Profiting From Third-Party Dupes: The Most Profitable Legal Trading Strategy In WoW (Patreon)

 Summary: In this article the issue of detecting and potentially profiting from third-party dupes, ie virtual item duplication that you are not carrying out personally, is discussed. Simple methods for analyzing when a dupe is taking place are explained, the advantages of investing in these situations are discussed, potential countermeasures are discussed, and some simple formulas on investment strategy are explained. 

 Note: This is probably the most powerful possible trading strategy in wow or mmo's generally. In short, it will make money faster than anything else other than actually creating dupes yourself.

 Introduction To The Concept

 With the price of Spirits Of Harmony spiking in value recently I thought it would be a good time to discuss this as buying into someone else's dupe is the most advantageous gold making situation in the game.

 Detecting and profiting from some kind of rigged market without actually doing the rigging is a mainstay of virtually every form of speculation I've participated in or even studied. The mathematics of it are compelling: in any efficient market profit margins are going to be extremely low or non-existent but when a third party is attempting to manipulate a market that changes things dramatically. Note that with wow it isn't necessary to be right 100% of the time that a market is being manipulated in some way, anything over 50% is useful information. 

Simple Methods To Detect Dupes

 Items rise and fall in value all the time and it is important to be able to differentiate between random fluctuations in the market, a high-roller dumping stock and a dupe. Generally speaking the emergence of a dupe has certain commonalities:

 1) The price of an item falls significantly by over 2/3 of its value over a non-trivial time period (ie weeks or months, not hours/days). 

 2) Items are duped in significant volume by a specific player or player/s in a specific realm, and also the dupe will usually be carried out cross-realm. This distinguishes dupes from normal market volatility which tends to become less significant with a large sample size.

 3) Any attempt at a market reset is unsuccessful.

 4) Duping occurs within a 3-month ban cycle. 

 5) There is no obvious shift in the fundamental value of an item based on public information. 

 (Note that these basic detection principles can mostly be applied to mmo's other than wow, or indeed any type of market where the probability of some sort of legerdemain is high. Only 4) is specific to wow, and even outside wow there will usually be some pre-definable window of opportunity). 

 This is essentially an issue of statistics. A statistician would probably go further than I have here and probably use something like an npq standard deviation test, chi-square test or other similar method to detect anything weird in a data set. This can be a very rewarding and profitable area to study but for this article I have tried to keep things simple. 

 Some Historical Examples

 I first begin to look into this area a few years back during the start of the Legion expansion when it was briefly possible to use class-trial characters to virtually instantly receive a garrison quest which awarded a stone which would upgrade any battle pet to level 25. Pets flooded the market - for example multiple Shore Crawler battle pets were selling on my own realm for just 200G. Shortly after the dupe was fixed the price went up to tens of thousands, where it had been prior to the dupe. Today a level 25 Shore Crawler still sells for an average of over 25000 gold. Even considering the pet market is quite slow it is easy to see that buying these pets up for virtually nothing was a sound strategy in the long run. 

 More recently a dip in the price of Spirits Of Harmony in the run-up to Christmas saw them go down from around 350G per spirit to 80 gold on my realm. I bought up thousands of them and eventually sold most of them for 395 gold. 

 There are many other examples. 

 Advantages Vs Disadvantages

 Backing third-party dupes has several advantages over trying to find dupes yourself:

 -No knowledge of duping itself is essential beyond knowing what to look for.

 -Dupes are a closely guarded secret among hi-level exploiters who do not generally share information outside of their immediate circle. They are not easy to find. While they can be very profitable when they can be run, in between finds there is an awful lot of research and unprofitable downtime. Someone who is just looking for other people using dupes is efficiently backing the labour of multiple geniuses at one time and expending relatively little effort. 

 -Naturally the original duper is in the best position to exploit the commercial advantage of the dupe: but the individual who bets into third-party dupes has more opportunities. His income can exceed that of the non-prolific duper, and will generally be more stable.

