Guild Wars 2

Two Guild Wars 2 players ran a survey of Gem Store users and applied some good ol’ statistical analysis to the results. Here are their findings.

A very good read, very informative. This proves what a lot of us have been saying for months on these forums.

I think I hear a mob coming though, I swear I saw a pitchfork crest over the hill…

This is fantastic work!

Not even 400 people…..this is a survey? Come back when you have a few thousand.

killcannon.2576:

Not even 400 people…..this is a survey? Come back when you have a few thousand.

Ah, hail brother, FEAR THE SWARM.

killcannon.2576:

Not even 400 people…..this is a survey? Come back when you have a few thousand.

Someone hasn’t taken statistics.

Incoming Anet defence force.

Guns and Giblets.9308:

killcannon.2576:

Not even 400 people…..this is a survey? Come back when you have a few thousand.

Someone hasn’t taken statistics.

Someone doesn’t know the player base numbers for GW2 or what a proper sample size is or viewer bias or survey participation bias. If you want to push an agenda, at least back it up with something besides 359 people.

A survey needs to be blind, wide, and mandatory to have any meaningful statistics involved with it. This survey was pointless.

And I’m not an apologist. So keep your labels.

killcannon.2576:

Someone doesn’t know the player base numbers for GW2 or what a proper sample size is or viewer bias or survey participation bias. If you want to push an agenda, at least back it up with something besides 359 people.

Several of these were discussed in the follow up discussion to the piece I linked, so I’m not going to rehash them here.

As for the sample size, you can obtain a good confidence interval/level for 300,000-500,000 (or more) players with a survey of 359.

A survey needs to be blind, wide, and mandatory to have any meaningful statistics involved with it. This survey was pointless.

I don’t think those terms mean what you think they mean…

And I’m not an apologist. So keep your labels.

Never said you were.

Aeonblade.8709:

This proves what a lot of us have been saying for months on these forums.

I wish that it did. Unfortunately it proves nothing.

It would be interesting to see some analysis done by someone who does not demonstrate such a significant degree of bias. I am not sure that I can trust someone to be objective when their result report demonstrates a marked lack of that objectivity.

I am not attempting to defend ANet here. I am pretty unhappy with significant elements of the game and am concerned about its future…

But surveys and analysis of this sort make everyone who has doubts about the company and the game look like we are intent on inventing problems rather than attempting to point out and/or solve those that really do exist.

Guns and Giblets.9308:

killcannon.2576:

Someone doesn’t know the player base numbers for GW2 or what a proper sample size is or viewer bias or survey participation bias. If you want to push an agenda, at least back it up with something besides 359 people.

Several of these were discussed in the follow up discussion to the piece I linked, so I’m not going to rehash them here.

As for the sample size, you can obtain a good confidence interval/level for 300,000-500,000 (or more) players with a survey of 359.

A survey needs to be blind, wide, and mandatory to have any meaningful statistics involved with it. This survey was pointless.

I don’t think those terms mean what you think they mean…

And I’m not an apologist. So keep your labels.

Never said you were.

The terms mean exactly what I think they mean. Please regale us all with what you think they mean. Look at current research on surveys and see what you find. Now look at survey research for games and players.

You know who else relied on small survey populations to come to incorrect conclusions and market something?

Microsoft.

You just can’t use the same methods that have been used for the last twenty years anymore. They simply are no longer accurate.

killcannon.2576:

The terms mean exactly what I think they mean. Please regale us all with what you think they mean. Look at current research on surveys and see what you find. Now look at survey research for games and players.

You know who else relied on small survey populations to come to incorrect conclusions and market something?

Microsoft.

I’m not here to try and demonstrate superiority or something. I’m honestly trying to understand why it is you think this survey is “pointless.”

On that front, I’m not seeing any arguments here for me to interact with. Why is the sample size insufficient given how confidence intervals are usually calculated?

Guns and Giblets.9308:

killcannon.2576:

The terms mean exactly what I think they mean. Please regale us all with what you think they mean. Look at current research on surveys and see what you find. Now look at survey research for games and players.

You know who else relied on small survey populations to come to incorrect conclusions and market something?

Microsoft.

I’m not here to try and demonstrate superiority or something. I’m honestly trying to understand why it is you think this survey is “pointless.”

On that front, I’m not seeing any arguments here for me to interact with. Why is the sample size insufficient given how confidence intervals are usually calculated?

