0wl Posted February 4, 2018 Report Share Posted February 4, 2018 14 hours ago, Alex said: how did ub not win best clanwars clan lol 69-19 btw probably because the majority of wins came from an lpc scene where the biggest competition was bv and xl, until eop came in and ub went 0-6 vs eop, leading to the closure of ub 0ldschooler 1 Link to comment Share on other sites More sharing options...
Pikey Posted February 4, 2018 Report Share Posted February 4, 2018 13 hours ago, Mustafa said: Cant say anyone has a better core, well deserved award for supremacy clan members Link to comment Share on other sites More sharing options...
puppyslush Posted February 4, 2018 Report Share Posted February 4, 2018 4 hours ago, K===Big Bombz said: As a statistics undergrad obtaining my masters in data analytics, your logic behind nulling Eop/Aao votes is flawed. Utilising 95% confidence intervals to obtain outliers only applies to numerical/quantitative data sets, not qualitative ones. For example, if there was a poll asking who the best soccer player in the world was, and everyone was told to choose their top two, most people would pick Messi and Ronaldo. Just because most people would vote those two players does not quantify as an outlier when the MAJORITY of the people in the poll chose that option. @slushpuppy I disagree. We are only taking a sample size of the entire population of the pure community and for most community-esque awards, they tend to be skewed in favor of certain clans if we do a straight majority polling. Your football analogy is faulty comparison as a result; population measurement vis a vis sample measurement. The problems Sharkbrew faces is akin to doing your sample polling in a Brazilian village, there, they would say Naymar is the best football player. That is why all community sites implement a sort of anti biasing measure and for this award, using standard deviation seems like the painless way rather than trying to manually verify as a result, however I'd welcome any improvement on normalization techniques as I agree with you on the aspect that totally nulling isnt the way to go. Link to comment Share on other sites More sharing options...
Blue Posted February 5, 2018 Report Share Posted February 5, 2018 gz all!!! Link to comment Share on other sites More sharing options...
auwi Posted February 5, 2018 Report Share Posted February 5, 2018 1 hour ago, slushpuppeh said: I disagree. We are only taking a sample size of the entire population of the pure community and for most community-esque awards, they tend to be skewed in favor of certain clans if we do a straight majority polling. Your football analogy is faulty comparison as a result; population measurement vis a vis sample measurement. The problems Sharkbrew faces is akin to doing your sample polling in a Brazilian village, there, they would say Naymar is the best football player. That is why all community sites implement a sort of anti biasing measure and for this award, using standard deviation seems like the painless way rather than trying to manually verify as a result, however I'd welcome any improvement on normalization techniques as I agree with you on the aspect that totally nulling isnt the way to go. Eliminating the most popular voting option as a grounds of establishing equal normal assumptions in sampling is incorrect because this is a qualitative poll. You might as well not have these awards if you’re scared that a certain clan has too many members on your site as your procedure is very flawed (imagine if every voter who thought eop was the #1 clan voted eop/aao). You could try to do what Zybez does, they prohibit individuals from voting for their own clan. DULL and 0wl 2 Link to comment Share on other sites More sharing options...
Stl Arrow Posted February 5, 2018 Report Share Posted February 5, 2018 Niceee Link to comment Share on other sites More sharing options...
C0M Posted February 5, 2018 Report Share Posted February 5, 2018 Gratz EOP. Link to comment Share on other sites More sharing options...
C0M Posted February 5, 2018 Report Share Posted February 5, 2018 Gratz Fatality. Respect Apex. Link to comment Share on other sites More sharing options...
puppyslush Posted February 5, 2018 Report Share Posted February 5, 2018 3 hours ago, K===Big Bombz said: Eliminating the most popular voting option as a grounds of establishing equal normal assumptions in sampling is incorrect because this is a qualitative poll. You might as well not have these awards if you’re scared that a certain clan has too many members on your site as your procedure is very flawed (imagine if every voter who thought eop was the #1 clan voted eop/aao). You could try to do what Zybez does, they prohibit individuals from voting for their own clan. The community awards have some qualities of a qualitative poll, but I wouldn't outright define it as such. What I am trying to achieve here is proportional representation as seen in many voting system. Unfortunately no-one in their right mind would vote for EOP/AAO pair unless being told to by a divine or higher order. The results prove that and I am glad the current system caught it. Prohibit voting for their own clan is unfortunately ruins any motivation for members from top performing clans to vote. Link to comment Share on other sites More sharing options...
