Analysis of Flask Prices

We don’t normally break out individual server prices- preferring instead to look for and study trends identified by producing an average across servers. Examining an average allows us to normalize single-server anomalies. This is important as a server’s economy is a fragile thing and can be manipulated significantly by a single person (even, sometimes, unintentionally). Our averages correct for this imperfection by producing a composite index and, ideally, a trend. However, given the recent conversations on the reliability of averages on both this site and others, we decided, this time, to give the data a second look by breaking out the individual server reports. What we found was most interesting…

Recently, we tested the prices of the four most popular Northrend flasks. To be perfectly honest, after a brief two-week study we found the initial results to be fairly unremarkable. Outside of the normal slow deflationary movement and slight weekday vs. weekend cycle that we’ve observed and reported on previously (and this time to an even lesser degree) it would seem that there was really not much to report.

Here’s the data for a one of the flasks, the Flask of Stoneblood:

2 Week High: 29.345 on Saturday, May 25, 2009
2 Week Low: 24.1625 on Monday, June 1, 2009

Flask of Stoneblood- 2 weeks- 4 srvr avg

Fairly unremarkable right? However, upon closer inspection we find a few interesting observations. By taking away the four-server average function and looking at the flask’s performance on individual servers we start to see some fascinating patterns in price ranges . Take a look at the data yourself. Our observations follow.

Flask of Stoneblood- 2 weeks- 4 srvr unique

Our Observations:

  • Not all servers followed the weekend = higher prices cycle. One server completely bucked this trend by showing lower prices on the weekends and higher ones on weekdays.
  • As we observed with another recent analysis, prices were slightly higher on most servers (3 out of 4 this time) on the Alliance side than those of their opposing faction, the Horde. Of note, these were 4 different servers from the last set we saw the same phenomenon. Does it cost more to play Alliance? Does faction have an effect on prices?
  • The higher the price point, the wider the amount of variance in the prices. This makes sense, however, when we look at the variance in terms of percentages instead of hard amounts where the data reveals itself to be consistent.
  • Price points were consistent on servers on both faction sides. If a server had higher prices on the Alliance side, prices were also higher on the Horde side of that same server. This was observed in every single test and has been observed before. There is, seemingly, some correlation between prices on both sides of the fence for each server which is probably evidence of ample cross-faction sales.
  • There was certainly ample evidence, across both factions of all four servers that a persistent up and down pricing cycle exists. This should provide some trading opportunities for the astute trader.

We do our best to not speculate or draw opinionated conclusions of our tested data. We prefer to leave that part of the process to you, our readers. It should also be said that we ran similar tests on the other three flasks and are continuing our analysis internally, amongst our team members. We intend to report additional data on flasks in the coming weeks provided there is ample interest from the WoWenomics community.

A note on our testing methods:

As always, we use our unique formula for Going Rate as our measured price point as this allows for us to assess only realistic prices. Four servers were tested in this round. This means that we actually ran tests on 8 unique auction houses as both factions are tested and then compiled into a single, reported server average. Tests were run for 15 days, Monday, 18 May 09 to Monday, 1 June 09. The servers we used were selected for population and reported auction house activity factional equivalence. Tests were run at the same time of day, once daily in this case, for 15 consecutive days. Additional information on our testing methods can be found on our other Market Data Reports.



Filed under WoW Economy News, WoW Gold Making Tips, WoW Market Commentary, WoW Market Data

8 responses to “Analysis of Flask Prices

  1. taco

    Considering that frost lotus are going for 30-40g on my server. I’m not sure how the alchemists are making any gold out of this (maybe they rely on procs).

    It doesn’t make sense to me why anyone would sell flasks for these prices.

  2. I’d suggest to check the servers on

    On more raiding servers the flask prices are up on Tuesday-Thurstday (like the red line on the chart), as lazy raiders buy their consumables on the last day, while more casual servers have the standard weekday-weekend cycle.

  3. Merge

    Well think about that, if you the lotus goes for 30-40. and you make at least 2 flasks per lotus due to the recent changes. Factor in the Single flask going for the lowest being 24 gold…. still make 8 gold profit give or take, due to the AH cut. sell the two flasks you get from 1 lotus you make 48ish.. and that is if you sell it for around 24 a piece.. so yeah that is how they make profit

  4. Ewber

    Yeah my server (Azgalor-US, my guess is it would be a ‘raiding’ server, but I can’t click the link to check while at work) sees higher prices on most raiding related goods (flasks, enchant mats, food, herbs, etc.) on Tues.-Thurs. with a declining trend starting on the weekend just like your “red” server.

    I assume this trend occurs as more casuals come in and put their goods on the AH and as the more hardcore put all their ‘leftovers’ they have from clearing Ulduar on tues. and wed. night.

    Also, since pvp is pretty big on Azgalor, I assume the higher enchanting mat prices come from people spending arena points on new gear that needs gems/enchants.

    And also as Gevlon pointed out, on the weekdays many of the people who raid, while not necessarily ‘lazy’, just don’t have alot of time after they get off work (generally 5pm-6pm) to raid time start (7pm-8pm usually) and are in a hurry to get their flasks and enchants before the raid and dont really care if they are paying a 10-20% premium on prices.

    Great post guys, I love the trend analysis graphs.

  5. Buymetoys

    Although I understand tracking prices at the same time everyday for purposes of having consistent data, I’ve found that on my AH the numbers vary widely over the course of a day. Prices spike before raid time every major raid night (M-Th) and fall in between raid times.

    Generally, I’m able to sell flasks at about a 20% premium if I list shortly before a raid. I’d think that has something to do with motivated, price insensitive buyers and also something to do with some sellers listing at the ‘wrong’ time.

    Similar things happen with lotus prices and I suspect with other flask mats, although less so as the supply is generally there to better absorb demand.

  6. Interesting data. A few questions about your survey:
    – How accurate do you think it is given that you’ve only checked prices ones per day and only for two weeks. I don’t think you can derive weekly trends from this short experiment.
    – Were all servers in the same timezone? Were the raid times the same? This could explain the difference in hi/low weekend/weekday data. If the check time was before the raid on one server and during/after the raid on another, there would be differences in trends.

    • jederus

      Hi BBM,

      Well, the data is 100% accurate as far as it being a factual representation of the prices we observed. As far as it being an appropriate representation of the observations we’ve made, I’d agree that we could stand to do a longer test to become even more familiar with trends. In fact, as we mentioned in the post, we are still testing this (and the other three) flasks. That said, I’m pretty sure that we’ll still find that our observations ring true even with longer testing- hence the reason we posted it.

      Be sure to keep stopping by so you catch our next, more comprehensive, flask testing results. That said, I’m not sure yet if we are going to break each of the four flasks into individual server readings or not yet.

      All of the servers used were different time zones. The ‘time’ we refer to was time of day. Yes, there may be some disparity between prices due to raids starting up but, we feel, choosing servers from all the same time zone might produce results that were skewed in other ways.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s