Character Slots Purchase (Ideal price $3.88 per slot) link to analysis included.

Faugaun · May 2015

This is the fourth and final thread I will be making on the purchase of Character Slots. It began with:

http://forums.elderscrollsonline.com/en/discussion/169379/zos-why-can-i-not-buy-character-slots#latest

Which spawned the creation of two surveys:

http://forums.elderscrollsonline.com/en/discussion/169577/if-zos-sold-additional-character-slots-how-many-would-you-buy/p1

and

http://forums.elderscrollsonline.com/en/discussion/169573/if-zos-sold-additional-character-slots-what-is-the-most-you-would-pay-per-slot/p1

These survey data were analyzed and compared. When looked at together the ideal price for the population represented by the voters indicated that if ZoS were to sell character slots then they should sell them for $3.88 (100 crowns is $0.80 - $1.00 currently depending on which pack you purchase so at a high end 388 crowns and at a low end 312 crowns...maybe 350 is good?). At this price they could expect to earn an average of $4.80 per active player (this includes those players who would purchase 0 slots). Depending on how many active subscribers there are and how expensive it would cost to develop the system to allow character slot purchases this could result in a very popular and profitable project for ZoS.

A full explanation of the analysis is presented at the following link: http://asolutionaday.com/business-solutions/selling-character-slots-elder-scrolls-online/

Edit: Here's a linked image showing a hypothetical 100k active player based sales representation:

Nestor · May 2015

Sounds like a fair price. Which means we will probably never get to pay it. This is less than the cost of a cup of coffee. The bean counters will never let this happen this cheap.

See how I worked in bean counters and coffee in one argument?

Heruthema · May 2015

Why should I pay for something that should be free? Why is everyone so willing and ready to pay. I already pay 15 a month and for that I should be getting extra slots, priority queue, etc. Instead you are proposing I pay more for something I should already have.

Gidorick · May 2015

That's a lot of work for a pretty simple request. I don't think adding fluff to the request makes it more conniving. They know we want to be able to buy extra chatacter slots.

The price is a bit low though. I would say more in the 1000 Crown range. What you're purposing is less than the price of a costume.

Faugaun · May 2015

Heruthema wrote: »

Why should I pay for something that should be free? Why is everyone so willing and ready to pay. I already pay 15 a month and for that I should be getting extra slots, priority queue, etc. Instead you are proposing I pay more for something I should already have.

@Heruthema If this were a subscription only game then I might agree with you ...but it is no longer a subscription only game and thus not everyone pays a monthly fee. With your subscription you get 1500 crowns a month and can certainly purchase these (if you choose) for that benefit without any additional cost. At the end of the day if they cannot make money then we will not have a game to play and if they cannot implement a feature profitably then that feature will not come. Zeni is a for profit company with private investors who expect a return on their investment. To think that a millionaire and billionaire business moguls who are private investors will just give something for free is naive. That said they are more than happy to take risks knowing that somethings will not be successful but to propose an idea with zero return on investment...in other words free? Yeah...you're not gonna sell it to the financiers.

Nestor wrote: »

Sounds like a fair price. Which means we will probably never get to pay it. This is less than the cost of a cup of coffee. The bean counters will never let this happen this cheap.

See how I worked in bean counters and coffee in one argument?

@Nestor Haha...Mcdonalds has $1 coffee... btw...if the bean counters reject the most profitable pricing then perhaps that explains some things....

nastuug · May 2015

Heruthema wrote: »

Why should I pay for something that should be free? Why is everyone so willing and ready to pay. I already pay 15 a month and for that I should be getting extra slots, priority queue, etc. Instead you are proposing I pay more for something I should already have.

And what would you propose happens to the extra slots if you were to fail to pay the subscription one or more months? Freeze said characters and their inventory/enlightenment gains? Don't think you'll see that happen.

Faugaun · May 2015

Gidorick wrote: »

That's a lot of work for a pretty simple request. I don't think adding fluff to the request makes it more conniving. They know we want to be able to buy extra chatacter slots.

The price is a bit low though. I would say more in the 1000 Crown range. What you're purposing is less than the price of a costume.

@Gidorick The most amazing castle I visited in Ireland was owned by the government. It was donated to them from the former owner when he died. It's grounds covered an entire mountain, had hunting woods, a large garden and a full-time staff to maintain everything and keep it in order for the owner. The owner used this castle as a vacation house, he was an American business mogul and he flew his buddies to his castle about 4 weeks a year to visit, party, and have fun.