auwi Posted February 5, 2018 Report Share Posted February 5, 2018 32 minutes ago, slushpuppeh said: The community awards have some qualities of a qualitative poll, but I wouldn't outright define it as such. What I am trying to achieve here is proportional representation as seen in many voting system. Unfortunately no-one in their right mind would vote for EOP/AAO pair unless being told to by a divine or higher order. The results prove that and I am glad the current system caught it. Prohibit voting for their own clan is unfortunately ruins any motivation for members from top performing clans to vote. It is entirely a qualitative poll as it isnt measuring a numerical quantity. For example, if you wanted to measure the average time it takes for a person to return to a certain point in game (in the context of a return fight), the results are numerical and will follow a bell distribution in which the majority of the observed data will be clustered around a time frame (lets say 30-45 seconds). If you obtain data points that are in the 2-3 minute time range then you can discard that data as outliers as it occurs outside of the bell distribution (outside of the 95% level of confidence that you mentioned earlier). Having proportional representation is a flawed concept because people will just vote for their own clans and there are very little people in this community that browse the forums without actively participating in clanning (whereas in the main community there is a mixed of active and retired players). To outlaw certain voting patterns just because "a certain clan leader told his clanmates to vote a certain way" sets a precedent that a sharkbrew rank can choose to believe that a certain voting pattern is unacceptable and nullify any votes that correspond with that voting pattern. If you as the founder of Sharkbrew wish to go this route then thats fine, but just understand that it is statistically flawed and will skew future sample polling that you do. You claim that prohibiting voting for a person's own clan ruins any motivation for members from top clans to vote. Well, whats the point of nulling a majority of votes for a certain clan just because they have a larger presence on this website? That completely defeats the point that you made. You claim that its fishy for people to vote for Eop and Aao.... well who has the right to justify what clans deserve to be voted on as the best clans? Isn't that the whole point of your poll? To see what the community perceives as the best clans? Link to comment Share on other sites More sharing options...
0wl Posted February 5, 2018 Report Share Posted February 5, 2018 47 minutes ago, slushpuppeh said: Unfortunately no-one in their right mind would vote for EOP/AAO pair unless being told to by a divine or higher order. The results prove that and I am glad the current system caught it. Probably the most biased thing you could claim. How would you know if people would not normally vote for such clans? How do you know what everybody and anybody could be thinking? You're not basing your system off of fact, you're basing it off of speculation and your own personal opinion. With your own justification, you have ultimately given yourself and sharkbrew the final decision to allow or disallow certain clans from winning which renders the whole voting process useless as the decision for who wins does not lie with the votes, but your own personal opinions. The community is not divided equally (eop 20, foe 20, fi 20, etc), some communities take up a larger portion of the overall pure community. DULL 1 Link to comment Share on other sites More sharing options...
Halloween Posted February 5, 2018 Report Share Posted February 5, 2018 dead awards Link to comment Share on other sites More sharing options...
