Gold Capped: Tracking the most frequently bought and sold items
by Fox Van Allen 
Jan 16th 2012 at 7:00PM
One of my favorite topics here on Gold Capped is World of Warcraft's problem with inflation. If affects just about everyone in a very negative way, regardless of whether they're an Auction House maven or a casual player. Inflation makes any gold your character is holding worth less and less by the second, making work you do now far less valuable than work you do later. It even affects the way developers approach the economy, from the amount of gold you get for finishing a daily to the creation of new gold sinks.
By most anecdotal measures, in-game inflation is wildly out of control. And that's one of my problems as WoW Insider's other market follower; the only evidence of inflation we have is ancedotal. There's no real solid way for us to measure inflation in the game and understand what's working to control it and what's not.
The question got my mental gears turning. In the real world, inflation is measured using something called the Consumer Price Index. Creating an in-game version of the CPI intrigues me, but to figure out the best way to construct it, we need to first figure out the answer to another difficult question: What do people buy the most of in-game?
What's the Consumer Price Index?
I'll spare you the full economic lesson, but a Consumer Price Index is a pretty simple concept. You take a selection of items that the average American buys -- things like bread, clothes, electricity, cars, gas, and housing -- and keep tabs on the price of each. Then later -- say, in 10 years' time -- you can compare the new prices to the old prices and see how they've changed. The percentage that prices go up (and they almost always go up) is the rate of inflation.
Keeping tabs on price changes is the easy part, though. The harder part about crafting a CPI is taking into account the amount of each item a person buys on average over a year and to adjust the weighting of each item in the market basket of items we selected. If the price of bread increases by 10% and the price of a house decreases by 10%, we can't just average things out to zero, because people tend to buy more loaves of bread over the course of a year than they do houses. Unless they're a real estate investor with a gluten allergy. Hmm.
... and this has to do with WoW ... how?
Hey, don't rush me. I was getting to that part.
If we take the concept behind the real CPI and translate it into Azeroth, we can measure the effect of inflation on the World of Warcraft economy and identify the best places to invest our hard-earned gold. Websites like The Undermine Journal and AHSpy.com make it really easy to track prices. We just need to figure out what the typical player purchases most often in a way that represents what people buy the most. But what do people buy the most?
It's hard to track the number of items sold on the Auction House with the given limitations of the API. We'll have to keep track of the number of items listed, instead -- a far easier number to determine. Obviously, not all items listed will be sold, but economic theory suggests that they'll all sell or be used eventually once the right price is found.
So, enough set-up. Here are the items that are bought and sold the most.
Top seller: Raw materials
By far, raw materials make up the most common purchases on the Auction House. That's because just about everything else being sold on the Auction House (with a few exceptions) are comprised of these raw mats.
Herbs To represent herbalists, alchemists, and scribes, I'm going to include in our inflation-measuring market basket the two most traded herbs. Currently, this is Whiptail (1.89 million units globally) and Cinderbloom (1.68 million). We'll also include Peacebloom (0.19 million) to try and capture the impact of those leveling professions on the market.
Ore To represent miners, blacksmiths, and jewelcrafters, we'll keep tabs on the two most traded ores. Currently, this is Elementium Ore (3.86 million units globally) and Obsidium Ore (1.62 million). We'll also include Copper Ore to capture the leveling market (0.51 million).
Enchanting mats To represent enchanters, we'll have to look at enchanting material prices. We'll keep tabs on the current expansion's dust, Hypnotic Dust (2.21 million), and the current expansion's greater essence, Greater Celestial Essence (0.21 million), the most traded enchanting mats by far. We'll also keep track of the changes in the most basic enchanting mat, Strange Dust (0.36 million).
Cloth To represent tailors and those poor suckers who actually make and use bandages, we'll include the current expansion's common cloth, currently Embersilk Cloth (1.11 million). We'll also look at Linen Cloth (0.91 million).
Leather To represent skinners and leatherworkers, we'll look at the current expansion's leather, currently Savage Leather (0.70 million). It easily sells more than all the other leathers combined. We'll also follow movements in the price of low-level Light Leather (0.43 million).
Elemental And finally, many craftables require elementals. We'll keep track of the two most traded elementals, which are currently Volatile Life (3.05 million) and Volatile Earth (2.16 million).
Raiding goods
While your more casual player makes the raw material market go 'round, there's a segment of the market that more casual players are less influential in: the high-end raiding market. Surely, there are far fewer hardcore raiders, but the items they buy tend to be exponentially higher priced.