This business mogul made his money as one of three founders of the Tobasco company.

The point here is that in any economic system it is not necessarily the sale price that dictates the amount of money you make. Sell a lot of anything and then even low profit margins make sense. If you examine the analysis you will notice that they actually make less money as the price increases beyond $3.88. What then is the point of charging 1000 crowns when you then reduce the number of buyers to a fraction of how many you would otherwise have? I see no compelling reason for them to limit their profits by overpricing a product...especially when once that product is developed it has a nominal cost to create additional units.

Faugaun · May 2015

@Gidorick I added this in the analysis to help clarify with a visual representation. Will also add it to the original post.

nastuug · May 2015

Faugaun wrote: »

@Gidorick I added this in the analysis to help clarify with a visual representation. Will also add it to the original post.

Unfortunately, this was done with a very, very, very small sample size.

edit: need more 'very'

VictoriaRachel · May 2015

Ah, a graph.

One that is based on an unrepresentative incredibly small sample using a biased poll.

Must be 'analysis'.

*sigh*

Also if those people are all subscribers there is no additional profit at all as all extra slots can be bought with the crowns they are already receiving. You want to encourage people to spend more then they already do not just hand it out like candy.

Gidorick · May 2015

nastuug wrote: »

Faugaun wrote: »

@Gidorick I added this in the analysis to help clarify with a visual representation. Will also add it to the original post.

Unfortunately, this was done with a very, very, very small sample size.

edit: need more 'very'

This was my thought as well.

There is also the consideration that additional slots take SOME space on servers. So there is regular costs for ZOS there and also price can impact perceived value and if they provide it too cheaply people will buy without really even wanting it.

I commend your effort @Faugaun but it seems like a rube goldberg machine of information.

You could have just said "Id Pay X crowns for an extra chatacter slot!" and we would have had a conversation on that.

Faugaun · May 2015

...

VictoriaRachel wrote: »

Ah, a graph.

One that is based on an unrepresentative incredibly small sample using a biased poll.

Must be 'analysis'.

*sigh*

Also if those people are all subscribers there is no additional profit at all as all extra slots can be bought with the crowns they are already receiving. You want to encourage people to spend more then they already do not just hand it out like candy.

How are the polls biased? If you have a better poll then by all means let's see it. As for sample size in survey data, I'm working on locating a method for determining confidence and p-value since that is a statistical method I've never explored I don't have an answer I am satisfied with yet. obviously as with any sampling the larger your sample the more accurately it represents the population.

So far I have located a table that indicates a sample size of 110 is adequate to represent a population of 1,000,000 with 90% confidence. If you agree with the chart then a sample of 102 and a sample of 164 may in fact be adequate sample sizes.

The chart was not sourced and frankly I'm not satisfied with the numbers from the chart (Which may be perfectly accurate). I'm still looking for the actual formula which would determine sample size in this instance (without knowing the population size).

For now I will say, if you can get a better survey with a larger sample then by all means let's have it. In the mean time I will continue researching this aspect until I can say further. In any case they are descriptive in nature at the very least and as such they still provide useful information.

Nestor · May 2015

Gidorick wrote: »

You could have just said "Id Pay X crowns for an extra chatacter slot!" and we would have had a conversation on that.

He did, read the OP again, the threads are linked there.

Thymos · May 2015

Faugaun wrote: »

...

VictoriaRachel wrote: »

Ah, a graph.

One that is based on an unrepresentative incredibly small sample using a biased poll.

Must be 'analysis'.

*sigh*

Also if those people are all subscribers there is no additional profit at all as all extra slots can be bought with the crowns they are already receiving. You want to encourage people to spend more then they already do not just hand it out like candy.

How are the polls biased? If you have a better poll then by all means let's see it. As for sample size in survey data, I'm working on locating a method for determining confidence and p-value since that is a statistical method I've never explored I don't have an answer I am satisfied with yet. obviously as with any sampling the larger your sample the more accurately it represents the population.

So far I have located a table that indicates a sample size of 110 is adequate to represent a population of 1,000,000 with 90% confidence. If you agree with the chart then a sample of 102 and a sample of 164 may in fact be adequate sample sizes.

The chart was not sourced and frankly I'm not satisfied with the numbers from the chart (Which may be perfectly accurate). I'm still looking for the actual formula which would determine sample size in this instance (without knowing the population size).