puppyslush Posted February 5, 2018 Report Share Posted February 5, 2018 4 hours ago, K===Big Bombz said: It is entirely a qualitative poll as it isnt measuring a numerical quantity. For example, if you wanted to measure the average time it takes for a person to return to a certain point in game (in the context of a return fight), the results are numerical and will follow a bell distribution in which the majority of the observed data will be clustered around a time frame (lets say 30-45 seconds). If you obtain data points that are in the 2-3 minute time range then you can discard that data as outliers as it occurs outside of the bell distribution (outside of the 95% level of confidence that you mentioned earlier). One of the key difference of qualitative vs quantitative polling is the in the strength or viability of the data collected. The questions asked in a qualitative poll would be tailor made for the target audience. This isn't it. A proper qualitative polling would first establish a subject's inherent bias test then use appropriately crafted questions based on previous studies on that respective group of subjects. Quote Having proportional representation is a flawed concept because people will just vote for their own clans and there are very little people in this community that browse the forums without actively participating in clanning (whereas in the main community there is a mixed of active and retired players). To outlaw certain voting patterns just because "a certain clan leader told his clanmates to vote a certain way" sets a precedent that a sharkbrew rank can choose to believe that a certain voting pattern is unacceptable and nullify any votes that correspond with that voting pattern. If you as the founder of Sharkbrew wish to go this route then thats fine, but just understand that it is statistically flawed and will skew future sample polling that you do. Proportional representation could means we are giving everyone's vote a scaling factor with respect to the general population and believe it or not.. standard deviation is one of the methods used in to solve/mitigate this issue of voter imbalance.. For the past 2 annual awards, we observed that even when all clans heartily partake, clan members typically for for their own clan, and there is a variation when choosing the 2nd clan If such variation exists, then it is reasonable expectation that we can eliminate outliers using the voting pairs When we straight counted those awards without taking into account scaling factor, only exclusive club of clans won the awards You need to snap out of the straight majority mindset. This isn't. This voting process is a 2 part system. EOP members who didn't follow their templates had their votes counted, so did everyone else who had votes with variations with their 1st choice(FOE/UB/PX) and 2nd choice EOP. ^^ That is why EOP still won and I strongly believe this time round, the voting structure legitimatizes their win Quote You claim that prohibiting voting for a person's own clan ruins any motivation for members from top clans to vote. Well, whats the point of nulling a majority of votes for a certain clan just because they have a larger presence on this website? That completely defeats the point that you made. You claim that its fishy for people to vote for Eop and Aao.... well who has the right to justify what clans deserve to be voted on as the best clans? Isn't that the whole point of your poll? To see what the community perceives as the best clans? Read my above points I have acknowledged this system isn't entirely the best and I have asked you for solutions; But so far, despite claiming position of authority, you have given me nothing but rhetoric and catchy headline titles to stroke your own ego. I really hope that you give me something concrete in the next reply. 3 hours ago, Einstein said: Probably the most biased thing you could claim. How would you know if people would not normally vote for such clans? How do you know what everybody and anybody could be thinking? You're not basing your system off of fact, you're basing it off of speculation and your own personal opinion. With your own justification, you have ultimately given yourself and sharkbrew the final decision to allow or disallow certain clans from winning which renders the whole voting process useless as the decision for who wins does not lie with the votes, but your own personal opinions. The community is not divided equally (eop 20, foe 20, fi 20, etc), some communities take up a larger portion of the overall pure community. You are embarrassing the name of Einstein. Link to comment Share on other sites More sharing options...
iObsession Posted February 5, 2018 Report Share Posted February 5, 2018 On 04/02/2018 at 2:37 PM, Shoan said: @iObsession l000l <14:47:23> "Shoan" appears, was kicked from channel "[Foe] Minis" by "Bash Link to comment Share on other sites More sharing options...
Phebz Posted February 5, 2018 Report Share Posted February 5, 2018 How did FO recieve #1 Clan Wars, but never wins? Just curious. Grats everyone Link to comment Share on other sites More sharing options...
Shoan Posted February 5, 2018 Report Share Posted February 5, 2018 1 hour ago, iObsession said: <14:47:23> "Shoan" appears, was kicked from channel "[Foe] Minis" by "Bash opened .txt for reply but pc crashed lol Link to comment Share on other sites More sharing options...