Gear The best way to track the movement of raid-quality BoE gear will be to keep an eye on valor gear. To try and keep things somewhat consistent, we'll keep an eye on DPS plate (assuming all gear for other spec behaves similarly) because it's the most frequently listed item (846 global listings). In this case, we'll watch Heartcrusher Wristplates until a new class of BoE gear comes out next expansion.
Enchants and gems For the most part, the cost of enchanting and gemming that gear should be covered in the raw materials section. Just to be safe, though, we'll follow the costs of the highest-quality red gem available (currently, Queen's Garnet; 5,058 global listings) and the highest-quality enchanting mat available (currently, Maelstrom Crystal; 62k global listings).
Putting it all together
We'll keep track of the prices of these items over the upcoming months to try and identify some trends and to give us all a better idea of what to expect in the next expansion. And while the data accumulates, I'll keep up pressure on Editor-in-Chief Alex Ziebart to build me a multimillion-dollar, NASDAQ-style video wall.
So, what do I expect the data we're collecting to show? Well, if history is any judge, you should see purchasing power increase until Mists of Pandaria is released. Why? Decreased demand -- fewer and fewer people will need raw materials, gear, and the items used to enchant and gem that gear. As a result, most people will simply stockpile money.
When Mists of Pandaria is released, history tells us the dam will break. Players will finally have something to spend their pent-up money on, and since supply will (initially) be very tight, we'll see inflationary problems worse than ever before. As for fixing that problem -- well, we'll get to that issue very, very soon.
In the meantime, what do you think? Does my selection of items accurately represent the World of Warcraft economy? Is my market basket missing some crucial item? Also, do you have a massive gently used video wall for sale? PST.
Maximize your profits with more advice from Gold Capped. Do you have questions about selling, reselling, and building your financial empire on the auction house? Fox and Basil are taking your questions at fox@wowinsider.com and basil@wowinsider.com.
Filed under: Economy, Gold Capped
Reader Comments (Page 1 of 3)
rovingbandit21 Jan 16th 2012 7:13PM
Do not forget big ticket items, which in WoW's case would be mounts and pets. While not everyone buys them, Mounts and pets are corollary to Houses and Cars in the varied other indices that show inflation or the lack there of. I do like the idea of creating WoW's version of the Big mac index though. Find the most popular food on the AH and compare prices across servers to find the most depressed/expanded server economies.
morncreek Jan 16th 2012 8:28PM
Mounts would not work. There are only two non-TCG, non-BoP mounts in the game - Reins of the Crimson Deathcharger and Reins of Poseidus. Everything else is BoP from a boss, quest, or from a mount vendor. You would have to take into consideration two guild perks, faction reputation, and goblin racial. Those factors would be impossible to track when taking into account the size of WoW's player-base, and tracking the previously mentioned mounts would be too small a sample size because of rarity.
Pets, however, might work. Leaping Hatchling, Darting Hatchling, and the Argent Tournament pets are probably the best bets.
Bapo Jan 16th 2012 8:38PM
morn, you completely forgot the chopper and the drake from alchemy
morncreek Jan 16th 2012 8:47PM
Holy... I completely did. And now it's forever on the internet. T_T
Sahara Jan 16th 2012 11:54PM
The problem with including argent tournament pets is that their price increases with the scarcity of players willing to farm the seals required to buy them. Their inflation rate, taken separately, would be quite high, given how the AT became a ghost town once the next expansion was released. Wouldn't that skew the rest of the data?
Adam Jan 17th 2012 11:50AM
@Sahara
Those tourney pets have a chance to drop in tank and healing bags.
Adam Jan 17th 2012 11:52AM
@Sahara
Those tourney pets have a chance to drop in tank and healing bags.
Aykwa Jan 17th 2012 2:38PM
There are a few other considerations/problems with the tracking that is being attempted here:
Because of patches and expansions, there is a kind of built-in expiration date on most or all of these goods or services. For example, Whiptail will become comparatively useless when MoP goes live, and its price will decrease as we get close to that. When tracking inflation in RL, you don't have to worry about bread becoming obsolete in 6 months, and the inherent associated deflationary pressure that comes with that knowledge. As such, most of these goods suffer from some price decay. Thus a model that treats these goods more like a stock option rather than a commodity would be best.
