issue 26 Patch Notes for May 30th, 2019

[Nericus]

My only issue with people using DFB to level up so much, is that we have a lot of NEW players (somehow) who were sucked into it and have either nor idea how to actually play past that point, OR, people who haven't done the minimal amount of unlocks needed for certain things.

 

It's common to be recruiting for an ITF and someone has to ask "what zone" and we have to explain how to get the Midnighter badge. Or recruiting for an I-trial and people try to join without having Alpha unlocked. Or trying to get in on a Mothership raid when they haven't done Levantra's arc yet. Just today I did a DiB badge run, and someone asked (I'm not sure if joking or in exasperation) if they had to unlock a certain location or have a certain badge first.

 

That's the practical side of it. On a more personal note, I'm irked that so many NEW players are (inadvertently?) avoiding a lot of content. Yes, anything made before issue... 6 let's say, really kinda sucks. "Go to this warehouse. Click glowie, don't bother reading the clue. Kill skulz at the end. Run around outside, kil 10 skulz. Go to this warehouse, kill ALL the skulz. Repeat a few times" But a lot of the newer stuff like Twinshot's arc, the new Skull arc, are really good! A lot of the game is generic "kill dudes with flimsy justification", I get that. But we're doing people a disservice by pretending the game doesn't exist pre-50.

 

I mean hell, I've seen people with level 20+ characters asking how to use enhancements and other basic stuff! I don't wanna tell anyone how to play, but maybe some people would be better served by playing City of Heroes "right", instead of "Super Powered Sanitation Workers".

 

I have encountered veteran players, or at least they say they were as well as people new to the game and both ask questions as the vets have forgotten things and the new players do not know.

 

Once upon a Time me, you, myself, and many others were new to the game and I bet many of us when new got pulled into the latest farm craze and out leveled many things.  Ouroborus now fixed that as people can go back and be powered down to the correct level.

 

I also remind the returning vets and newcomers of sites like paragonwiki and my pal's CoH YouTube channel fo their information needs.

 

We cannot expect people to know everything right away be they new or returning.

 

Also there was always a problem as it were back in the first era of CoH with power leveled players that didn't know the game

[justicebeliever]

Actually, I ran a statistical inference and regression anova analysis on this from the responses in the thread. Because there were more than 30 distinct sample responses, it is a statistically viable pool. With the alpha at 5%, confidence level of 95%, it is statistically accurate that the community supports this, as demonstrated by the sample numbers.

 

Being very familiar with statistical analysis, a sample size of 30 or more has been shown to be statistically representitive of the population. Data, and proof.

 

You were saying?

 

I’m not arguing but for a population in 50-90k range, 30 seems like too small a sample size

[SwitchFade]

Actually, I ran a statistical inference and regression anova analysis on this from the responses in the thread. Because there were more than 30 distinct sample responses, it is a statistically viable pool. With the alpha at 5%, confidence level of 95%, it is statistically accurate that the community supports this, as demonstrated by the sample numbers.

 

Being very familiar with statistical analysis, a sample size of 30 or more has been shown to be statistically representitive of the population. Data, and proof.

 

You were saying?

 

I’m not arguing but for a population in 50-90k range, 30 seems like too small a sample size

 

When I first began learning statistics, I felt the same. Mathematical evidence has shown that 30 or more is accurate, at around a 95% confidence with a margin of error of +/- 3%.

 

Ipsos and all other polls use the same mathematical standards proofed over the last 200 years or so.

[Saikochoro]

 

The changes to DFB/MSR we’re not improvements in the sight of all. Nor were they supported by the entire community as this thread alone has shown. Several people do indeed support the changes, and several people do not. The community is split on the issue. The devs just happened to agree that change was needed. However, it is wrong to say the community needs and supports it, as only part of the community sees it that way. There are always two sides.

 

A quick count of people disagreeing with the changes in this 11 page thread is 6 vocal people everyone else seems fine with the Dev's balancing efforts.

 

Thanks for all your hard work Dev's

 

There are plenty of thread about it all with differing opinions. The original comment I responded to implied that the community in whole needed and supported this change. That is completely false. There is a part of the community that does not support it. I never said one was bigger than the other. Only that it is wrong to assume the community supports it as a whole. It is better to say, “part of the community supports this”. That is undoubtedly true. Saying the community needs this however is fully subjective and shouldn’t really be said at all. I’m not going to say the community needs DFB nor will I say the community needed it to be nerfed. Either statement is purely based on opinion. I will says that the nerf was definitely wanted by some and very much opposed by others.

 

Actually, not false, validated by data. 6 people disagree. Out of multiples of 10 agreeing. So, the majority approve, therefor in a community, that is majority support.

 

Not everyone agrees with mandatory seatbelt laws. However, the majority approve, thus, the community supports it.

