The results are in from the World of Warcraft Gold Survey, and I’d like to thank all that participated, in all 2,531 usable results were collected! The survey has produced interesting results which I hope can provide insight to the in-game economic condition of the player base, and allow gold theorycrafters to better understand the economy and it’s players.
The results of the survey however may be misleading to those who do not fully understand the implications of the data, so I’d like to take the time to walk readers through exactly what this survey did, and the implications of the data.
I make no claim to have professional or even an amateur Economic or Statistics background. This study was done purely for the sake of knowledge and was not done for any organization or intended for any professional study. I make no guarantee to the accuracy of this data. My apologies to any statistics or economics majors that I give an aneurysm as a result of reading this study.
The survey started the evening of Saturday, October 29th and ran until the afternoon of Saturday, November 5th and collected 2,531 usable results. Requests for survey results ran on The Consortium, US and EU Official Forums, MMO-Champion forums, WoW Insider’s Classifieds, and several independent gold blogs. Entries which contained data impossible due to the nature or limits of the game were discarded, though such contamination was limited. The survey was hosted on site using LimeSurvey software.
The goal of the survey was to establish average wealth, wealth distribution, and demographics in the World of Warcraft Player base. This survey and report were inspired by Sinshroud’s of The Consortium post on Can The Consortium Be Considered the “1%” of WoW?.
As with any research the results of this survey was subject to a certain amount of error and the very nature of the survey led to some unique potential sources of error:
- Opt In Survey This survey was simply advertised for and was not a targeted survey, that means someone can be biased towards completing it or not, and those with more gold in this case would be much more likely to complete the survey than those with little. Many such sites were ones dedicated to making gold and will produce higher than average results. This skews the results upwards.
- Limited Distribution Since it was impossible for me to go in-game and ask people individually about their wealth I had to rely on various sites to get my survey out there, meaning players who do not frequently visit World of Warcraft related sites would likely not be exposed to the survey. This skews the results upwards.
- Entry errors This survey was a self-completed survey and is susceptible to data entry errors, fat fingers etc. It is difficult to determine what effect this would have on results, but it is most likely that this would skew data upwards due to added digits.
- False Data Many people might feel compelled to lie and inflate their numbers on a survey such as this, as with any survey centred around a potential ego issue. Several players, based on responses in several forum threads, did not seem to understand that the question was asking for liquid only, and entered their worth value. This would skew results upwards.
- Rounding Errors Rounding conventions were not given for the survey and so subjects were left to their own rounding conventions. I believe these errors could swing either way and will for the most part cancel out.
- Trolls Many entries contained impossible data and had to be excluded, however some other entries may be blatant corruption on a more reasonable scale and did not get caught by my filters. The filters on the questions plus the likely hood of trolls entering too-low data makes it reasonable to assume errors due to this will cancel out
- Site Trouble Twice during the survey, once due to faulty hardware and once due to server load, the survey became unavailable, potentially cutting of some subjects mid-response, leading to a high number of incomplete surveys. A problem with the survey software prevented some players from continuing the survey.
- Perceived Security Issues Several players reported not taking the survey due to worries that the information would cause them to be targeted for hacking and account theft.
These errors would create an inflated result, producing higher than actual result.
Subjects who completed the survey were asked to complete 4 questions, how many years since they’ve started playing World of Warcraft, which activities they enjoy ingame, which out of game resources they frequent, and how much gold they have.
Subjects were asked if they raided, participated in PvP, RP’d, played the auction house, or farmed gold (limited to manual labour type activities). Subjects were asked to answer to these with Yes, No, or Uncertain. Subjects were also asked if they visited various online resources: The Consortium, Just My Two Copper, MMM-Champion, Gold Capped from WoW Insider, or miscellaneous gold blogs. They were given just two answer options for this, Yes or No. None of these questions were mutually exclusive and could answer Yes, No, or Uncertain to any number of questions.
The final question was how much liquid gold the subject had across all characters, realms, and accounts not including the worth of items in their inventory.
Realm or faction specific questions were not asked to help alleviate security concerns and the scope of this survey did not include realm or faction numbers.
With 2,531 results the subjects had between them 408,738,617 gold for an average of 161,493 gold per subject. Anecdotal evidence suggests that this number is way too high, so further data and analysis must be considered when determining how much gold an average player has.
The median wealth was 35,000 gold meaning that 50% of the subjects had 35,000 gold, significantly lower than the average wealth of the sample data.
The average time since subjects first started playing World of Warcraft was 4.4 years.
The results were sorted by wealth and plotted into a Lorenz curve, seen on the right. The red area is the line of equality, if everyone had the same amount of gold, while the blue area is the Lorenz Curve, the actual wealth distribution, where the x-axis is the bottom % of household, while the y-axis is the % of income. This graph shows a drastic discrepancy between the poorer players, and the mega wealthy.