Also a similar problem, there is almost no real way to carry value over from one expansion to another EXCEPT for gold itself. In the real world inflation greatly benefits 2 types of people: Borrowers (as their debts become relatively smaller, due to monetary inflation), and those who hold physical goods which do not depreciate. In a typical RL inflationary scenario, if I borrow every penny I possibly can to buy property, gold, pork bellies, oil, and oranges, I can expect the value of those items to move up with the inflation, and thus the value of my borrowed money (my debt) to go down. I'd make a killing. People who have lots of cash and little/no physical commodities would get crushed as his cash would become worthless. BUT, in WoW-style inflation, there is no way to carry value past an expansion in commodities. Those commodities will have severe price erosion, and so people all know they have to get rid of their goods and turn them into in-game gold in order to not lose large amounts of value.
This inherently means that the supply of gold in the hands of players will tend to increase as you get close to the end of an expansion (driving up monetary inflation) and that the value of commodities will tend to decrease as you get close to the end of an expansion (driving down commodity inflation). Thus the model for inflation in WoW is very different from the one we see in RL, and I believe that a totally different model would have to be created to more accurately measure in-game inflation. One for which we may not have access to enough data to be accurate with.
On top of all that, wow markets are highly fragmented (due to non-connected realms and AHs) with massive pricing differences between realms that simply don't exist in RL. Each realm will have its own inflationary rates dependent on many factors, such as population, average player laziness, number of gold-farmers/bots on that realm, number of players hacked on that realm, number of illegal gold-buyers on that realm, percentage of players who are market-savvy on that realm (willing and able to manipulate the prices of goods on the AH). For example, for an Inferno Ruby, right now you would pay about 550 gold on one realm, 100 gold on another. That is a massive spread on the price, of about 5.5 times. In the US, if that imbalance existed for one oft-traded good between say New York and California, people would be buying up all the product in Cali and shipping it to NY, paying a little shipping, and making a huge profit. They'd do it until the prices all over the country became much more comparable. But in WoW, that can't happen right now (maybe someday in the future). If everyone in game were on the same AH, prices would not only equal out, but the average price of all Inferno Rubies would likely drop, as the market becomes more liquid and more efficient.
So yeah, I've probably gone on way too long here, sorry. But the point is that there are some serious problems with trying to use a CPI strategy that works in RL to calculate numbers for the in-game economy. To really do so we'd need long-term, real-time access to Blizz data on actual holdings (gold AND items) for a representative number of players. Then we could calculate the net worth of the average player (across all their toons) through time, and really have a stab at calculating in-game inflation. I would guess that someone at Blizz is already doing this. And the only way for us to do it would be to have someone create an addon that would be installed by many players (hopefully a representative population) and would track all of that and report it back to one central location for processing/analysis.
Ben Jan 17th 2012 6:04PM
In-reply to Aykwa :
Thanks for the fascinating analysis.
However, I think the point Fox makes could still be evaluated in short-term, and might show already some trend over a mere few weeks.
The assertion is that the trend would be roughly the same across all realms, because the root cause is shared regardless of the average population, save perhaps on realms with a population largely focused on PvP, or with a very low-level population.
Fox Van Allen Jan 17th 2012 7:18PM
"Because of patches and expansions, there is a kind of built-in expiration date on most or all of these goods or services. For example, Whiptail will become comparatively useless when MoP goes live, and its price will decrease as we get close to that. When tracking inflation in RL, you don't have to worry about bread becoming obsolete in 6 months, and the inherent associated deflationary pressure that comes with that knowledge. As such, most of these goods suffer from some price decay. Thus a model that treats these goods more like a stock option rather than a commodity would be best."
Yeah, this project is a major challenge for the reasons you laid out. But I think you missed a point about how I'm going to proceed.
I'm not actually tracking Whiptail and Cinderbloom. What I'm actually tracking is the "two most heavily traded herbs." Unfortunately, those quotation marks were removed in copy editing. But they're important -- they signify that the specific item I'm tracking will change from expansion to expansion, and in some cases, patch to patch.
It's somewhat similar to how the CPI addresses things like car prices. Obviously, the change in the price of a 1966 Ford Fairlane says little about what's happening with regard to inflation. So instead of following that specific car, the CPI would instead follow the average mid-sized vehicle of the current model year -- say, a 2012 Ford Taurus.
They're very different cars, but they're as similar as you're going to get for the purposes of measuring a CPI.