 

You clearly have a problem understanding representative statistical samples. This thread is not representative of the population and cannot be used to accurately gauge the community as a whole. It can however show that the community is at least split on this issue purely due to the fact that there is more than one opinion expressed. Even if all posts were positive or all posts were negative it still would not be an appropriate statistical sample. That’s why I didn’t say, the community is not in support of the DFB change. I said, part of the community doesn’t support it. That is a true statement.  Saying the majority of the community supports it is false and is not backed up by real data.

 

Also, comparing a computer game to seatbelts is ridiculous. It’s just like in the open letter announcement someone comparing computer game opinions to gun safety. One is real world and one is a game. One has no influence on real world safety. One can result in real world death.  That is not a good faith comparison in the slightest.

 

Actually, I ran a statistical inference and regression anova analysis on this from the responses in the thread. Because there were more than 30 distinct sample responses, it is a statistically viable pool. With the alpha at 5%, confidence level of 95%, it is statistically accurate that the community supports this, as demonstrated by the sample numbers.

 

Being very familiar with statistical analysis, a sample size of 30 or more has been shown to be statistically representitive of the population. Data, and proof.

 

You were saying?

 

You can perform any test you want on a certain sample, but if the sample is itself is not representative all of the outputs are not reliable. Before you can even begin to pull a representative sample you must first clearly define the population and then be sure that the population is complete. Before you can define the population and test for completeness you have to do a proper risk assessment to set the proper parameters of the test. Then you must set materiality based on various risk factors after a thorough analysis of the population. Once you have done a risk assessment, set materiality, set the parameters, defined the population, and tested it for completeness you can then pull your sample. Pulling the sample itself has various methods, but it must first be based on a complete population and have proper parameters set.

 

All that is to even have a populationto pull the sample. Sample size is directly related to risk. Then the sample itself must be tested for accuracy before it can be relied on.

 

Neither of us has all of the data points necessary from this thread to have a complete and accurate population, let alone a reliable sample to form a reasonable conclusion with any degree of accuracy. They way in which the sample is pulled and how parameters are determined is most important.

 

Edit: deleted unnecessarily rude and demeaning sentences. I would give myself -1 inf I I could.

[SwitchFade]

 

The changes to DFB/MSR we’re not improvements in the sight of all. Nor were they supported by the entire community as this thread alone has shown. Several people do indeed support the changes, and several people do not. The community is split on the issue. The devs just happened to agree that change was needed. However, it is wrong to say the community needs and supports it, as only part of the community sees it that way. There are always two sides.

 

A quick count of people disagreeing with the changes in this 11 page thread is 6 vocal people everyone else seems fine with the Dev's balancing efforts.

 

Thanks for all your hard work Dev's

 

There are plenty of thread about it all with differing opinions. The original comment I responded to implied that the community in whole needed and supported this change. That is completely false. There is a part of the community that does not support it. I never said one was bigger than the other. Only that it is wrong to assume the community supports it as a whole. It is better to say, “part of the community supports this”. That is undoubtedly true. Saying the community needs this however is fully subjective and shouldn’t really be said at all. I’m not going to say the community needs DFB nor will I say the community needed it to be nerfed. Either statement is purely based on opinion. I will says that the nerf was definitely wanted by some and very much opposed by others.

 

Actually, not false, validated by data. 6 people disagree. Out of multiples of 10 agreeing. So, the majority approve, therefor in a community, that is majority support.

 

Not everyone agrees with mandatory seatbelt laws. However, the majority approve, thus, the community supports it.

 

You clearly have a problem understanding representative statistical samples. This thread is not representative of the population and cannot be used to accurately gauge the community as a whole. It can however show that the community is at least split on this issue purely due to the fact that there is more than one opinion expressed. Even if all posts were positive or all posts were negative it still would not be an appropriate statistical sample. That’s why I didn’t say, the community is not in support of the DFB change. I said, part of the community doesn’t support it. That is a true statement.  Saying the majority of the community supports it is false and is not backed up by real data.

 

Also, comparing a computer game to seatbelts is ridiculous. It’s just like in the open letter announcement someone comparing computer game opinions to gun safety. One is real world and one is a game. One has no influence on real world safety. One can result in real world death.  That is not a good faith comparison in the slightest.

 

Actually, I ran a statistical inference and regression anova analysis on this from the responses in the thread. Because there were more than 30 distinct sample responses, it is a statistically viable pool. With the alpha at 5%, confidence level of 95%, it is statistically accurate that the community supports this, as demonstrated by the sample numbers.

 

Being very familiar with statistical analysis, a sample size of 30 or more has been shown to be statistically representitive of the population. Data, and proof.

 

You were saying?