For now I will say, if you can get a better survey with a larger sample then by all means let's have it. In the mean time I will continue researching this aspect until I can say further. In any case they are descriptive in nature at the very least and as such they still provide useful information.

The polls are biased, because there is basically 1 option for no, and several for yes. You need a very complex formula to do the analysis of polls that have so many choices to come up with the overall results of the range of what could be earned by ever active player. You also need more control of your sample if you have 2 samples that are directly related. Like, are you sure that everyone that voted in one poll, voted in the second poll. You're creating more variables when you don't have that control.

You also ignore the fact that a subscriber would get 1,500 crowns each month they have subscribed. That means they have the option to buy 3 character slots in a month with their stipend. Which does not net ZOS anymore than what they are already selling the subscriber.

90% confidence has a large gap for error also, you should be aiming for 95% if not 99%. You should also be seeing that the ideal price and the average earning per player would be of a range, not just the middle. Averages are often something taken from known numbers, probability is what you get from unknown numbers.

If you wanted to get a decent look of what 100,000 people wanted to know with a 95% confidence level, with a 5% margin of error, you would need to poll 383 people.

Naivefanboi · May 2015

Well if come console launch we have only 3 slots and the rrst in crown store, we know who to thank. xD
zos needs more money! Just look at how content has picked up since the crown store! Yes thats sarcasim.

nastuug · May 2015

Naivefanboi wrote: »

Well if come console launch we have only 3 slots and the rrst in crown store, we know who to thank. xD

LOL

Faugaun · May 2015

Thymos wrote: »

Faugaun wrote: »

...

VictoriaRachel wrote: »

Ah, a graph.

One that is based on an unrepresentative incredibly small sample using a biased poll.

Must be 'analysis'.

*sigh*

Also if those people are all subscribers there is no additional profit at all as all extra slots can be bought with the crowns they are already receiving. You want to encourage people to spend more then they already do not just hand it out like candy.

How are the polls biased? If you have a better poll then by all means let's see it. As for sample size in survey data, I'm working on locating a method for determining confidence and p-value since that is a statistical method I've never explored I don't have an answer I am satisfied with yet. obviously as with any sampling the larger your sample the more accurately it represents the population.

So far I have located a table that indicates a sample size of 110 is adequate to represent a population of 1,000,000 with 90% confidence. If you agree with the chart then a sample of 102 and a sample of 164 may in fact be adequate sample sizes.

The chart was not sourced and frankly I'm not satisfied with the numbers from the chart (Which may be perfectly accurate). I'm still looking for the actual formula which would determine sample size in this instance (without knowing the population size).

For now I will say, if you can get a better survey with a larger sample then by all means let's have it. In the mean time I will continue researching this aspect until I can say further. In any case they are descriptive in nature at the very least and as such they still provide useful information.

The polls are biased, because there is basically 1 option for no, and several for yes. You need a very complex formula to do the analysis of polls that have so many choices to come up with the overall results of the range of what could be earned by ever active player. You also need more control of your sample if you have 2 samples that are directly related. Like, are you sure that everyone that voted in one poll, voted in the second poll. You're creating more variables when you don't have that control.

You also ignore the fact that a subscriber would get 1,500 crowns each month they have subscribed. That means they have the option to buy 3 character slots in a month with their stipend. Which does not net ZOS anymore than what they are already selling the subscriber.

90% confidence has a large gap for error also, you should be aiming for 95% if not 99%. You should also be seeing that the ideal price and the average earning per player would be of a range, not just the middle. Averages are often something taken from known numbers, probability is what you get from unknown numbers.

If you wanted to get a decent look of what 100,000 people wanted to know with a 95% confidence level, with a 5% margin of error, you would need to poll 383 people.

I won't bother going line for line here but there are no yes and know answers within the surveys. They all represent positive values within a numeric range. Within the selections are valid selections for all possible positive numeric values, even ridiculously high values. Reference subscribers, that is a valid point (kind of) subscribers do get crowns and they may spend the crowns from the subscription on slots. However they may already be purchasing other things with their crowns and some players may get a subscription simply because they will buy slots. The fact remains though that no matter how the players got the crowns providing an additional purchase will result in more crowns being spent. Every character slot sold is actually an additional sale that would not have occurred had there not been the option to purchase character slots.

