation of the skill system, and it were to show that games created with FET do have balanced outcomes, would you still deny it and continue to protest against the system? Would you still think that your personal gut feeling invalidates objective data?
I'm just trying to get inside the mindset of people like you.
I'd be fine with it, still doesn't change the issue of games being curb stomps. And this isn't gut feeling, it's observation of a fet happening and then a rookie throwing the balance out because he's a 0 value. prem div players with lower hive scores than me being assigned to one team.
An observation? Show me the data. Otherwise It's a gut feeling. You'd be surprised how many things you are "sure" about without writing them down are actually very much different from what you thought. How often did you sneeze today? Sneeze once in a crowded room and it feels like you've been sneezing a lot more than on other days, because you don't really pay attention to the times you sneeze alone at home, and you forget them. You could swear that you're coming up with a cold. There is so much room for psychological bias if you just try to remember how many times something has happened.
I'm not saying the system is perfect, because I don't have the data, and I don't know. This was a hypothetical question. And your answer very much showed me that you value your personal subsample more than comprehensive statistics.
A hypothetical question to you: if someone were to show you a comprehensive dataset of every single NS2 game played since the implementation of the skill system, and it were to show that games created with FET do have balanced outcomes, would you still deny it and continue to protest against the system? Would you still think that your personal gut feeling invalidates objective data?
Yes! Fight da system, maan! :P
ok, DO IT, I want those data now! :P Seriously, it would be nice to have it. How would you determine the outcome is balanced? Round duration?
It's difficult. I don't know what is available, but I think the system desperately needs data to analyse. And I'm not entirely sure how to determine if the outcome is balanced, either. One way to test the system would be to take a sample of every game that is NOT balanced according to the system, and see if the distribution of wins is statistically different from that predicted by the system. Evaluating balanced games would be more difficult, since there would be no separating characteristic of either team, and by default, it would be a 50-50 distribution.
A hypothetical question to you: if someone were to show you a comprehensive dataset of every single NS2 game played since the implementation of the skill system, and it were to show that games created with FET do have balanced outcomes, would you still deny it and continue to protest against the system? Would you still think that your personal gut feeling invalidates objective data?
I'm just trying to get inside the mindset of people like you.
I'd be fine with it, still doesn't change the issue of games being curb stomps. And this isn't gut feeling, it's observation of a fet happening and then a rookie throwing the balance out because he's a 0 value. prem div players with lower hive scores than me being assigned to one team.
Can you link to one of these premium division players?
50-50 between team1 and team2 will simply mean the both teams have same chance of being stacked. You would be showing the system predicts correctly only on average(same as random prediction).
50-50 between team1 and team2 will simply mean the both teams have same chance of being stacked. You would be showing the system predicts correctly only on average(same as random prediction).
That's what I said: we couldn't draw much inference from analysing matches with balanced predictions. Way to cut out the beginning of the quote.
If we concentrate on, say, a sample of matches that the system predicts have a 90-10 probability of outcome, we can analyse whether the team that has a 90% chance of winning actually wins 90% of the time or not. If yes, then we have evidence (though not proof) that the system works.
That's what I said: we couldn't draw much inference from analysing matches with balanced predictions. Way to cut out the beginning of the quote.
If we concentrate on, say, a sample of matches that the system predicts have a 90-10 probability of outcome, we can analyse whether the team that has a 90% chance of winning actually wins 90% of the time or not. If yes, then we have evidence (though not proof) that the system works.
Sorry bout that, just expanding what you said... But the same apply to 90-10 too(only proving the result is correct on average). Only way would be to force the same teams to play like 100 games together and then judge from that(Well looking at the hours some players have, this could be doable :P).
Well, if the team that the system predicts having a 90% chance of winning a match wins 90% of the time, I think everything is perfect. It means that any given match with that difference between the skill ratings is going to end up in the stronger team winning with a 90% certainty, and before the game starts, people in the teams can say "okay, I've got a 90%/10% change of winning this". I don't really understand what else there is to consider after that.
You don't need to have the exact same teams to get a sample, you can use every observation where the system predicts such probability of outcome. Many statistical and scientific studies are made with this principle. For example, in finance, people study how stocks that have performed well in the previous month (high increase in prices) behave during the month after. They study this effect by ranking stocks based on their past-month performance EACH MONTH, and then getting a sample of, for example, 120 portfolios of stocks that had the best performance for that given month, one for each during a 10-year window, and they use the data on those 120 portfolios to do research. These portfolios contain different stocks every month, but each portfolio is considered an observation of the same entity, namely the "winner portfolio". They do not require each portfolio to always contain the same stocks.
The point of this thread was to canvass opinion about the skill system, not to degenerate into a train wreck regarding evolution or financial markets (both of which, by the way, have more data than a piddling 200 or so regular players can provide for this game), and a rehashing of what has been said numerous times for 15 or more pages on another thread.
UWE neglected this aspect of the game, and no doubt this helped drive a lot of players away from the game. The optimization issues were the cause for the rest to go.