 

Still clearly do not understand representative statistical samples. You can perform any test you want on a certain sample, but if the sample is itself is not representative all of the outputs are not reliable. Before you can even begin to pull a representative sample you must first clearly define the population and then be sure that the population is complete. Before you can define the population and test for completeness you have to do a proper risk assessment to set the proper parameters of the test. Then you must set materiality based on various risk factors after a thorough analysis of the population. Once you have done a risk assessment, set materiality, set the parameters, defined the population, and tested it for completeness you can then pull your sample. Pulling the sample itself has various methods, but it must first be based on a complete population and have proper parameters set.

 

All that is to even have a populationto pull the sample. Sample size is directly related to risk. Then the sample itself must be tested for accuracy before it can be relied on.

 

Neither of us has all of the data points necessary from this thread to have a complete and accurate population, let alone a reliable sample to form a reasonable conclusion with any degree of accuracy. Using buzzwords does not change the fundamentals of statistical analysis. I highly doubt you are familiar with statistical analysis based on your few posts about it and if you were I would never hire you to fill my team. Your posts show you clearly lack a basic understanding of stastival analysis.

 

I appreciate your attempt to explain your assumptions. I don't appreciate, however, your rather rude comments about my understanding of statistics. I would ask that you please be more polite in the future and refrain from making unfounded assumptions of which you know naught.

 

Your statements about statistics are based in fact, but misinterpreted. I will demonstrate with a few examples.

 

Population is a set of all, in this case the players of the game.

 

Sample is a set within a population.

 

Population: N

 

Sample: n

 

Now, sample bias may be occuring, yes, which you sort of attempt to explain. To account for bias, a MARGIN FOR ERROR is considered, as I have mentioned.

 

Margin for error: +/-

 

If the sample size is >30, the sample is valid and statistically accurate.

 

If the r square value is greater than 80%, the regression CLEARLY shows that one variable affects another. In this case this condition is met.

 

If the value in the bell curve falls under 5%, the alpha, then the hypothesis has been "validated." Condition met.

 

Even then, a sample can be proven WITHOUT a known population, as the median of a sample is the median of the population, this is mathematically known and proven. Further, central limit and laws of large numbers are mathematically observed laws. Additionally, although we must calculate sample size when population is unknown, as I have, this condition is ALSO MET.

 

Population known. Sample size large enough. R square condition significant. Value under curve below alpha. Sampling error does occur, ACCOUNTED FOR.

 

Statistically, the community IN MAJORITY (51% or greater) supports the change. If you want to debate "majority," please feel free. I would encourage you to then seek appropriate venues, such as Miriam Webster or any other such official publication concerning the commonly accepted meanings of English words

 

If you want to debate statistics, there is no debate, simple mathematical rules, all established, and met in this case.

 

Please refrain from incendiary statements concerning intellectual capacity. I assume you are highly intelligent, I ask you do reciprocate.

 

Thank you for your respectful participation.

[Saikochoro]

 

The changes to DFB/MSR we’re not improvements in the sight of all. Nor were they supported by the entire community as this thread alone has shown. Several people do indeed support the changes, and several people do not. The community is split on the issue. The devs just happened to agree that change was needed. However, it is wrong to say the community needs and supports it, as only part of the community sees it that way. There are always two sides.

 

A quick count of people disagreeing with the changes in this 11 page thread is 6 vocal people everyone else seems fine with the Dev's balancing efforts.

 

Thanks for all your hard work Dev's

 

There are plenty of thread about it all with differing opinions. The original comment I responded to implied that the community in whole needed and supported this change. That is completely false. There is a part of the community that does not support it. I never said one was bigger than the other. Only that it is wrong to assume the community supports it as a whole. It is better to say, “part of the community supports this”. That is undoubtedly true. Saying the community needs this however is fully subjective and shouldn’t really be said at all. I’m not going to say the community needs DFB nor will I say the community needed it to be nerfed. Either statement is purely based on opinion. I will says that the nerf was definitely wanted by some and very much opposed by others.

 

Actually, not false, validated by data. 6 people disagree. Out of multiples of 10 agreeing. So, the majority approve, therefor in a community, that is majority support.

 

Not everyone agrees with mandatory seatbelt laws. However, the majority approve, thus, the community supports it.

 

You clearly have a problem understanding representative statistical samples. This thread is not representative of the population and cannot be used to accurately gauge the community as a whole. It can however show that the community is at least split on this issue purely due to the fact that there is more than one opinion expressed. Even if all posts were positive or all posts were negative it still would not be an appropriate statistical sample. That’s why I didn’t say, the community is not in support of the DFB change. I said, part of the community doesn’t support it. That is a true statement.  Saying the majority of the community supports it is false and is not backed up by real data.

 

Also, comparing a computer game to seatbelts is ridiculous. It’s just like in the open letter announcement someone comparing computer game opinions to gun safety. One is real world and one is a game. One has no influence on real world safety. One can result in real world death.  That is not a good faith comparison in the slightest.

