Monday, 29 August 2011

Diablo 3: The economy in numbers

It may perhaps come as no surprise to regular readers of this blog that I've been thinking about numbers.

In particular I've been thinking about this comment of Loque's:
"Now... for what concerns the "making money in D3" I don't really think it will happen for 99.9% of the players."

Why 99.9? Why not 80% or 99.99999% Can we attempt to figure it out?

Here goes

The virtual currency market is currently estimated at over $2 billion dollars annually and predicted grow to $5 billion by 2011. (Source)

Older but more trustworthy data states:

"  Figures

While reliable figures for gold farming are hard to come by,[11] there are some estimates of the market for in-game currency.
In 2005 The New York Times stated that there were over 100,000 professional, full-time gold farmers in China alone.[5] And in 2006 sales of such virtual goods were thought to amount to somewhere between 200[12] and 900 million USD.[4]
Another estimate, drawn from 2005/2006 data, valued the market at not less than USD200 million per year[13] and suggested that over 150,000 people were employed as gold farmers with average monthly earnings of USD145.[13] This same report estimated that 80% of all gold farmers were from China;[11] a fact which has led to prejudice towards Chinese players.[14] 2008 figures from the Chinese State valued the Chinese trade in virtual currency at over several billion yuan, or nearly USD300 million.[15]"

An academic analysis by Richard Heeks includes mention of speculation that as early as 2006 the market might be larger than $1b. And this is a market that has seen and is seeing non-stop growth.

Updating his figures in 2009 Hecks mentions that Nick Ryan put the Chinese RMT industry alone at $10b:

"Estimates of 400,000 people working in China on gold farming/trading are seen as very realistic, and the number could be 500,000 up to 1,000,000 (the latter figure is also estimated in Nick Ryan's recent articles on RMT, which claims total turnover for Chinese RMT operations of US$10bn per year). There are many brokerages in all major cities in China (Ryan's article estimates 60,000 in total)."

So in Diablo 3's first year is it fair to suggest that at a conservative estimate the global RMT market might be somewhere between $2b to $5b?

What's Diablo 3's cut?

I think there are a lot of reasons why D3 players will be more inclined to buy items than players of other games. First it's legal. Second there's the allure of believing that if you make yourself just that bit more awesome you'll be able to recoup your investment. Third it's the successor to WoW, Blizzard's first new avatar-based game since 2004 so it will inherit a lot of players who bought gold items and characters in World of Warcraft. Fourth the game provides significant blocks (difficulty level changes, pvp) that many people will only overcome by buying items.

So let's guess $5b gets spent and D3 players spend a fifth of it.

Diablo 2 sold 4 million copies. I strongly suspect that Diablo 3, launching on the back of WoW will sell more than that. There are simply more gamers around these days as well as a much higher number of older gamers. Also Blizzard has done really well with opening up new markets for WoW in places like Brazil, Russia and particularly China. Add in the allure of being able to get paid for playing, whether it's naive or not, which will draw in many more customers and I think it's not a stretch to say a big chunk of PC gamers will be playing D3. I'd guess about 10 million. I think at least 5 million is almost certain.

So if D3's real money auction house is going to turn over $1b in its first year and it has 10 million players that's $100 per player each year.

It won't distribute like that.

Hardcore mode (permadeath) players will get nothing through the legal AH. There will of course be a secondary market. (I suspect the existence of this black market will eventually push Blizzard into authorising RMT in D3 Hardcore).

The majority of players will never reach endgame.

Not all farmers are equal. Roughly a farmer's income = monsters killed * magic find bonus worth of magic items plus monsters killed * gold find bonus worth of gold.

Luck plays quite a big part, causing profits to deviate from the average that any particular playstyle might suggest.

Game knowledge can also play a big part, for instance a better method of killing fast (such as Javazons in the Cow Level in D2) or a better farming route.

Cheating (duping etc) can play a spectacularly big part but overall diminishes the game as there's less point being so interested in a game where people are able to cheat and hack and dupe. Hopefully Blizzard will keep cheating under control.

Game difficulty will play a big part. If the game can be farmed by under-geared characters in a relatively brain-dead way then uneducated Chinese labourers are perfectly effective. If the game is more difficult then players will need more grounding and understanding of the game. I'm speculating it may be hard for the typical Chinese gold farmer we saw in WoW to do this, ie it will change from a blue collar job to a white collar job. I'm sure they'll get up to speed in time but gold farming businesses that are not prepared may have a very rough first year. The current system of sharing characters in shifts with the guy on the next shift selling anything that isn't nailed down will be completely useless for farming Inferno.

All forms of arbitrage aren't actually significant in relation to the turnover figures. Arbitrage moves profit from one economic agent to another, it does not actually increase turnover itself.

Conclusion: if you're a skilled player who farms Inferno a lot with magic find gear on I think you'll earn over $1000 per year playing Diablo 3. If you also arbitrage effectively you will earn more. It won't support any but a few as full-time professionals but it's a financially significant hobby.

