Most of you are likely to have noticed the big QA vs Auctioneer "battle" Gevlon caused with
a few of his posts not long ago which resulted in a huge argument on wether QA or Auctioneer's batch post function was best for glyph posting. Gevlon claimed that QA was "an automated noob" which could only undercut by 1 copper. I wont point out the flaws in his arguments, enough people have done that so there's no need to.
This was almost a week ago and most people have stopped caring about it, so why do I bring it up? Well, the way Gevlon defended his strategy got me thinking that if he cares so much about it it must work, so I took a look at it in detail.
As seen in
this and
this post he has already explained his methods twice, but I didn't really pay much attention to them. Now I notice that what is very special about this way of posting is that instead of crafting a bunch of every glyph and throw them up on the AH while trying to keep your threshold as low as possible to avoid being undercut is that it uses the Auctioneer market value, both to decide if the glyph is worth making and if it should be posted. He set a minimum value they need to have to make them and then starts crafting. Auctioneer is configured to not post glyphs at less than 60% of their market value. This makes sure he gets a good profit on close to every glyph he makes and gives a very good gold/hour ratio, but it can't be compared to carpet bombing when it comes to pure gold gain. So I started thinking, I don't got all that much time for working the AH, so why not try to optimise gold per hour?
I started thinking of trying the goblin method, but I saw a few flaws.
1. It takes a while to choose all glyphs that are going to be made (can be solved with KTQ though)
2. It's very dependant on regular scanning of the AH which I sometimes forget.
3. It relies on market values that doesn't instantly show changes to the price. A glyph with a market value of 20g may only sell for 7.
4. Market value is a subject to change over time. A glyph that is worth 20g according to auctioneer maybe only got a market value of 4g in a few weeks if it hasn't sold, resulting in a loss.
5. Many pointed this problem out in the discussion, auctioneer posts very slow and it takes really long time to get all glyphs up.
So why don't improve it? If I use QA instead of Auctioneer I can rule out most of these points, it's fast removing point 5, I can set it to a specific threshold to never go with a loss and scanning doesn't matter. That leaves only point 1 and 3, 1 can easily be solved by using KTQ to craft all glyphs, and only recraft those I have sold. That way 3 is also solved, I don't focus on any specific glyphs but go for them all, no matter value knowing that they will sell in the end.
By then having a high threshold and a low fallback I will put every glyph on the AH in that area unless I'm undercut below my threshold. So how will this work then? It's simple, the AH goes in cycles. You've all heard of the price cycles for example flasks and such over the week, but glyphs are slightly diferent. They don't depend on buyers as much as sellers. Lets say a glyph sells for 35g, I then come and lower it to 17g(made up fallback, pretty realistic though) where I will get undercut by a certain amount of silvers/copper which results in the undercutter himself getting undercut ending up in a spiral of doom leading the price downwards until it finally reaches my threshold.
At this time I will jump out and stop posting that glyph while my competition stays and fight for the profit of this glyph. The prices then reaches the point where only one person is posting as he's below everyone else's threshold which will result in no glyphs of that kind on the AH once they expire/sell causing someone (maybe me) posting the glyph at their fallback making the whole thing start over again. So as you can see I'm giving sales away at those low prices while focusing on only selling glyphs at a good profit margin. I wont tell you my prices but in a worst case scenario I make about 400g/hour selling glyphs and a lot more normally. This is including the time spent posting, something I only do once or twice a day, maybe three if I have time (unlike Gevlon who is only posting once every 48 hours).
So far doing this has proven very effective and I really start to see Gevlons point even if I don't completly agree with it.