 

Actually, I ran a statistical inference and regression anova analysis on this from the responses in the thread. Because there were more than 30 distinct sample responses, it is a statistically viable pool. With the alpha at 5%, confidence level of 95%, it is statistically accurate that the community supports this, as demonstrated by the sample numbers.

 

Being very familiar with statistical analysis, a sample size of 30 or more has been shown to be statistically representitive of the population. Data, and proof.

 

You were saying?

 

Still clearly do not understand representative statistical samples. You can perform any test you want on a certain sample, but if the sample is itself is not representative all of the outputs are not reliable. Before you can even begin to pull a representative sample you must first clearly define the population and then be sure that the population is complete. Before you can define the population and test for completeness you have to do a proper risk assessment to set the proper parameters of the test. Then you must set materiality based on various risk factors after a thorough analysis of the population. Once you have done a risk assessment, set materiality, set the parameters, defined the population, and tested it for completeness you can then pull your sample. Pulling the sample itself has various methods, but it must first be based on a complete population and have proper parameters set.

 

All that is to even have a populationto pull the sample. Sample size is directly related to risk. Then the sample itself must be tested for accuracy before it can be relied on.

 

Neither of us has all of the data points necessary from this thread to have a complete and accurate population, let alone a reliable sample to form a reasonable conclusion with any degree of accuracy. Using buzzwords does not change the fundamentals of statistical analysis. I highly doubt you are familiar with statistical analysis based on your few posts about it and if you were I would never hire you to fill my team. Your posts show you clearly lack a basic understanding of stastival analysis.

 

I appreciate your attempt to explain your assumptions. I don't appreciate, however, your rather rude comments about my understanding of statistics. I would ask that you please be more polite in the future and refrain from making unfounded assumptions of which you know naught.

 

Your statements about statistics are based in fact, but misinterpreted. I will demonstrate with a few examples.

 

Population is a set of all, in this case the players of the game.

 

Sample is a set within a population.

 

Population: N

 

Sample: n

 

Now, sample bias may be occuring, yes, which you sort of attempt to explain. To account for bias, a MARGIN FOR ERROR is considered, as I have mentioned.

 

Margin for error: +/-

 

If the sample size is >30, the sample is valid and statistically accurate.

 

If the r square value is greater than 80%, the regression CLEARLY shows that one variable affects another. In this case this condition is met.

 

If the value in the bell curve falls under 5%, the alpha, then the hypothesis has been "validated." Condition met.

 

Even then, a sample can be proven WITHOUT a known population, as the median of a sample is the median of the population, this is mathematically known and proven. Further, central limit and laws of large numbers are mathematically observed laws. Additionally, although we must calculate sample size when population is unknown, as I have, this condition is ALSO MET.

 

Population known. Sample size large enough. R square condition significant. Value under curve below alpha. Sampling error does occur, ACCOUNTED FOR.

 

Statistically, the community IN MAJORITY (51% or greater) supports the change. If you want to debate "majority," please feel free. I would encourage you to then seek appropriate venues, such as Miriam Webster or any other such official publication concerning the commonly accepted meanings of English words

 

If you want to debate statistics, there is no debate, simple mathematical rules, all established, and met in this case.

 

Please refrain from incendiary statements concerning intellectual capacity. I assume you are highly intelligent, I ask you do reciprocate.

 

Thank you for your respectful participation.

 

My tone was incindiary and mocking. The last sentence I used was uncalled for and I am editing my original post. As it has been quoted already, it can serve as a lesson for others to not be rude. I apologize for the attack on intellectual capacity. You responded in a very professional and impressively calm manner given the tone of my post. +1 inf.

 

You do seem an intelligent person. Perhaps it is our education and perhaps professional field differences that have led to this (and emotion on my part). In my profession mathematics rule in the actual results of a sample. But if the population is found to be in error in any manner it absolutely cannot be relied upon. Even though sample is based on risk, which we both agree upon, and even though math may state that the risk is accounted for, it still cannot be relied upon. The population must be complete and accurate first and foremost. If it is found to be in any way incomplete or inaccurate the sample is considered to be unreliable and must be repulled with the new information without exception. That is a lessen I had to very painfully learn early on in my career resulting in many long nights and weekends at work. I’m talking getting home at 3am and getting back up at 5am to go to work. That is most likely what colored my response and why I very quickly became heated as it was also something I had to teach my interns and staff in turn (resulting in even more long nights and weekends). I literally got grey hairs in my 20s due to the stress involved and almost lost my marriage. So it brought me back to a dark place in my life. Still that is not an excuse for rudeness. That said, I have always thought my profession did certain things wrong. Perhaps this is one of them. If this is the case, I apologize for setting it as absolute.

 

I will do my best to refrain from personal attack such as what I did. I do apologize and hope that you accept my apology. Thank you for being civilized and calmly reprimanding my behavior.

[SwitchFade]

Saiko,

 

You are truly a gentleman, and we both are clearly very conversant with mathematical theory and application. Your expertise surely outshines mine.