DULL Posted February 5, 2018 Report Share Posted February 5, 2018 9 hours ago, slushpuppeh said: One of the key difference of qualitative vs quantitative polling is the in the strength or viability of the data collected. The questions asked in a qualitative poll would be tailor made for the target audience. This isn't it. A proper qualitative polling would first establish a subject's inherent bias test then use appropriately crafted questions based on previous studies on that respective group of subjects. Proportional representation could means we are giving everyone's vote a scaling factor with respect to the general population and believe it or not.. standard deviation is one of the methods used in to solve/mitigate this issue of voter imbalance.. For the past 2 annual awards, we observed that even when all clans heartily partake, clan members typically for for their own clan, and there is a variation when choosing the 2nd clan If such variation exists, then it is reasonable expectation that we can eliminate outliers using the voting pairs When we straight counted those awards without taking into account scaling factor, only exclusive club of clans won the awards You need to snap out of the straight majority mindset. This isn't. This voting process is a 2 part system. EOP members who didn't follow their templates had their votes counted, so did everyone else who had votes with variations with their 1st choice(FOE/UB/PX) and 2nd choice EOP. ^^ That is why EOP still won and I strongly believe this time round, the voting structure legitimatizes their win Read my above points I have acknowledged this system isn't entirely the best and I have asked you for solutions; But so far, despite claiming position of authority, you have given me nothing but rhetoric and catchy headline titles to stroke your own ego. I really hope that you give me something concrete in the next reply. You are embarrassing the name of Einstein. As I told you in irc pm the results aren't realistic lol. It seems to me and majority of others you just gave awards to other clans to please them so they don't call you "EoP biased" That happens regardless man. There is no way you can claim fo is best clan wars clan when they haven't won a p2p prep in 2 months lul. Strongest core sup? No offence but strongest core of what? pulling 20 people for few months? eop's been doing that but with 80 people. Funniest one is strongest rank team tie with px lol. Sorry not sorry there's no comparison. End of the day whether you like it or not you gave out awards to other clans just to please..and not be called eop biased...sad.. Link to comment Share on other sites More sharing options...
0wl Posted February 5, 2018 Report Share Posted February 5, 2018 Looks like the rest of the awards were drawn out of a hat because somebody doesn't want other clans to realize they all lost to eop again l0l! That would be bad for business since it would make other clans feel sad and not want to post here... Looks like it's just proving to everyone that the focus of sharkbrew is business over the community lol Link to comment Share on other sites More sharing options...
auwi Posted February 5, 2018 Report Share Posted February 5, 2018 12 hours ago, slushpuppeh said: One of the key difference of qualitative vs quantitative polling is the in the strength or viability of the data collected. The questions asked in a qualitative poll would be tailor made for the target audience. This isn't it. A proper qualitative polling would first establish a subject's inherent bias test then use appropriately crafted questions based on previous studies on that respective group of subjects. Proportional representation could means we are giving everyone's vote a scaling factor with respect to the general population and believe it or not.. standard deviation is one of the methods used in to solve/mitigate this issue of voter imbalance.. For the past 2 annual awards, we observed that even when all clans heartily partake, clan members typically for for their own clan, and there is a variation when choosing the 2nd clan If such variation exists, then it is reasonable expectation that we can eliminate outliers using the voting pairs When we straight counted those awards without taking into account scaling factor, only exclusive club of clans won the awards You need to snap out of the straight majority mindset. This isn't. This voting process is a 2 part system. EOP members who didn't follow their templates had their votes counted, so did everyone else who had votes with variations with their 1st choice(FOE/UB/PX) and 2nd choice EOP. ^^ That is why EOP still won and I strongly believe this time round, the voting structure legitimatizes their win Read my above points I have acknowledged this system isn't entirely the best and I have asked you for solutions; But so far, despite claiming position of authority, you have given me nothing but rhetoric and catchy headline titles to stroke your own ego. I really hope that you give me something concrete in the next reply. You are embarrassing the name of Einstein. The point is you're not using any sort of scaling factor or method to normalize the vote. You're completely getting rid of a certain type of vote because you believe it doesnt align with YOUR PRECEPTION of the "best clan." This in itself is the fault in your polling methods. To be honest, I dont really care about the outcome of the polls, and regardless if it affects eop or not, it's just a pixelated community award. But the procedure in which you obtained your results is incredibly faulty and would uphold the notion that many feel that this site is "biased." I'm not saying that you are, but in order to have a proper procedure to adequately appropriate certain awards, you cant just delete a certain demographic of votes. I don't have a majority mindset, and I certaintly wouldn't be arguing with you on this issue if I didnt have a Statistics degree itself. Because people would just vote for their clans anyway and you are (for whatever reason) removing votes because these votes have a common second answer, why not just null the second choice and include the first choice? But then again this is a biased process because this means that sharkbrew ranks would have identify among themselves which clans are worthy to be voted the best clans and which vote is "troll." Again this is biased, but if this is the route you are going for then so be it. In qualitative polling, it is important to understand the context of voting variances. In most qualitative studies, the data is left as it is but the researcher provides his thoughts and analysis of the results with voting. No one in the scientific community just "nullifies" a certain vote because of something that they believe is faulty. This is akin to studies regarding smoking in the USA, where tobacco lobbyists would pay off researchers to omit certain forms of data from health studies to establish tobacco as a "healthy" practice. Like I've mentioned many times before, this type of polling is FLAWED. I'm not gonna comment on your ego or anything, I am purely criticizing the faultiness of your qualitative sampling methods. If you think my word usage is "catchy" then thats up for you to decide. DULL and 17_ 2 Link to comment Share on other sites More sharing options...