90% is a pretty big gap, in my research field (which doesn't use surveys) we shoot for 95% confidence. Biomedical research shoots for 99.99% confidence. However, there are other scientific (and non-scientific) research fields which accept 90% confidence as significant. These topics are published in respectable peer reviewed literature with only 90% confidenc. Look at the appendix in any statistics book they all list values to 90% and lower in some. The reason is because this value can provide useful information and researchers and staticians know that 90% is 90% and it means exactly that. Even a lay person understands when someone says that the represented results have a 90% chance of being statistically significant.

Reference your 95% confidence numbers instead of hiding the information and using a lower count than what I used (1,000,000) be more productive and provide the numbers for the same 1,000,000...it's only 4xx in the lower 400's and for conservative analysis whatever the benchmark used should be above the actual player count to have at least the desired level of confidence. It really sounds like you're good at Google and undergrad statistics but need more education .

Finally as always if you have a better data set and a bigger sample lay it on us. I put out what I have and detailed the analysis and the data source I'm not saying it's perfect. People can look at what I did and make their own interpretations.

It's easy to critique others but until you or someone else produces something better then this is still the best information we have if you don't agree go make your own surveys and post your results.

BBSooner · May 2015

Gidorick wrote: »

That's a lot of work for a pretty simple request. I don't think adding fluff to the request makes it more conniving. They know we want to be able to buy extra chatacter slots.

The price is a bit low though. I would say more in the 1000 Crown range. What you're purposing is less than the price of a costume.

This.

If they added the feature it would very likely just get stamped with either 500 crowns or 1000 crowns. Either amount would be alright, imo.

nastuug · May 2015

Faugaun wrote: »

Thymos wrote: »

Faugaun wrote: »

...

VictoriaRachel wrote: »

Ah, a graph.

One that is based on an unrepresentative incredibly small sample using a biased poll.

Must be 'analysis'.

*sigh*

Also if those people are all subscribers there is no additional profit at all as all extra slots can be bought with the crowns they are already receiving. You want to encourage people to spend more then they already do not just hand it out like candy.

How are the polls biased? If you have a better poll then by all means let's see it. As for sample size in survey data, I'm working on locating a method for determining confidence and p-value since that is a statistical method I've never explored I don't have an answer I am satisfied with yet. obviously as with any sampling the larger your sample the more accurately it represents the population.

So far I have located a table that indicates a sample size of 110 is adequate to represent a population of 1,000,000 with 90% confidence. If you agree with the chart then a sample of 102 and a sample of 164 may in fact be adequate sample sizes.

The chart was not sourced and frankly I'm not satisfied with the numbers from the chart (Which may be perfectly accurate). I'm still looking for the actual formula which would determine sample size in this instance (without knowing the population size).

For now I will say, if you can get a better survey with a larger sample then by all means let's have it. In the mean time I will continue researching this aspect until I can say further. In any case they are descriptive in nature at the very least and as such they still provide useful information.

The polls are biased, because there is basically 1 option for no, and several for yes. You need a very complex formula to do the analysis of polls that have so many choices to come up with the overall results of the range of what could be earned by ever active player. You also need more control of your sample if you have 2 samples that are directly related. Like, are you sure that everyone that voted in one poll, voted in the second poll. You're creating more variables when you don't have that control.

You also ignore the fact that a subscriber would get 1,500 crowns each month they have subscribed. That means they have the option to buy 3 character slots in a month with their stipend. Which does not net ZOS anymore than what they are already selling the subscriber.

90% confidence has a large gap for error also, you should be aiming for 95% if not 99%. You should also be seeing that the ideal price and the average earning per player would be of a range, not just the middle. Averages are often something taken from known numbers, probability is what you get from unknown numbers.

If you wanted to get a decent look of what 100,000 people wanted to know with a 95% confidence level, with a 5% margin of error, you would need to poll 383 people.

I won't bother going line for line here but there are no yes and know answers within the surveys. They all represent positive values within a numeric range. Within the selections are valid selections for all possible positive numeric values, even ridiculously high values. Reference subscribers, that is a valid point (kind of) subscribers do get crowns and they may spend the crowns from the subscription on slots. However they may already be purchasing other things with their crowns and some players may get a subscription simply because they will buy slots. The fact remains though that no matter how the players got the crowns providing an additional purchase will result in more crowns being spent. Every character slot sold is actually an additional sale that would not have occurred had there not been the option to purchase character slots.

90% is a pretty big gap, in my research field (which doesn't use surveys) we shoot for 95% confidence. Biomedical research shoots for 99.99% confidence. However, there are other scientific (and non-scientific) research fields which accept 90% confidence as significant. These topics are published in respectable peer reviewed literature with only 90% confidenc. Look at the appendix in any statistics book they all list values to 90% and lower in some. The reason is because this value can provide useful information and researchers and staticians know that 90% is 90% and it means exactly that. Even a lay person understands when someone says that the represented results have a 90% chance of being statistically significant.