 

I am humbled to be able to participate in the forums and the game with people such as yourself, and I truly believe this community needs you.

 

I hope we may team in game in the future, and I look forward to learning from your wisdom.

 

There is no fault in your posts, and no conversation exists from a single perspective, thus, if I have somehow caused distress I TOO am responsible, and apologize profusely.

 

Thank you for being part of our community and allowing me to learn from you.

[Healix]

The last two posts are examples of why the CoH community has remained one of the best.

[vallerainth]

The last two posts are examples of why the CoH community has remained one of the best.

 

I agree. Though we may have drastically different views at the end of the day we can step back and appreciate the other person's perceptive, as well as question our own. Then we put on our tights and go take care of business.

[Sanguinesun]

 

 

I appreciate your attempt to explain your assumptions. I don't appreciate, however, your rather rude comments about my understanding of statistics. I would ask that you please be more polite in the future and refrain from making unfounded assumptions of which you know naught.

 

Your statements about statistics are based in fact, but misinterpreted. I will demonstrate with a few examples.

 

Population is a set of all, in this case the players of the game.

 

Sample is a set within a population.

 

Population: N

 

Sample: n

 

Now, sample bias may be occuring, yes, which you sort of attempt to explain. To account for bias, a MARGIN FOR ERROR is considered, as I have mentioned.

 

Margin for error: +/-

 

If the sample size is >30, the sample is valid and statistically accurate.

 

If the r square value is greater than 80%, the regression CLEARLY shows that one variable affects another. In this case this condition is met.

 

If the value in the bell curve falls under 5%, the alpha, then the hypothesis has been "validated." Condition met.

 

Even then, a sample can be proven WITHOUT a known population, as the median of a sample is the median of the population, this is mathematically known and proven. Further, central limit and laws of large numbers are mathematically observed laws. Additionally, although we must calculate sample size when population is unknown, as I have, this condition is ALSO MET.

 

Population known. Sample size large enough. R square condition significant. Value under curve below alpha. Sampling error does occur, ACCOUNTED FOR.

 

Statistically, the community IN MAJORITY (51% or greater) supports the change. If you want to debate "majority," please feel free. I would encourage you to then seek appropriate venues, such as Miriam Webster or any other such official publication concerning the commonly accepted meanings of English words

 

If you want to debate statistics, there is no debate, simple mathematical rules, all established, and met in this case.

 

Please refrain from incendiary statements concerning intellectual capacity. I assume you are highly intelligent, I ask you do reciprocate.

 

Thank you for your respectful participation.

 

I have a question then to pose to both you and Saiko concerning this discussion:

 

We have a representation of a subset of a subset of the player population here in this discussion:

 

A. People who are playing(The whole)

B. People who have decided to participate on the forum(Subset 1)

C. People who have decided to voice their for/against concerns related to the DFB changes on these forums. (Subset of Subset 1)

 

People not represented(but implied) are those who:

 

a) do not have a forum account.

b) have an account but have not seen this discussion

c) have seen this discussion but do not proffer a dislike or like on the forums about DFB.

d) do not offer a like or dislike in any medium.

e) do not have an opinion that is applicable for what ever reason(leveled differently etc etc)

 

 

And I am sure there are other variables that I am not accounting for in the broader sense as well in these regards.

And I am sure there are other variables that I am not accounting for in the more microcosmic level.

 

So given the uncertainties and number of variables that perhaps not a small number of people see, here is my question:

 

How does one, with any confidence, come up with -only- a margin of error of +/- 3% to the statistical analysis of who is for/against these changes? 

 

The unaccountable variables would seem to skew the analysis to the point of being unrepresentative and thus making the evidence rather compelling that one should question the validation of the conclusion.

 

 

 

1. Bias then

[_NOPE_]

a) Is gone, because to be able to play the game, you have to have a forum account... though you don't have to use it.

[Sanguinesun]

a) Is gone, because to be able to play the game, you have to have a forum account... though you don't have to use it.

 

I understand your desire to participate in this(as you general do seem to like post rather prolifically in many different discussions) but I was mainly directing this to the two aforementioned people. But to address your dismissiveness, would anyone prior to the start of it being public be subjected to that?  In addition would one truly be considered "have an account" in regards to those who just simply wanted to play the game and ignore the account truly count in what I was attempting to say "ie people who play but don't really have the account in their minds/use the account/"  as a distinction from b, those having an account who do participate on the forums but just not in this thread due to having not read it. etc etc.

 

Thus the distinction made with a) has validation as a variable just perhaps not in the worded meaning that you wish to be so quickly dismissive of.

 

 

[WolfWings]

I just wish the DFB xp didn't start to decline until a little closer to 22 so that it's easier to get into level 25 IOs. Once I get those I stop worrying about money entirely and just play for fun/content up till around 42, then start looking for IO sets and set bonuses.