 -Countermeasures are less likely to be directed at opportunists cashing-in than the original source of the dupe. *

 How should you invest when you discover an apparent duping situation?

 The short answer: very aggressively. 

 Longer answer: Books have actually been written on this. The basic formula comes from an obscure mathematical paper by a man named John L Kelly over sixty years ago in a publication known as the Bell System Technical Journal. This paper forms the basis of much modern gambling and investment theory.

 https://onlinelibrary.wiley.com/doi/abs/10.1002/j.1538-7305.1956.tb03809.x (this is not easy reading). 

 To greatly simplify what Kelly wrote here: in order to profit from a gambling/investment situation without unacceptable risk the player should bet the % of their total wealth equivalent to their advantage. So, if you have $10,000 and a 10% advantage on an even-money bet you would bet $1000** (10% of $10,000). This grows your money at the fastest possible rate whilst keeping the probability of encountering enough bad luck to wipe you out very low. 

 Because real-life situations tend to be more complex it isn't generally possible to apply Kelly strictly, but it gives a very good general guide. 

 I estimated that in the above pet dupe scenario that shore crawler pets would likely return to their original 10K gold value after my dupe had caused them to sell for as low as 200g on the ah. The risk on buying these pets assuming an open-ended time-frame was very small: limited to only scenarios where the pet crashed permanently or wow itself did-in which case gold would likely become valueless anyway. The upside if I was correct was 5000%. In this scenario risking close to or  your entire capital (keeping some amount for essential game-playing purchases) is clearly justified under Kelly's criterion.

 Let's look at a more complex example where the Kelly criterion is more useful as the correct amount to investment isn't as intuitively obvious. 

 In the recent Spirit Of Harmony scenario I had access to multiple spirits at an individual price of 80G. My guess was that they would return at least to a baseline of 200G (in actuality this turned out to be very conservative but let's go with this for now). That represents a return of 150% on investment. 

 Again, there is technical risk. It would be possible for the price to tank at lower than 80 gold for a prolonged period-but it is difficult to see how how this would happen unless Blizzard made some fundamental alteration  to the game which would make spirits worthless. Such things have happened but do not happen often. Even if the probability of losing my entire investment was 50% (far too high imo) then as long as my projected profit if successful the other 50% of the time is 150% my average profit is 50%. A Kelly bet would still be half of all my gold. 

 In practice you would almost never be betting less than 50% of your bankroll and if you aren't committing 75% you may be doing something wrong. The exceptions would be if you had some idea a dupe was taking but not that much confidence in your assumption-in such scenarios your optimal investment tails off sharply.

 Observe that, even if you are completely wrong about the existence of dupe: this may not actually matter. If a high-roller is just dumping all his inventory over a protracted period and this happens to coincide with a duping pattern, it doesn't necessarily matter for mathematical puropses: it is more a question of philosophy than anything of practical importance.

 * For high-rollers this still applies with less force. In theory a small-time dupe operation could be dwarfed by the capital of a big trader who detects it. In this scenario it is likely that Blizzard would have some kind of automated correlation trigger in place that would prioritize the trader as the source of the dupe, regardless of the actuality. 

 I considered titling this article "The Most Profitable  (Quasi-)Legal Trading Strategy" for this reason, but this places too much emphasis on a risk which would only in my opinion affect a tiny minority of gold investors in niche scenarios.

** Note that this example assumes an even-money wager. For more complex payout structures you adjust by dividing the ratio of a winning wager to a losing wager. So if a winning payout in the $10,000, 10% advantage scenario is 10 for 1, a Kelly bet would be 10% of 10000 divided by 10, or 1%. 

Note also that we are not dealing with a purely mathematically deterministic scenario like the toss of a balanced coin or a fair deal from a deck of cards. My (or your) estimates are just that: estimates, I can be wrong. This "variance around the estimate" or, to use less pretentious language "chance I messed up" is difficult or impossible to quantify, but most experts seem to agree that some conservative modifier is best used when calculating  the size of your investment in marginally profitable situations.  Fortunately when investing on the auction house you will not generally be looking for marginal situations.