Thursday, May 19, 2011

The G/Hour Misconception

One thing that really bothers me is when people talk about gold per hour. That is, the amount of gold you get in a given time frame doing a specific set of tasks. You can also relate this theory, and it is only just a theory, to the 80/20 rule where 80% of your gold must come from only 20% of your effort. Unfortunately both of these theories are almost impossible to quantify with any degree of accuracy.

Mind you I am not in any way saying that guessing your G/hour is pointless or always going to be wrong or a piece of information that has no meaning. No, I'm only going to try today to clear up an extremely popular misconception about the whole idea. One which just about everybody in the AH game likes to either ignore it or simply take g/hour as the word of god.

The only exception to g/h being immeasurable is farming a very specific group of mobs and skinning while alone or something along the lines of the shuffle when you vendor absolutely everything. You spend X time prospecting, Y time cutting gems, and Z time vendoring. Each one of those steps has an exact value attached to it which you can easily multiply and add accordingly. Hmmm, but what if you AH the rare gems? What if you like to DE carnelian spikes and turn them into scrolls to sell? Now it's suddenly impossible to calculate with any degree of certainty what so ever. This is where the problem comes in.

Let me give you the typical fallacy of the g/h myth. Actually it's more of a misuse of the theory than it is a misconception is certain cases. But that's just semantics, moving on. Here's the example: buy a widget from a vendor for 1g (copper rod, local cooking recipe, etc.) and place it on auction for 10g. That has taken you about 30 seconds to do so far. The items then sells for 11g, which gives you a 10g profit. What the typical person does here is instantly jumps forward in time to "omg I just made 10g in only 30 seconds! That means my gold per hours is 1200! Go me I'm gonna be rich tomorrow!"

Ok here's the first problem in why this is entirely wrong. You're completely skipping the 38 hours it took for that copper rod to sell. so rounding up here, that means that your g/hour from selling copper rods is somewhere in the range of... less that a friggin copper. Should you still sell copper rods? Of course. Is it great gold per hour? Obviously not.

The second problem here is another that is so staggeringly obvious that I'm shocked nobody has mentioned it before. These copper rods, these vendor recipes, these pets, you don't buy them from one vendor for 5g and sell them to another vendor for 20. You sell them to a person. A person who is either a moron, lazy, uninformed, or plain doesn't care what they spend on what. That person is not always there, the vendor that buys your common gems for 5g is always there.

So not only do these pro tips of slow selling items not make even a fraction of the gold per hour as they claim, they don't always have a person to sell to. Once you take these two things into account you'll have a much better chance of improving your actual gold per hours. Look a little bit deeper and you'll see that this whole time I've been repeating another one of my many AH motto's. Know your market. Know your customer.

To really make the g/hour theory applicable to anything is to add in one other part to the equation: repetition. If you get a world boe from an instance that you sell for 20k, your g/h is NOT 20k per dungeon run. So maybe make it like (gold / hour) / T where T is how much time you'd have to spend so that it can be repeated.

In the end, what you should be doing is what I keep telling you to do every Tuesday, look at the bottom lines first. Take the amount of time total that you spend working the AH, crafting, posting, collecting and everything else that goes into it. Tally that all up and then total up all of the gold you made for that day or week and put them together. That is what your real gold per hour is. Is that easy to do? No not really actually, I know because I've done it. But is it accurate? It is as honest about your g/h as you are. So when somebody claims an amazingly simple method of 600g/hour be wary as it is practically guaranteed that it isn't consistent.

Thanks for stopping by!


  1. Good post, although it's subjective how one figures their gold/hr compared to another. Someone who can only play an hour/day and then has RL things to do the other 23 hours would find the 'sell an item that takes a day to sell' of equal value to having received that gold instantly.

    I agree though, people do need to be aware of how gold/hour relates to each gold making strategy in blogs.

  2. Awesome stuff, Stok. I addressed this a bit in an earlier post, mainly aimed at "limited supply runs." The problem with the GPH figure from those (15min to make 1.2k, that's 4.8k gph!) was that it's not a sustainable figure. And, like you pointed out, you can't extrapolate from an isolated incident. I think I ended up reaching the same conclusion as you: take the total gold you made in a week and divide it by the total number of hours you spent on making that gold (e.g. AH scans, prospecting, crafting, cancel/relisting, barking in trade). That's going to give a much more accurate account of your GPH, or rather, gold per hour of effort. It's a much more useful calculation too, as you might be able to find ways to streamline your process to shorten the time you spend "making" gold while still earning the same amount.

  3. GPH is a flawed comparison tool, however - it is currently the best we have available. When used properly, GPH can help you to make comparison decisions about what would be the most effective use of your time.

    GPH figures shouldn't be used in order to extrapolate your way to becoming rich and famous, but you CAN use it in order to more effectively prioritize your time and make sure your routine is as efficient as possible.

    I would draw an example from yesterday's post, where you mentioned that +5 stats IS worth 2K because of the way that it adds up. GPH can work in much the same way. An extra 100 GPH might seem like a silly thing to worry about, but the point is the same: if you maximize the way you spend your time, it WILL add up, and GPH can be an effective way to ensure you are maximizing your time.

    I'm not sure if you have ever been to my site, Break the Bank, but I have been theorizing a "better" comparison tool for awhile now. It is currently not moving forward because it requires accurate, large-scale sales data in order to be fully effective, but I would appreciate any feedback on the concept itself.

  4. GPH is a single vector model with only a double input. It has great POTENTIAL to be greatly misleading, and Great SIMPLICITY for fast evaluation.

    ONLY YOU CAN VALUATE YOUR TIME... (and your pixels)
    If GPH is your tool great!
    If a more complex model and tool is your preference... roger!
    If you dont care, you just play, have fun and you think have more gold now than you did last time you looked... woot! you win!

    The person in the mirror has all the answers, the writers in the blog-o-sphere have perspective, fresh views, tips, knowledge and more to help speed up the conversation you have with that person in the mirror. You do talk to yourself? right?

  5. GPH for flipping a copper rod is measurable and useful. Include the time it takes to log in and out, and you'll see that it is not worth doing by itself.