Tannie Posted February 6, 2018 Report Share Posted February 6, 2018 wow i won something finally for once Link to comment Share on other sites More sharing options...
puppyslush Posted February 6, 2018 Report Share Posted February 6, 2018 13 hours ago, K===Big Bombz said: The point is you're not using any sort of scaling factor or method to normalize the vote. You're completely getting rid of a certain type of vote because you believe it doesnt align with YOUR PRECEPTION of the "best clan." ...... I'm not saying that you are, but in order to have a proper procedure to adequately appropriate certain awards, you cant just delete a certain demographic of votes. I have admitted that at the beginning in my original post that this particular method of completely nulling certain segment of the votes isn't the best case, but you keep insisting on giving me suggestions that are either: stupidly impractical to implement straight majority that does not reflect the true sentiment of the pure community Your suggestions are akin to polling Chinese people in China asking them which food they like best but telling them that they can't vote for rice or noodles because it would be bias lolzz. 13 hours ago, K===Big Bombz said: I don't have a majority mindset, and I certaintly wouldn't be arguing with you on this issue if I didnt have a Statistics degree itself. Because people would just vote for their clans anyway and you are (for whatever reason) removing votes because these votes have a common second answer, why not just null the second choice and include the first choice? But then again this is a biased process because this means that sharkbrew ranks would have identify among themselves which clans are worthy to be voted the best clans and which vote is "troll." Again this is biased, but if this is the route you are going for then so be it. Why would we introduce contingent voting system if we are going to eliminate one of the 2 votes??? And as I have fucking explained before, it is incredibly difficult to determine the clan of a poster as clan allegiance change weekly which begs the question which 1 to remove? Isn't that more bias? 13 hours ago, K===Big Bombz said: In qualitative polling, it is important to understand the context of voting variances. In most qualitative studies, the data is left as it is but the researcher provides his thoughts and analysis of the results with voting. Qualitative studies don't need to touch with the data BECAUSE the questions prepped for the subject audience. Moreover you keep interchangeably use qualitative voting/study/data when they all mean 3 difference things. Make up your damn mind. 13 hours ago, K===Big Bombz said: No one in the scientific community just "nullifies" a certain vote because of something that they believe is faulty. This is akin to studies regarding smoking in the USA, where tobacco lobbyists would pay off researchers to omit certain forms of data from health studies to establish tobacco as a "healthy" practice. Like I've mentioned many times before, this type of polling is FLAWED. I have already acknowledged that eliminating a whole segment of votes isn't the best way forward and I have asked you twice to give me a reasonable suggestion, but you keep insisting on bringing that point up to try and put the system down because there is absolutely no creative thoughts in that cranium of yours. It is an honest assessment of your behavior. Despite your "statistics degree", you cannot argue on or refute my technical points, instead choosing to keep relying on emotions and rhetoric, which makes me wonder whether you got your degree confused with that of a social science one.. Moreover, you seem to keep harping on your educational qualifications instead of refuting or suggesting solutions to this problem inspite of me explaining the depth of the problem I face and the thought process that was used to create this solution. Link to comment Share on other sites More sharing options...