First off, you need a population. You don’t have a population. You have a subsection of a subsection of a population. The survey wasn’t taken in game, it was taken on a forum, and by a subsection of a forum at that. So to get an accurate population you have to figure out “What types of players use forums?”, “Out of the players that use forums, how many players from GW2 use forums?”, “How many GW2 players that use forums use Reddit and GW2 Guru?”, “How long was the survey up?” “Who gave the survey?”, “Do they have an agenda?” “Do they have a following?” and on and on and on.

This is selection bias at it’s finest. I could go on with how surveys no longer show accurate representations of populations anymore because the population has changed and that they draw a certain crowd when they are not mandatory. There just isn’t enough information.

Statistics based upon polls just don’t have the same veracity they used to.

100% of the people polled said Killcannon is a god among men and a gift to womankind on earth. I got the statistics to prove it.

I don’t see anything in the survey that’s particularly surprising. Lots of people buy gems, some don’t. Enough people in the high end of gem purchases, if they continue to buy gems, can support the game. Probably the people spending “big money” in the gem store do it for RNG or to convert to gold.

No surprises at all here, for me anyway.

I took a quick look at the link and didn’t really see anything that is really informative for us as players. ANet will have much more accurate data regarding gem sales and what they are spent on, so it’s really of no use to them either.

As killcannon said, this survey can’t really be taken as accurate due to selection bias and we also have no idea if the demographic breakdown is a reasonable facsimile of the game’s population. I didn’t look for the original survey, but it is also worth noting that the precise wording of the survey questions can skew the results (though I’m not saying this is the case here)

Permanent Instant Repair Canister

I’ll sell all my minis into a lifetime of enslavement for one.

killcannon.2576:

Not even 400 people…..this is a survey? Come back when you have a few thousand.

You’re actually kind of right for this poll but wrong in that 359 people can be a statistically relevant population if the sample is selected well enough. In this particular case it wouldn’t have been so the poll is not really relevant.

Emperor.1360:

What a biased “study”, obviously pushing an agenda. Ridiculous.

All it boils down to is basically this: “I want stuff from the Gem Store, but I’m too cheap/poor/uptight to buy it. Therefore I’ll whine about it and try to morally justify my pathetic whining so I don’t sound like a moron and might even get some sycophantic appreciation from my equally worthless peers.”

I`m glad you and many others feel this way.

GW2 is a perfect game and has a perfect gem store

How long till anet locks the thread? i give it a day.

Guns and Giblets.9308:

As for the sample size, you can obtain a good confidence interval/level for 300,000-500,000 (or more) players with a survey of 359.

Yes, true. But not if the participants self select themselves into a survey, which is the case here. Tons of articles out there on self selection bias.

This bit right here:
“48,9% of the subjects between ages “16-24” bought a number “Bellow 800” Gems.”
confuses me and makes me distrust the survey even ignoring all its other faults. The minimum amount of gems you can buy is 800 at a time. So basically what this says is 48.9% of those surveyed who said they bought gems either lied or forgot.

From the data I pull following conclusions:

People with money spend money.
Because clearly people aged 24 and up buy the majority of gems.
Is that surprising? No. You didn’t analyze the gem store, but rather income levels vs age, that’s the underlying variable. At age 16 you correctly assumed these people live from pocket money, and don’t spend on gems. At age 20 these people have college to fund. At age 30 people have a job. Given age spreads, nothing else is proven. That part of the survey literally has nothing to do with the gem store. If you had made the same survey but based it around income levels, you’d have the exact same result.

Quality over quantity
I’m extremely happy with that conclusion, and I hope A.net will make it too. The Horn, permanent tools and services go in that direction. Boxes have to go.

Over half of a subscription
I do feel we get a good return for that money. In WoW, I never progressed beyond the second raid. As such, most WoW patches were void of any content for me. In GW2 I can actually play the new content. That conclusion doesn’t hold water for me.

Quality of life stuff
We knew this would happen, and I’m 100% content with it. Feels to me like I get something in return. Merely paying a subscription is a complete waste of money, especially when the so-called content is not made for me but only for the hardcore raiders. I never got a content update in WoW. Could you imagine Molten Facility only being available after beating FotM48? That’s what happened in WoW.

Lies, kitten ed lies, Statistics
Once again you prove that statistics can be bent to prove literally anything. With the same data, I come to an entirely different conclusion.

I hate statistics.

At least the fanart looked nice. Kinda…

Kailden.5943:

I hate statistics.

At least the fanart looked nice. Kinda…

I agreed with the fanart. Kinda…

His statistics prove two things:
- people with money spend money
- casual people play casual marketed game

kitten