Reference your 95% confidence numbers instead of hiding the information and using a lower count than what I used (1,000,000) be more productive and provide the numbers for the same 1,000,000...it's only 4xx in the lower 400's and for conservative analysis whatever the benchmark used should be above the actual player count to have at least the desired level of confidence. It really sounds like you're good at Google and undergrad statistics but need more education .

Finally as always if you have a better data set and a bigger sample lay it on us. I put out what I have and detailed the analysis and the data source I'm not saying it's perfect. People can look at what I did and make their own interpretations.

It's easy to critique others but until you or someone else produces something better then this is still the best information we have if you don't agree go make your own surveys and post your results.

It think it's more to the point that you're polling of 172 unique forum goers (51% of which said they wouldn't buy any extra slots) is skewed since this is just a drop in the bucket of the player base. It's not a fair representation of the mass's opinion. Hell, it's not even a fair representation of the majority of forum goers.

As @Gidorick stated, I commend you for your efforts, but these polls never paint an accurate picture.

Faugaun · May 2015

nastuug wrote: »

Faugaun wrote: »

Thymos wrote: »

Faugaun wrote: »

...

VictoriaRachel wrote: »

Ah, a graph.

One that is based on an unrepresentative incredibly small sample using a biased poll.

Must be 'analysis'.

*sigh*

Also if those people are all subscribers there is no additional profit at all as all extra slots can be bought with the crowns they are already receiving. You want to encourage people to spend more then they already do not just hand it out like candy.

How are the polls biased? If you have a better poll then by all means let's see it. As for sample size in survey data, I'm working on locating a method for determining confidence and p-value since that is a statistical method I've never explored I don't have an answer I am satisfied with yet. obviously as with any sampling the larger your sample the more accurately it represents the population.

So far I have located a table that indicates a sample size of 110 is adequate to represent a population of 1,000,000 with 90% confidence. If you agree with the chart then a sample of 102 and a sample of 164 may in fact be adequate sample sizes.

The chart was not sourced and frankly I'm not satisfied with the numbers from the chart (Which may be perfectly accurate). I'm still looking for the actual formula which would determine sample size in this instance (without knowing the population size).

For now I will say, if you can get a better survey with a larger sample then by all means let's have it. In the mean time I will continue researching this aspect until I can say further. In any case they are descriptive in nature at the very least and as such they still provide useful information.

The polls are biased, because there is basically 1 option for no, and several for yes. You need a very complex formula to do the analysis of polls that have so many choices to come up with the overall results of the range of what could be earned by ever active player. You also need more control of your sample if you have 2 samples that are directly related. Like, are you sure that everyone that voted in one poll, voted in the second poll. You're creating more variables when you don't have that control.

You also ignore the fact that a subscriber would get 1,500 crowns each month they have subscribed. That means they have the option to buy 3 character slots in a month with their stipend. Which does not net ZOS anymore than what they are already selling the subscriber.

90% confidence has a large gap for error also, you should be aiming for 95% if not 99%. You should also be seeing that the ideal price and the average earning per player would be of a range, not just the middle. Averages are often something taken from known numbers, probability is what you get from unknown numbers.

If you wanted to get a decent look of what 100,000 people wanted to know with a 95% confidence level, with a 5% margin of error, you would need to poll 383 people.

I won't bother going line for line here but there are no yes and know answers within the surveys. They all represent positive values within a numeric range. Within the selections are valid selections for all possible positive numeric values, even ridiculously high values. Reference subscribers, that is a valid point (kind of) subscribers do get crowns and they may spend the crowns from the subscription on slots. However they may already be purchasing other things with their crowns and some players may get a subscription simply because they will buy slots. The fact remains though that no matter how the players got the crowns providing an additional purchase will result in more crowns being spent. Every character slot sold is actually an additional sale that would not have occurred had there not been the option to purchase character slots.

90% is a pretty big gap, in my research field (which doesn't use surveys) we shoot for 95% confidence. Biomedical research shoots for 99.99% confidence. However, there are other scientific (and non-scientific) research fields which accept 90% confidence as significant. These topics are published in respectable peer reviewed literature with only 90% confidenc. Look at the appendix in any statistics book they all list values to 90% and lower in some. The reason is because this value can provide useful information and researchers and staticians know that 90% is 90% and it means exactly that. Even a lay person understands when someone says that the represented results have a 90% chance of being statistically significant.