Some points of interest not well represented in the graph include:
- The bottom 75% owns 14.41% of the wealth (the bottom 25% and 50% owns 0.88% and 4.37% respectively)
- The first quartile was 12,000 gold while the third quartile was 108,198 gold, and as previously mentioned the median was 35,000 gold
- “The 99%” owns only 75.75% of the wealth, leaving 24.25% to the top 1% of players
- 19% of the subjects had under 10,000 gold
- 72% had under 100,000 gold
- 82% of players were under the reported sample average
- 96% were under 1 million gold
These benchmarks illustrate the disparity between the top gold makers in the game and the more casual players and how top heavy the population is in terms of wealth when 82% of the player base is “below average”.
The calculated Gini Index (the measurement of inequality where 0 is perfect equality and 100 is perfect inequality; lower is better basically) for the World of Warcraft population, based on the sample data, is 65.5. To put that in perspective, if World of Warcraft was a country, it would be ranked the second highest inequality, just below Namibia at 74.3 and just above Lesotho at 63.2, while Canada and the United States weigh in at 32.6 and 40.8 respectively, to give some scale to the disparity WoW players face.
Two of the questions included in the survey were designed to collect demographic information about the subjects regarding which activities they frequented in World of Warcraft and which resources they made use of out of game, with the following results:
|Do you raid?||Do you participate in PvP?||Do you Role Play?||Do you “play” the Auction House?||Do you farm gold?|
90% of the players answering the survey claimed to participate in at least one of these activities, while 10% answer “No” or “Uncertain” to all the activities.The average wealth of players who claimed they participated in the various activities were compared to the overall sample average:
|Raid||PvP||RP||Auction House||Farm Gold|
The activities usually associated with high competitiveness, raiding and PvP, tended to have more gold, with the more casual of the two having a smaller increase. Counter-intuitively players who claimed to farm gold had less gold on average, while those who claimed they played the Auction House had the highest average of all the activities. The small percentage of players that claimed they role-played had the lowest average gold of the activities. The players who participated in any activity had an average 8% higher than the sample average, while players reporting no activity in the listed fields had an average 75% below the sample average.
The following are the results from which resources the subjects used:
|The Consortium?||Just My Two Copper?||Gold Capped @ WoW Insider?||MMO-Champion?||Misc. Gold Blogs?|
In total 92% of subjects reported using at least one of the listed resources. The average wealth of players who reported using the various online resources in the survey were compared to the overall sample average:
|Consortium||JMTC||Gold Capped||MMO-C||Gold Blogs|
As the target audience of a resource trends towards the more casual warcraft players the average wealth of said audience declines. The Consortium, where content is more focused around discussion of advanced topics and not just a one way flow of information had the highest average, 294% above the general sample average. Gold Capped, where the target audience is more novice and is an almost 100% one way flow of information had a penalty, the average was 11% below sample average despite having the second largest audience. MMO-Champion stands out in despite not being oriented towards gold making, rather general game knowledge, had a larger wealth average in their readers than Gold Capped did. Players who claimed to follow at least one of the resources had a marginal 1% advantage, while those who claimed to not follow any had a 9% penalty.
Length of Play
The number of years since the subjects started playing World of Warcraft with the following results:
|Percent of Players||5%||9%||17%||19%||20%||19%||11%|
|Difference from sample average||-40%||+1%||-31%||+28%||-19%||+7%||+43%|
The trend from this data (seen in graph on left) indicates that as a player gathers more experience ingame, the more likely the player is to have an above average level of gold. This data threatens to be invalidated due to inflation of gold over the years, however since accumulated gold does not earn interest and most inflation happens at set points for everyone (expansions, patches etc.) it is safe to leave this data in for comparison.
The results of this report highlights a large gap in wealth between the upper and lower class players in World of Warcraft. With such a large percentage of players below the mean wealth level, the whole ingame economic system is an extremely top-heavy one, leaving large sums of gold sitting idle in the top players pockets kept isolated from contributing to the economy, which is especially detrimental if one considers that a server economy is the simple flow of gold and materials. One can argue however that by holding such large sums of gold such players can more easily influence and correct markets, keep inflation in check by keeping the sums of gold out of the hands of the masses, and provide buyers for big ticket items such as TCG Mounts.
It is also clear from the results of this report that activities that tend to attract more competitive players, Raiding and PvP, tended to have larger wealth averages amongst it’s participants while more docile activities, Role Playing, had a significantly lower average.Similarly, resource aimed at intermediate to advanced subjects and audiences provided a marked increase in wealth levels, even if the resource was not focused in gold making.In-game “age” proved to have a positive effect on the average wealth levels of players, though whether this is due to pure time to accumulate said gold or from the experience of playing so long, is yet to be determined.