 

I'd second this, tweaking things to start at 12 and drop to 50% XP at 22 (or even 15/25?) would taper off smoothly into level-25-IO range.

 

Otherwise this whole patch is wonderful, ESPECIALLY the "immob doesn't block knock* from working" stuff! :D My Earth/Thorn Dom thanks ya'all greatly for that part. =^.^=

[Saikochoro]

 

 

I appreciate your attempt to explain your assumptions. I don't appreciate, however, your rather rude comments about my understanding of statistics. I would ask that you please be more polite in the future and refrain from making unfounded assumptions of which you know naught.

 

Your statements about statistics are based in fact, but misinterpreted. I will demonstrate with a few examples.

 

Population is a set of all, in this case the players of the game.

 

Sample is a set within a population.

 

Population: N

 

Sample: n

 

Now, sample bias may be occuring, yes, which you sort of attempt to explain. To account for bias, a MARGIN FOR ERROR is considered, as I have mentioned.

 

Margin for error: +/-

 

If the sample size is >30, the sample is valid and statistically accurate.

 

If the r square value is greater than 80%, the regression CLEARLY shows that one variable affects another. In this case this condition is met.

 

If the value in the bell curve falls under 5%, the alpha, then the hypothesis has been "validated." Condition met.

 

Even then, a sample can be proven WITHOUT a known population, as the median of a sample is the median of the population, this is mathematically known and proven. Further, central limit and laws of large numbers are mathematically observed laws. Additionally, although we must calculate sample size when population is unknown, as I have, this condition is ALSO MET.

 

Population known. Sample size large enough. R square condition significant. Value under curve below alpha. Sampling error does occur, ACCOUNTED FOR.

 

Statistically, the community IN MAJORITY (51% or greater) supports the change. If you want to debate "majority," please feel free. I would encourage you to then seek appropriate venues, such as Miriam Webster or any other such official publication concerning the commonly accepted meanings of English words

 

If you want to debate statistics, there is no debate, simple mathematical rules, all established, and met in this case.

 

Please refrain from incendiary statements concerning intellectual capacity. I assume you are highly intelligent, I ask you do reciprocate.

 

Thank you for your respectful participation.

 

I have a question then to pose to both you and Saiko concerning this discussion:

 

We have a representation of a subset of a subset of the player population here in this discussion:

 

A. People who are playing(The whole)

B. People who have decided to participate on the forum(Subset 1)

C. People who have decided to voice their for/against concerns related to the DFB changes on these forums. (Subset of Subset 1)

 

People not represented(but implied) are those who:

 

a) do not have a forum account.

b) have an account but have not seen this discussion

c) have seen this discussion but do not proffer a dislike or like on the forums about DFB.

d) do not offer a like or dislike in any medium.

e) do not have an opinion that is applicable for what ever reason(leveled differently etc etc)

 

 

And I am sure there are other variables that I am not accounting for in the broader sense as well in these regards.

And I am sure there are other variables that I am not accounting for in the more microcosmic level.

 

So given the uncertainties and number of variables that perhaps not a small number of people see, here is my question:

 

How does one, with any confidence, come up with -only- a margin of error of +/- 3% to the statistical analysis of who is for/against these changes? 

 

The unaccountable variables would seem to skew the analysis to the point of being unrepresentative and thus making the evidence rather compelling that one should question the validation of the conclusion.

 

 

 

1. Bias then

 

Your question is the reason I was quick to discount the earlier statistic. In my profession, this is solved in the risk assessment and population definition phase. We would first do an risk assessment:

 

-who will even know about this change

-who will have an opinion on this change

-who is able to express an opinion on this change

-who is likely to actually express AN opinion on this change

-who is likely to express their own opinion on this change

-who is likely to express an opinion that aligns with the actual change regardless of their actual opinion

-who is likely to express an opinion with the majority regardless of their actual opinion

-who is likely to express an opinion to be in the minority regardless of their opinion

-who is likely to actually express their own opinion regardless of any circumstance

-who would have expressed their opinion, but decided to refrain because they saw they were outnumbered

-who would not have expressed their opinion, but did either to support the apparent minority or majority

- queue as many other relevant question to determine what actually defines the population

 

Then you have to get into the “why” for every one of those questions. Then you have to weigh how much risk each question has in influencing the opinion on anyone chosen in the sample. The risk of how often it may occur and the magnitude of the influence have to both be taken into account for each variable. Frequency and magnitude are different. A high frequency risk would be that the variable may affect day 10 people, but only slightly. A high magnitude risk is might only affect one person, but affect them so much that they make multiple posts posing the same point of view. This in turn can influence other variables and other subsets of the population.

 

The risk assessment is used to identify the most risky subset of the population. In my profession we would test each subset separately as it they are defined differently and have a different risk assessed.