Or to put it another way if you're a reasonably competent gamer you will pay back the cost of the box at least if you do some magic finding in Inferno.


  1. Well this one was a very interesting analysis, good reading (I love blogs). That said, the only problem I see here is just one: too many assumptions, speculations, guesses and (clever) theorycrafting.

    I understand you love this kind of stuff (as you explained us, you love numbers) but this specific analysis would need a lot more solid/verified numbers.

    I'm glad I could poke you and stimulate your mind :)

  2. Guilty as charged, it's a guess, just a guess really.

  3. Let me contribute using a WoW example and let's see how many people are interested in spending real dollars for virtual ingame items (vanity pets).

    According to this link

    the vanity pet raised over $1.9 million in sales. Each pet sold for 10$ and it was account-bound. So we have a total of 190.000 units sold.

    How many people play WoW? According to this article:

    we have 11.1 million subscriptions (worldwide) by the end of June, so 1 month before the pet's sales.

    So, in the end, how many customers decided to spend 10$ for the ingame pet?

    11.1 mil / 190.000 => 1 player every 58 decided to buy it.

    So, in the end, 1.7% is the amount of subscribers (worldwide) who actually spent real dollars for the charity pet.

    Considering that "charity" gives a lot of motivation (so you can say YES, I spent real money but I did that for Japan!), I imagine that common virtual items (non charity-related) will sell much less.

  4. you are comparing apples to oranges Loque, to test the market in WoW you would need to have loot of the current heroic raid for sale.
    And i get more people would want that, than a cute cosmetic pet.

  5. I know. We're theorycrafting here, using the sources we have.

  6. I do think though, that despite it's being guesswork the statement that 99.9% of players will never make anything does not apply.

    One in three is a widely accepted figure for how many people illegally buy stuff in games like WoW. There are enough buyers to ensure that if you find a half-decent legendary you don't want early on in D3 you'll get more than a couple of cents for it.

    So the playing model of - never buying, keeping suitable items, selling off top end items you don't need will probably pay for the box.

  7. What I consider an interessting question, how will the influx of money into the economy be structured?

    Lets say we have 10% players (called buyer, the other groups will be called normal players and fast levelers) willing to spend money, who buy in at the start of the game.
    As I cant imagine someone spending money on common drops, I guess they will buy three things: characters, gold and exceptional items ( uniques/epics from all the game and rare inferno level drops).
    In WoW as in Diablo2 and nearly all RPG games, gold drop rates are growing fast with levels, so most of the gold in the economy will be farmed by high level characters either in inferno or if inferno is to though for them, farming weak monsters fast in hell.
    Together with the fact of the best items dropping in inferno, but being usable by hell chars of the right level, leveling fast to lvl 60 and farming like a madman should be very lucrative, espacially if the "Game difficulty will play a big part." comment about proffessional farmers is true.
    So the big crowd wont be able to sell gold and wont be willing to sell their character. That leaves Unique/Epic item drops for the normal players. The gained money will be valuable to them and they will only use it for the mentioned above "valuable" items and gold. I forgot the runes, if they are rare like runes in D2 (i remember to have read that the best ones only drop in inferno) they will too be "valuable" items. So the normal players will buy these vauable goods from the fast levelers or unique/epic drops from eachother. The fast levelers dont need any items the normals have to buy, as i assume that a lvl 60 inferno rare sword drop should be better than say a lvl 40 unique sword.

    Conclusion: in the time until most players have a max lvl inferno running character, the circuit of money should look this way:

    Buyers: deposit money at blizzard, make most purchases of fast levelers, a smaller amout of purchase from the normal players.

    Normal players: sell valuable random epics, spend that money later for stuff the fast levelers farmed or other normal players got at random

    fast levelers: all the money ends up here, $$$

    I dont know how long this golden state for powerlevers will last, as I dont know how many hours and what skill level it takes to reach inferno farm status. My guess would be around one to three monthes.

    The posibility of buy/selling chars isnt figured in, but the expected profits of the first inferno farm chars will doubtlessly drive the price up.

  8. I think that's essentially right Christof. That's what I was saying last week, that I'd aim to get straight to farming Inferno as soon as possible.

  9. me too, I remember leveling a Deathknight to server 5th or something, and it was a blast.
    Unlike the level grind at a wow expanision, which is more a sprint I expect this to be more like a marathon.

    Which offers the question, run or let run?
    If i would own a chinese farming agency, i would try to let my unskilled workers grind chars up in shifts and sell them to people wanting to get a rested start into inferno. With this business approach I could use the skill level of each worker optimal, as i would have teams for say 1-20, 20-40, 40-60 for each class, so each worker gets experience only in the level range he needs.

  10. If D3 is as big, as transformative, as I think it's going to be we'll see a radical change in the Asian RMT industries.

    Most likely is that a lot of the unskilled gold farmers get laid off and replaced by skilled professionals who work in synergistic teams. Unskilled workers will get made redundant because in the new paradigm they are redundant.

    Character leveling is though, as you say, a task that could remain viable for the unskilled.