auwi Posted February 6, 2018 Report Share Posted February 6, 2018 7 hours ago, slushpuppeh said: I have admitted that at the beginning in my original post that this particular method of completely nulling certain segment of the votes isn't the best case, but you keep insisting on giving me suggestions that are either: stupidly impractical to implement straight majority that does not reflect the true sentiment of the pure community Your suggestions are akin to polling Chinese people in China asking them which food they like best but telling them that they can't vote for rice or noodles because it would be bias lolzz. Why would we introduce contingent voting system if we are going to eliminate one of the 2 votes??? And as I have fucking explained before, it is incredibly difficult to determine the clan of a poster as clan allegiance change weekly which begs the question which 1 to remove? Isn't that more bias? Qualitative studies don't need to touch with the data BECAUSE the questions prepped for the subject audience. Moreover you keep interchangeably use qualitative voting/study/data when they all mean 3 difference things. Make up your damn mind. I have already acknowledged that eliminating a whole segment of votes isn't the best way forward and I have asked you twice to give me a reasonable suggestion, but you keep insisting on bringing that point up to try and put the system down because there is absolutely no creative thoughts in that cranium of yours. It is an honest assessment of your behavior. Despite your "statistics degree", you cannot argue on or refute my technical points, instead choosing to keep relying on emotions and rhetoric, which makes me wonder whether you got your degree confused with that of a social science one.. Moreover, you seem to keep harping on your educational qualifications instead of refuting or suggesting solutions to this problem inspite of me explaining the depth of the problem I face and the thought process that was used to create this solution. First of all, I'm enjoying this discussion because it just goes to show how hurt you get when someone intellectual criticizes your methods. You continue to attack me personally while I try to keep things on topic. Whose the one stroking his own ego? In reference to your main response, I haven't provided any of my ideas because each time I point out a blatant error in your topic, you defend it by providing false information that no researcher would ever implement in his data. Since you're confused about my usage of the words data/study/voting, in this context it is basically the same and you seem disheveled by my usage of those words. Why not focus on what I'm saying and not point out inconsistencies in my word usage? You keep claiming that I'm spewing rhetoric, when in reality I'm sufficiently counter-arguing your points. For example, you said that you implemented a scaling factor in the poll. I point out to you that your scaling factor (removing a certain demographic of votes) is not in fact a scaling factor. In response, you freak out, claiming that I haven't offered a viable solution and that my Statistics degree is made up. You're the founder of a community site yet you're acting like a child. A qualitative study is one that seeks to understand or learn a certain group of people's opinions, ideas, and circumstances. If you were running a TRUE qualitative study you would accept the votes that were given on the topic instead of trying to come up with a method of normalizing the data. I'm not saying this is a full proof method, but the outcome of the votes would undoubtedly be based on which clan has the highest number of posters on this website. What I'm stating is that its very hard to determine which clans deserve which award, but in the context of a community poll, you cannot just delete observations because they are "biased." This is like trying to fix a wrong with a wrong, trying to put out a fire by using fire, or trying to fight ddos with ddos. It just doesn't work, and something this website should avoid for it to remain an un-biased community forum. 17_ 1 Link to comment Share on other sites More sharing options...
K0nt Posted February 6, 2018 Report Share Posted February 6, 2018 Holydreams lmfao . gz on all the awards Link to comment Share on other sites More sharing options...
Pker4life Posted February 6, 2018 Report Share Posted February 6, 2018 Ob baby Link to comment Share on other sites More sharing options...
Pker4life Posted February 6, 2018 Report Share Posted February 6, 2018 Cool Link to comment Share on other sites More sharing options...
Recommended Posts
Create an account or sign in to comment
You need to be a member in order to leave a comment
Create an account
Sign up for a new account in our community. It's easy!
Register a new accountSign in
Already have an account? Sign in here.
Sign In Now