 

Then you would test the accuracy and completeness of the population. This would be very difficult in a public forum and would require a lot of information gathering. Then once you are sure each population is complete and accurate you would pull a statistical sample based on the size of the population and it’s assessed risk. Then each opinion gathered in said sample would have to be tested for accuracy, which in this case would be anything biasing the person into expressing an opinion they don’t actually hold. Almost impossible to prove.

 

Then if you were able to conclude that enough of the sample was accurate you could also conclude that the population it represents also follows the same conclusion. But it would only be for that particular subset of the population.

 

Of course this is just how it works in my profession. Testing it in that manner would be next to impossible. But as Switchfade proved, as a purely mathematical proof there is a simpler way. He would have to fill you in on how to answer your questions in that view point.

[Saikochoro]

Saiko,

 

You are truly a gentleman, and we both are clearly very conversant with mathematical theory and application. Your expertise surely outshines mine.

 

I am humbled to be able to participate in the forums and the game with people such as yourself, and I truly believe this community needs you.

 

I hope we may team in game in the future, and I look forward to learning from your wisdom.

 

There is no fault in your posts, and no conversation exists from a single perspective, thus, if I have somehow caused distress I TOO am responsible, and apologize profusely.

 

Thank you for being part of our community and allowing me to learn from you.

 

Thank you for accepting my apology. I do not think that you have anything to apologize. I leapt to conclusions and was very rude. That did deserve an apology. Thank you for your kind words. I am glad to have someone like you in our community. I currently only play on indomitable. Global handle @Saiko. Feel free to add me as a friend if you have any on that server. I’m in AK time zone though so probably only on really late compared to most of the world.

[SwitchFade]

I have a question then to pose to both you and Saiko concerning this discussion

 

Hi, apologies, I snipped out most of the response quote to remove the growing thread response monster.

 

Without regurgitating an actual multi factor anova, with r square, p values, z value, standard error, upper and lower limits, mean squares and all that crap, I'll summarize as best I can, and refer you to excel and data sets to run a sample analysis...

 

The population can be considered known. The sample size is the participants in the thread.

 

The argument that forum participants are not representative of the whole population can be accurate is the sample size is under 30. As sample size grows over 30, this is a non issue.

 

Sample bias does exist, accounted for by confidence level and margin for error.

 

As confidence level rises, standard deviation changes and upper and lower limits widen.

 

Margin of error is calculated by 'critical value x S.E. of the statistic'

 

There aren't subsets of subsets, as a sample is a portion of a population, here represented by the thread. Why?

 

Because in the thread, data points are those against the change and everyone else, because the change has already been made. If it had not been made, it would be those in favor and everyone else. This is because of the postule of the "null hypothesis"

 

This all indicates an analysis with two variables, at most three. An analysis run with either shows how one or two variables affect another, independent and dependent.

 

This goes beyond a simple poll of data, in a vacuum. Why? A set data poll within itself is a population, not a sample. A sample is used to model a population and must be run through statistical formulae.

 

A regression analysis will model that data set, accounting for error and confidence level over a population size, or without, as we can leverage central limit, Cheb's theorem and the rule of large numbers.

 

In essence, the thread has enough data to accurately model the population with a confidence level of 95% and a margin of error of +/- 3%, supporting that the population in majority (51% or greater) is in favor of the change.

[SwitchFade]

Also, I forgot to mention the two outcomes the analysis renders:

 

Did the change cause unhappy.

 

So...

 

Change=unhappy

 

Or

 

Change=absence of unhappy

 

Notice that absence of unhappy and happy are not the same, again, see null hypothesis

 

Edit: addition

 

If we were to run an analysis the other way before the change, the variables would be slightly different, becoming those harmed by the change, vs those not harmed. In this case we would see that the percentage supporting the change (here seen as 'not harmed') are still the majority.

 

This is because we are looking at a sample after the change, where the sample is the thread. Before the change there was no thread, there for, we would have to generate a sample. In such a sample we would then measure NOT who uses dfb, but who is made unhappy vs absence of unhappy. By gathering a new sample, and running an analysis, we would see responses that indicate "dooooooom" vs "eh + yaaaaay."

[Saikochoro]

I have a question then to pose to both you and Saiko concerning this discussion

 

Hi, apologies, I snipped out most of the response quote to remove the growing thread response monster.

 

Without regurgitating an actual multi factor anova, with r square, p values, z value, standard error, upper and lower limits, mean squares and all that crap, I'll summarize as best I can, and refer you to excel and data sets to run a sample analysis...

 

The population can be considered known. The sample size is the participants in the thread.

 

The argument that forum participants are not representative of the whole population can be accurate is the sample size is under 30. As sample size grows over 30, this is a non issue.

 

Sample bias does exist, accounted for by confidence level and margin for error.

 

As confidence level rises, standard deviation changes and upper and lower limits widen.