Reference your 95% confidence numbers instead of hiding the information and using a lower count than what I used (1,000,000) be more productive and provide the numbers for the same 1,000,000...it's only 4xx in the lower 400's and for conservative analysis whatever the benchmark used should be above the actual player count to have at least the desired level of confidence. It really sounds like you're good at Google and undergrad statistics but need more education .

Finally as always if you have a better data set and a bigger sample lay it on us. I put out what I have and detailed the analysis and the data source I'm not saying it's perfect. People can look at what I did and make their own interpretations.

It's easy to critique others but until you or someone else produces something better then this is still the best information we have if you don't agree go make your own surveys and post your results.

It think it's more to the point that you're polling of 172 unique forum goers (51% of which said they wouldn't buy any extra slots) is skewed since this is just a drop in the bucket of the player base. It's not a fair representation of the mass's opinion. Hell, it's not even a fair representation of the majority of forum goers.

As @Gidorick stated, I commend you for your efforts, but these polls never paint an accurate picture.

That I will agree with it is only representative of the population that the sample represents (Which I have stated multiple times). How similar the actual population and the population that the samples describe is a completely different debate. My guess is that there are many differences in forum goers and non forum goers. Is payment amount and quantity reference character slots? We don't know. We could find out if we asked a random sample of non-forum goers and then use anova to compare the groups to answer that. I don't have anyway to collect that sample...as ZoS is the only one with access to the means to even ask the question. However, if ZoS asks the question then the responses will undoubtedly be lower than the truth for obvious reasons.

That said this is at current the best sample avaliable if the sample size grows enough I will be happy to recompute the numbers for better accuracy. That will not address the difference in the sample population and the actual population. But, at current it would be fair to say that there is a 90% chance that this reflects the population of forum visitors that look and vote at polls in the general tab on Wednesdays. How that small subset models the greater population of tamriel is a whole bunch of data collection on many differenr factors and several books worth of writing. Probably a good project for a thesis in an appropriate field though.

Gidorick · May 2015

You must really love statistics, eh @Faugaun?

Faugaun · May 2015

Gidorick wrote: »

You must really love statistics, eh @Faugaun?

hehe ... I get bored...and statistics and games are my hobby

not many people that derive equations when they are bored...somethings wrong with me i'm convinced

Audigy · May 2015

Heruthema wrote: »

Why should I pay for something that should be free? Why is everyone so willing and ready to pay.

Well said.

Additional slots are a P2W mechanism and have no place in a free MMO. As soon the crown shop offers important items or content that a player can not gain by playing, ESO will follow the evil road of milking its customers for things that should be free.

Say NO to additional slots in the crown shop!

Faugaun · May 2015

Audigy wrote: »

Heruthema wrote: »

Why should I pay for something that should be free? Why is everyone so willing and ready to pay.

Well said.

Additional slots are a P2W mechanism and have no place in a free MMO. As soon the crown shop offers important items or content that a player can not gain by playing, ESO will follow the evil road of milking its customers for things that should be free.

Say NO to additional slots in the crown shop!

@Audigy What would you then propose? Also, is buying a second account p2w? Cause I can do that all day long, third account, fourth account..

Gidorick · May 2015

I 100% disagree that additional slots are pay to win. I actually can't even fathom why this would even be considered p2w. It's a system function like a name change or server change. It has absolutely no impact on game-play whatsoever.

Faugaun · May 2015

Gidorick wrote: »

I 100% disagree that additional slots are pay to win. I actually can't even fathom why this would even be considered p2w. It's a system function like a name change or server change. It has absolutely no impact on game-play whatsoever.

The best attempt I saw was the argument that more characters equals more hirelings, more bank space, more inventory, more guild slots are more trader guilds and you have more characters with your account wide cp.

My response was that its actually P2L because they spend more time levelling more toons and thus get less cp for time played. Regarding the guilds...well we are talking character slots...though I would like guild slots also for rp reasons, more bank space...whoopdeedo I don't use the bank space I currently have....more hirelings, someone else pointed out 5-7 min of node hopping in crag is about equivalent...so yeah I don't see it either...

VictoriaRachel · May 2015

Faugaun wrote: »

How are the polls biased? If you have a better poll then by all means let's see it. As for sample size in survey data, I'm working on locating a method for determining confidence and p-value since that is a statistical method I've never explored I don't have an answer I am satisfied with yet. obviously as with any sampling the larger your sample the more accurately it represents the population.