It is amusing that while this appears to be the only game for which a fair, consistent, team-creation or skill measurement system cannot be developed, there are so many snake oil salesmen on these forums proferring their various types of pseudo-intellectual babbling waffle. I think Hugh was correct when, a few years ago, he described them as "professional commentators" with no skills outside of a keyboard. No amount of echo-chamber auto-justification of a flawed system can bring this game back to life though.
It is even disturbing how seriously people are talking about a skill system for a video game which, in the grand scheme of things, no one plays. I and many others think the system is no good. That is my opinion. About an aspect of a game. In response we get all kinds of essays, abuse and passive aggressive attacks demanding "explanations" about "mindsets." This is a game. There is no need for nuance. None of the theses I've seen on this and the other thread actually matter. Because this game doesn't matter.
In the end, circular arguments by a few verbose people cannot solve the problem which is that the skill system in use now is unsatisfactory, and in fact is a failure, with the other main underlying problem being the dearth of players.
Comments
It is possible to hold the position that the skill system is accurate, but doesn't assign fair teams.
It is possible to hold the position that the skill system is not accurate, but does assign fair teams.
What are you supposed to vote for if either of those is your position? It's just a terrible poll man.
No we couldn't have. The first option =/= "the system is good".
If the skill system does not help create balanced teams, the system is flawed. Some of the zealots here seem to be re-writing history and giving the impression that the skill calculation system was created just for the sake of it. The skill system was not created for people to flaunt their giant scores.
Even if there is a 50-50 equal split in opinion, that is enough to show that not enough of the community (at least the subset represented by forum go-ers) has confidence in the system to justify its continued usage. If you go out into the wild, you will find that in the general population there's even less faith in the scoring system as is.
There's no contradiction here. The point of the skill system was to enable better team creation. If you think the current system does not help create balanced teams, then the system is flawed and needs to be corrected or scrapped.
Very easy to vote as required.
It is possible for the skill system to produce inaccurate results and it have NOTHING to do with whether it is "good" or "broken" or not.
No, no it is not. If you don't understand why that is by now, after it being spelled out to you in incredible detail, then you never will.
I also find it contradictory that you say people are rewriting history, yet you tell us all the ways we can vote in your poll even if we can't agree with it entirely. For example, I find it very useful in balancing purposes. I in no way think it gives a perfect representation of skill. The skill system does work, but it also could be better.
The community does have a negative view of it for a variety of reasons in my opinions, not all of which have to do with it being balanced. FET votes are hard to get passed
One thing you should ask your self; How do you define a balanced game? Does it have to be 50/50?
People have biases. One could easily be inclined to consider a game where they themselves were dominating; "fair", while a game where they were only slightly disfavored; "unfair", just because of their own bias.
People also have confirmation biases, they might forget all the games that purport to their agenda and remember the ones that do.
In addition. NS2 is a game that unfortunately is designed in a way that it heavily snowballs. Snowballing games leave an illusion that you are getting stomped, because a small favor grows bigger and bigger. So a team that gains an early lead by fair means, can dominate the game and make it look really one-sided.
So anecdotes aren't very useful if you don't have some real data to back it up and some non-ambiguous definitions of what you're talking about.
This is why people like moultano repeatedly ask you to link to examples where you think the system is inaccurate.
I don't mean to single you out Das19, I'm writing this because others have raised similar anecdotes, you're simply the most recent example.
By how many games it actually balances, which for me is less than 1 in 4. It doesn't do the job it is there for.
Can you describe more about what those games are like? Were there obviously excellent players that it didn't split up or was it something more subtle? Was it one race that was winning in all of the bad games or even split?
Also how often do you find games that don't use the skill system to be balanced? Is it more or less than 1 in 4?
As for the unbalanced games, it's a mix. can be just a curb stomp on both teams from the get go with no apparent reason why(i.e. no particular special player). had games where we've had one exceptional player but he will just run off and Rambo while the rest of the team gets slaughtered and we will lose all territory. new one recently is where a rookies gets balanced to one team and I find that team will be far far stronger even with the rookie. Other games where we get an exceptional player (hive wise) in the comm chair and they tend to end up slaughter fests. The biggest one i find though is when you see comp players being put on the same side, thankfully they don't last long as playing as either side is just dull.
Personally I just think it's not a problem that can be fixed with such a small playerbase.
It's why I've been focused lately on trying to find ways to lessen that snowball
Two things you can do:
1. Allow simple majority surrender at any time of game
2. Untie cysting and power nodes from building requirements
A hypothetical question to you: if someone were to show you a comprehensive dataset of every single NS2 game played since the implementation of the skill system, and it were to show that games created with FET do have balanced outcomes, would you still deny it and continue to protest against the system? Would you still think that your personal gut feeling invalidates objective data?
I'm just trying to get inside the mindset of people like you.
Just those two options, eh?
Besides you never elaborating on your proposal regarding powernodes, despite Nordic pointing that out, I wholeheartedly disagree - there are many more methods.