Fox Van Allen Jan 17th 2012 7:44PM
Oh, and one point -- I intentionally did not include the Chopper or Vial of the Sands, because those two items are relatively fixed in price due to the bulk of the mats coming from a vendor. Even if inflation is 500% over the next expansion, the price of the Chopper and Vial should stay at a 10 or 20% premium over the cost of mats.
scherbaddie Jan 17th 2012 8:45PM
Perhaps I'm misunderstanding the theory behind this, but are you saying:
(Price now - Price 10 years ago) / Price 10 years ago = inflation
Eg. Bread 10 years ago = $1
Bread now = $1.10
(1.1 - 1) / 1 = 10% inflation
I'm not sure tracking the price of raw mats like this is going to provide a similar method of measuring inflation. Prices on the AH generally spike after a patch and come down over time as demand for the mats drops. Bread does not do this (Banana's do in flood/cyclone season in Australia, but I hate bananas anyway..).
We all know how it works by now: as demand for enchants drop, enchanters buy less essences, more sit on the market and get undercut and the price drops. Because the essences/dusts sell for less money, I respond my lowering my maximum buy price for Elementium/Saronite Ore and the price of the ore drops too.
Essences just after 4.3: 100g
Essences now: 70g
Inflation = (70-100)/100 = -30%
So are we saying the value of gold has gone up since 4.3? Sure i can buy more essences for the same amount of gold, but I don't want to buy essences, I want to sell them.
I thought inflation was to do with how rich you are compared to your peers (in wrath the average player had 10k gold, now the average player has 20k gold, so gold is worth 1/2 of what it was in wrath).
scherbaddie Jan 18th 2012 12:25AM
Actually I can see how this would work in the long run expansion to expansion. Market price of Saronite Ore is 3g now, Market price of Pandamite Ore is 4.5g in 5.0.6, so inflation is 50% (or "I can buy 2/3rds as much ore with the same gold than i could last expansion", so gold is worth less than it was).
So when MoP launches, will you keep tracking Saronite/Whiptail/Cinderbloom/etc. until we get to the point where MoP mats overtake the Cata ones in terms of volume listed on AH?
Would flasks be a good raiding consumable to track? They're probably more of a casual consumable if most raiding guilds use cauldrons.
Aykwa Jan 18th 2012 5:55PM
Fox,
I'm glad you're taking on this project, and I understand there will be limitations. I'm sorry if I came off a bit negative, mostly I was hoping to address challenges, but I didn't do a good job in communicating that.
In regards to:
"It's somewhat similar to how the CPI addresses things like car prices. Obviously, the change in the price of a 1966 Ford Fairlane says little about what's happening with regard to inflation. So instead of following that specific car, the CPI would instead follow the average mid-sized vehicle of the current model year -- say, a 2012 Ford Taurus."
Yes, that case can be made. Caution, however, should be taken because it isn't quite comparable. For as we approach the end of an expansion life cycle, both supply and demand drop off quite a bit. The fewest players are typically playing, and those who do aren't often crafting items with that expansion's mats, etc. While herb or ore or enchanting mats prices may drop by half the last weeks or months before a new expansion, the same cannot be said for new car prices right before the next model year comes out. You might get a 5% or 10% discount for a 2011 Ford Taurus as the new 2012s arrive, but you're never going to see it drop to half the week before, and to almost nothing the week after the 2012s arrive on the lot.
Of course nothing in life is perfect, and I'm sure you'll do your best. I wish you the best of luck on the project. The results will be interesting to see.
Meighan Jan 16th 2012 7:15PM
I find that valor BoEs behave very differently from actual drops, because the supply of them is far greater.
Scuac Jan 16th 2012 7:36PM
I think the whole idea is excellent. The only thing I am not convinced of is the BoE valor gear. As we saw through these last three tiers, they are not always for the same slot. If in the next tier the VP BoE is a chest piece, it is naturally going to be more expensive than bracers regardless of inflation.
clundgren Jan 16th 2012 7:41PM
Is there anyway to include green quality items? These have always been a significant market, more so since transmogrification.
Nate Jan 16th 2012 7:48PM
As an economics student myself, this sounds fascinating. Due to the volatility of WoW compared to real life, it will be hard to get precise data, but it should be great for getting a general picture. I think all of the items you chose were great; the only addition I would consider would be one of the most popular fish/cooking materials that are used by raiders for the well fed buff (weighting it lowly but including it nonetheless). I'm interested in hearing how this turns out.
Fox Van Allen Jan 17th 2012 7:29PM
That's a really good point. I'll try to include a fish component.
Bumblebee Jan 16th 2012 7:51PM
Something that should be noted for this is Repair Costs. To have any kind of success in game everyone needs to keep their gear in good shape. Could this be considered as a tax of sorts?