To make the poll less biased in future limit it to figures and not opinion. The "I would pay more than $6.25 because I want to support the game and money isn't that important." is CLEARLY your own opinion on paying more than that. It is leading, it is encouraging other people not to vote for it as they do think money is important but it is something they would still pay for it due to other reasons that you are eliminating.

You are sampling a small number of a small sample. You are not sampling a representative cross section of the community, you are only looking at those that use the forum. You are then only sampling a really small number of those users. Basically it not enough to be considered meaningful.

Faugaun wrote: »

So far I have located a table that indicates a sample size of 110 is adequate to represent a population of 1,000,000 with 90% confidence. If you agree with the chart then a sample of 102 and a sample of 164 may in fact be adequate sample sizes.

The chart was not sourced and frankly I'm not satisfied with the numbers from the chart (Which may be perfectly accurate). I'm still looking for the actual formula which would determine sample size in this instance (without knowing the population size).

If it does not have a source I can not see it being worth the bytes needed to show it. We have no idea what that table is talking about, we do not know if it is a reliable source, we do not know what caveats that table comes with. Therefore it is not useful information for us and lets be honest sounds horrifically suspect. Look at the opinion polls in the UK Elections that even far larger polls do not actually give accurate results.

Faugaun wrote: »

For now I will say, if you can get a better survey with a larger sample then by all means let's have it. In the mean time I will continue researching this aspect until I can say further. In any case they are descriptive in nature at the very least and as such they still provide useful information.

I am happy for you to continue to research it is just presenting it as meaningful analysis to which I object.

Faugaun · May 2015

VictoriaRachel wrote: »

Faugaun wrote: »

How are the polls biased? If you have a better poll then by all means let's see it. As for sample size in survey data, I'm working on locating a method for determining confidence and p-value since that is a statistical method I've never explored I don't have an answer I am satisfied with yet. obviously as with any sampling the larger your sample the more accurately it represents the population.

To make the poll less biased in future limit it to figures and not opinion. The "I would pay more than $6.25 because I want to support the game and money isn't that important." is CLEARLY your own opinion on paying more than that. It is leading, it is encouraging other people not to vote for it as they do think money is important but it is something they would still pay for it due to other reasons that you are eliminating.

You are sampling a small number of a small sample. You are not sampling a representative cross section of the community, you are only looking at those that use the forum. You are then only sampling a really small number of those users. Basically it not enough to be considered meaningful.

Faugaun wrote: »

So far I have located a table that indicates a sample size of 110 is adequate to represent a population of 1,000,000 with 90% confidence. If you agree with the chart then a sample of 102 and a sample of 164 may in fact be adequate sample sizes.

The chart was not sourced and frankly I'm not satisfied with the numbers from the chart (Which may be perfectly accurate). I'm still looking for the actual formula which would determine sample size in this instance (without knowing the population size).

If it does not have a source I can not see it being worth the bytes needed to show it. We have no idea what that table is talking about, we do not know if it is a reliable source, we do not know what caveats that table comes with. Therefore it is not useful information for us and lets be honest sounds horrifically suspect. Look at the opinion polls in the UK Elections that even far larger polls do not actually give accurate results.

Faugaun wrote: »

For now I will say, if you can get a better survey with a larger sample then by all means let's have it. In the mean time I will continue researching this aspect until I can say further. In any case they are descriptive in nature at the very least and as such they still provide useful information.

I am happy for you to continue to research it is just presenting it as meaningful analysis to which I object.

You didn't read the analysis at all where I specifically detail the representitve sample....and I don't see the $6.25 option as leading since that would be the market cap based on the cost of a second account divided by 8. Anything higher is simply donation to a for profit cooperation (perhaps it would have been more defensible worded slightly differently, but I don't think it is creating a bias, obviously without another poll this cannot be proved one way or the other).

I think I expressed similar sentiment regarding that table, and for much the same reasons, I'm not going to present a table that isn't sourced but in the sake of disclosure to what I had uncovered that was it so far.

I think there is a lot more subjectivity to a pole like "what is your favorite fruit" or "politician" than "what price would be the most that you would pay" as such I don't think the UK statement is relevant. The population is much larger, and what i have found so far with surveys is that the larger the underlying population the larger your sample needs to be. Its like comparing apples and oranges, they are both round fruits but once inside they are completely different. In the context of the analysis it is meaningful, again if you read it I clearly articulate that it is relevant to the population that the sample represents, which may be but is not necessarily the same as the entire forum population or the forum population on patch day or the weekend forum population or the whole tamriel population.