 

Margin of error is calculated by 'critical value x S.E. of the statistic'

 

There aren't subsets of subsets, as a sample is a portion of a population, here represented by the thread. Why?

 

Because in the thread, data points are those against the change and everyone else, because the change has already been made. If it had not been made, it would be those in favor and everyone else. This is because of the postule of the "null hypothesis"

 

This all indicates an analysis with two variables, at most three. An analysis run with either shows how one or two variables affect another, independent and dependent.

 

This goes beyond a simple poll of data, in a vacuum. Why? A set data poll within itself is a population, not a sample. A sample is used to model a population and must be run through statistical formulae.

 

A regression analysis will model that data set, accounting for error and confidence level over a population size, or without, as we can leverage central limit, Cheb's theorem and the rule of large numbers.

 

In essence, the thread has enough data to accurately model the population with a confidence level of 95% and a margin of error of +/- 3%, supporting that the population in majority (51% or greater) is in favor of the change.

 

I much prefer this way. Wish I could have done it this way. Would have prevented all those 80+ hr work weeks. But Im done with that job anyway.  In a much better job now.

[Solarverse]

Snip

 

Did you guys do something to fix the rubberbanding? I have not had an issue with rubberbanding ever since this patch. I honestly do not believe it to be coincidence.

[Abraxus]

Just a thought for those that objected to the DFB changes on the basis of using it to level-up toons quickly to try out different powersets for viability.  I believe that has now been addressed with the addition of the beta server.

[Sanguinesun]

Just a thought for those that objected to the DFB changes on the basis of using it to level-up toons quickly to try out different powersets for viability.  I believe that has now been addressed with the addition of the beta server.

 

That and the irony that people are already slowly returning to using the AE again since many missions there are now better exp at the moment than the DFB runs whilst using 100% boosts.  Its all about exp per hour with little down time.

 

So its rather wince worthily come full circle once more.

[Abraxus]

Proof positive that if XP is really all you're after, there are still ways to do it.  I'm sure there are those who will point out that it isn't the same, which is true.  But, in total it will make the difference between taking less than a day to get to from 1 to 50, and possibly adding a few more hours to the total.  Might seem like light years worth of difference, but in the grand scheme of things, it's really not.

[Sanguinesun]

Proof positive that if XP is really all you're after, there are still ways to do it.  I'm sure there are those who will point out that it isn't the same, which is true.  But, in total it will make the difference between taking less than a day to get to from 1 to 50, and possibly adding a few more hours to the total.  Might seem like light years worth of difference, but in the grand scheme of things, it's really not.

 

The only concern though is that then because the ping pong ball has bounced back over into the ae side of the net, that they'll further pop it etc etc etc.  I get server stability concerns that were had due to the AE map server stuffs but the congregating to the trial was the direct result of that change.  Limiting choices in the short term is what leads to folks being funneled into something else.    Speed of leveling should not be a worry for the team, really outside of server related issues. We're not in the epistemic bubble days of the NC folks needing to figure out how to eek out as much time to play in the game to maintain transactions and subs etc.  Its a different perspective now.  Personally, my only concern is that there is a desire to continue to maintain nc's previous thinking (or a reluctance to let that go for what ever reason).  We're seeing a few of these decisions now and while trying to think of them in the context of the now, it still is compelling to have concerns of the direction that it may continue to go towards.  *shrug*.

[ELECTR0]

Hope the DFB nerf fixes the recent problem of people not forming teams for any other content below 20. The graphical updates look great as well.

[WumpusRat]

The only concern though is that then because the ping pong ball has bounced back over into the ae side of the net, that they'll further pop it etc etc etc.  I get server stability concerns that were had due to the AE map server stuffs but the congregating to the trial was the direct result of that change.  Limiting choices in the short term is what leads to folks being funneled into something else.    Speed of leveling should not be a worry for the team, really outside of server related issues. We're not in the epistemic bubble days of the NC folks needing to figure out how to eek out as much time to play in the game to maintain transactions and subs etc.  Its a different perspective now.  Personally, my only concern is that there is a desire to continue to maintain nc's previous thinking (or a reluctance to let that go for what ever reason).  We're seeing a few of these decisions now and while trying to think of them in the context of the now, it still is compelling to have concerns of the direction that it may continue to go towards.  *shrug*.

 

Speaking of AE, the rewards for running content through there seem to have dropped off substantially as well. I was running some maps prior to the patch, and it would get roughly 220-240 tickets per run. Post-patch, those same maps are giving 150-170 per run. Larger/longer maps are giving the same reduced rewards, or about a 30-40% reduction in what you used to get. I don't know if this was an intentional nerf, or something that just slipped in, but it wasn't documented in the patch notes, so if it's another "undocumented nerf", that's a bit annoying. I'd hope it's simply a bug, or if it's intended, that they'd announce it.