Just give it some thought, it doesn't need to be as black and white as this poll.
There's an interesting distinction there that I think is worth making. In many of the cases you describe, the games were not balanced because important players weren't doing what was most effective to win. This doesn't necessarily mean that the teams were unbalanced. If your best player decides to fool around, you will probably get stomped even if the teams are perfectly composed. There's nothing that even a perfect system could do about this.
As much i dont want to derail this thread into a reduce-the-snowball one, I still think the best idea is dimishing returns on RT resource generation.
This way the team with more RTs has an advantage, but not a huge one.
The numbers would have to be balance tested, but:
Something like each additional rt after 1 is worth 10%-20% less resourse/sec. So if one team controls the whole map but the other can hold their naturals, they would still be disadvantaged, but not as much as now.
Half resource points are already used if a team has 0 RTs, so i know fractions of a resource point are possible.
anyway, thats all for my derailing!
every room on a map would have a res node
someone make a mod
I'd be fine with it, still doesn't change the issue of games being curb stomps. And this isn't gut feeling, it's observation of a fet happening and then a rookie throwing the balance out because he's a 0 value. prem div players with lower hive scores than me being assigned to one team.
ok, DO IT, I want those data now! :P Seriously, it would be nice to have it. How would you determine the outcome is balanced? Round duration?
even with a magic skill system that evaluates skill accurately, the player that goes commander after teams are assigned has a large impact on the team. A good/bad commander will affect the team accordingly, especially on marines. Also, unless the skill level on the alien team is even, the choice of lifeform of individual players will be significant. If the one exceptional player goes gorge/commander, the team is probably done for on "balanced teams".
@moultano
I think FET is pretty much hit and miss. There have been stomps and even games, but I don't know if it is better than if we had forced random. As I mentioned before, I think the 1300-2000 range is pretty murky. It is just my experience though.
But It is definitely better than stacking that has a 90% chance of other team struggling to hold naturals.
An observation? Show me the data. Otherwise It's a gut feeling. You'd be surprised how many things you are "sure" about without writing them down are actually very much different from what you thought. How often did you sneeze today? Sneeze once in a crowded room and it feels like you've been sneezing a lot more than on other days, because you don't really pay attention to the times you sneeze alone at home, and you forget them. You could swear that you're coming up with a cold. There is so much room for psychological bias if you just try to remember how many times something has happened.
I'm not saying the system is perfect, because I don't have the data, and I don't know. This was a hypothetical question. And your answer very much showed me that you value your personal subsample more than comprehensive statistics.
It's difficult. I don't know what is available, but I think the system desperately needs data to analyse. And I'm not entirely sure how to determine if the outcome is balanced, either. One way to test the system would be to take a sample of every game that is NOT balanced according to the system, and see if the distribution of wins is statistically different from that predicted by the system. Evaluating balanced games would be more difficult, since there would be no separating characteristic of either team, and by default, it would be a 50-50 distribution.
That's what I said: we couldn't draw much inference from analysing matches with balanced predictions. Way to cut out the beginning of the quote.
If we concentrate on, say, a sample of matches that the system predicts have a 90-10 probability of outcome, we can analyse whether the team that has a 90% chance of winning actually wins 90% of the time or not. If yes, then we have evidence (though not proof) that the system works.
You don't need to have the exact same teams to get a sample, you can use every observation where the system predicts such probability of outcome. Many statistical and scientific studies are made with this principle. For example, in finance, people study how stocks that have performed well in the previous month (high increase in prices) behave during the month after. They study this effect by ranking stocks based on their past-month performance EACH MONTH, and then getting a sample of, for example, 120 portfolios of stocks that had the best performance for that given month, one for each during a 10-year window, and they use the data on those 120 portfolios to do research. These portfolios contain different stocks every month, but each portfolio is considered an observation of the same entity, namely the "winner portfolio". They do not require each portfolio to always contain the same stocks.
UWE neglected this aspect of the game, and no doubt this helped drive a lot of players away from the game. The optimization issues were the cause for the rest to go.
It is amusing that while this appears to be the only game for which a fair, consistent, team-creation or skill measurement system cannot be developed, there are so many snake oil salesmen on these forums proferring their various types of pseudo-intellectual babbling waffle. I think Hugh was correct when, a few years ago, he described them as "professional commentators" with no skills outside of a keyboard. No amount of echo-chamber auto-justification of a flawed system can bring this game back to life though.
It is even disturbing how seriously people are talking about a skill system for a video game which, in the grand scheme of things, no one plays. I and many others think the system is no good. That is my opinion. About an aspect of a game. In response we get all kinds of essays, abuse and passive aggressive attacks demanding "explanations" about "mindsets." This is a game. There is no need for nuance. None of the theses I've seen on this and the other thread actually matter. Because this game doesn't matter.
In the end, circular arguments by a few verbose people cannot solve the problem which is that the skill system in use now is unsatisfactory, and in fact is a failure, with the other main underlying problem being the dearth of players.
Lighten up people!
I'm out.