Again though if you really wanna refute it please run your own polls and present your findings. I would love to see what results others come up with. Plus with multiple different methods we can pick the best and go from there.

Here's a sampling idea. Get on your VR 14 create a random table consisting of all exploration points in tamriel assign a sequential number to each one. The use a random number generator to select which places and the order you will go. Go to each location and poll the first player you see at each place. Repeat this 500 times and then we will have a real study if we can get 10 people on board then we can define the question and we each would need to poll 50 people. We might also need to generate random times to account for players on different parts of the globe. With careful question design and being able to ask multiple questions of the same person and link the responses. Then we could probably run this as a parametric test since the questions could be continuous positive numbers.

Some ideas for questions:

What is the most you would pay for an extra character slot?
How many slots do you think you might eventually purchase?
How long have you played the game?
What level is your highest character?
How old are you?
Are you male or female?
What country do you live inside?

Due to the increased error introduced from the addition of more questions sample size may need to be increased especially if you wanna go multivariate.

I'm not interested enough to go through that though, I'll just buy a second account and drop my sub if they don't release slots or price them above 500 crowns.

nastuug · May 2015

Audigy wrote: »

Heruthema wrote: »

Why should I pay for something that should be free? Why is everyone so willing and ready to pay.

Well said.

Additional slots are a P2W mechanism and have no place in a free MMO. As soon the crown shop offers important items or content that a player can not gain by playing, ESO will follow the evil road of milking its customers for things that should be free.

Say NO to additional slots in the crown shop!

I wouldn't be so quick as to label this as P2W. Just because it is something that cannot be acquired with in-game currency doesn't mean it's P2W. There's little competitive advantage. These additional character slots mostly appeals to the altoholics/crafters anyways.

gw2only1b14_ESO · May 2015

Id personally had thought about and brought it up before crown store and sure would love extra slots but now that its B2P I can just use a defunct account and make 8 more on it. but if they do then I think they should charge $5 to $7 each for the slots and go up to 16

VictoriaRachel · May 2015

Faugaun wrote: »

You didn't read the analysis at all where I specifically detail the representitve sample....and I don't see the $6.25 option as leading since that would be the market cap based on the cost of a second account divided by 8. Anything higher is simply donation to a for profit cooperation (perhaps it would have been more defensible worded slightly differently, but I don't think it is creating a bias, obviously without another poll this cannot be proved one way or the other) .

I have read everything you have written, I just do not agree with a lot of it. I do not think it is representative of any group.

As for not seeing that final option as leading, I think that is just a case of being blinded by bias. I understand the reason you do not want that option but that feeling as been placed into the poll skewing the result. Making only one option contain opinion rather than simply a price , especially one make claims about the value of money in general is biased and could easily have been avoided by taking a more measured approach. If the final option had simply stated “$6.25 or more” I would agree that it was not biased.

Faugaun wrote: »

I think there is a lot more subjectivity to a pole like "what is your favorite fruit" or "politician" than "what price would be the most that you would pay" as such I don't think the UK statement is relevant. The population is much larger, and what i have found so far with surveys is that the larger the underlying population the larger your sample needs to be. Its like comparing apples and oranges, they are both round fruits but once inside they are completely different. In the context of the analysis it is meaningful, again if you read it I clearly articulate that it is relevant to the population that the sample represents, which may be but is not necessarily the same as the entire forum population or the forum population on patch day or the weekend forum population or the whole tamriel population.

I do not know why you think a “what is your favourite price” poll would be any less subjective. I also think that like the UK election, which is why I referenced it, what people tell pollsters and what their actions will be are hugely different. You have to remember you are asking “What do you want to pay for…?”. People are always going to give lower values then they will spend when it is offered. It is human nature to want things cheap. People do not want to pay a lot for anything, but they will when that is their only option. So a subjective poll like this will never give a real result.

Faugaun wrote: »

Again though if you really wanna refute it please run your own polls and present your findings. I would love to see what results others come up with. Plus with multiple different methods we can pick the best and go from there.

I am not going to run a poll. I do not think polls are very helpful in this situation.

What people want to pay and what people will pay is very different. In my opinion any poll, no matter how representative will give a lower result than the price we will actually see these listed for in the store and people will pay that higher value.