Monday, April 4, 2011

Point of Diminishing Returns

The point of Diminishing Returns is an economic concept that, like opportunity cost, has facinated me for quite some time. The simplest explanation I can give for it is that at some point, additional input won't really change the output in any meaningful way. To illustrate, think of workers making widgets (I still have no idea what a "widget" is, so good luck actually visualizing this). Say it takes one worker 30 minutes to make one widget, two workers 15 minutes, three workers 10 minutes, four 7.5 minutes, and so on. At some point, adding one more worker doesn't substantially change how quickly widgets are made. Certainly the difference between one worker and two is huge (30min compared to 15min), but the difference between ten and eleven is quite small (3min compared to 2.73min). And of course, there's no guarantee that this model will hold true for each worker you add. There's probably some point where the widget just can't be made any faster, and adding workers will do absolutely nothing. The actual "point" where the returns start diminishing is going to be somewhat subjective. Shaving .27 min off the time it takes to make a widget will add up if you're making thousands of them and certainly be worth the cost of the extra worker, but there's going to be a point where the cost of that extra worker just won't justify what little time you'll save.

So what on earth does this have to do with gold making?

Making gold takes time, and as we all know, "time is money, friend!" Be it spending hours farming, hours camping your glyphs, hours spent shuffling Obsidium, or hours spent leveling alts and new professions, the more time you put in, the more gold you'll make. BUT there is going to be a point of diminishing returns in how much time you invest versus how much gold you make (i.e. the gold you make from an extra hour of work doesn't add significantly to how much gold you're already making in the time you're already spending).

This relies heavily on the somewhat tricky to calculate gold-per-hour (GPH). I've never quite understood GPH, simply because I'm never sure what to count as my time investment. Is it the time it takes for a sale? The time spent crafting? Should I factor in leveling my toon and leveling its professions? Likewise, I'm not even sure how to calculate the gold part! Is it the potential net-worth of what I've crafted, or what gold has actually found its way in to my pockets?

To keep it simple, I'm just going to say your GPH is the amount of gold you made at the end of the week divided by the time you spend on gold-related activities (excluding leveling of toons or professions). So literally, the gold you've made per hour spent making gold. You could do gold made per day over time spent per day, but per week will help smooth out any outliers. E.g. there are some days that I'll sell a full set of PvP armor and rake in 10k, but that doesn't happen every day. My "GPH" that day would be huge, but not reflective of the actual average over the week.

Ok, so on to the actual point of the article. When do diminishing returns start to factor in? I believe it will be different for everyone, but for me, it's usually flipping vendor recipes and pets. To illustrate this point, take the gold I normally make per week and compare it to the additional gold I'd make from doing a hypothetical Outlands Limited Supply Run like Sinshroud over at the Consortium Forums illustrated a few months ago. First off, I don't think his times are accurate, simply because he seems to do a lot of porting to and from Shattrath, and not all of us have the luxury of a high-level mage. I'm not certain how long it would take for everyone else to do a circuit of the Outlands, but I'd guess about 30 minutes (though as we shall see in a moment, it doesn't make that big of a difference). I'll assume the amount earned is about accurate, bumping it up to 1.6k just to be safe (and make up a bit for the additional time). So that's 1.6k earned for 30 min of work. Not bad. Now let's throw everything else in to the mix and see what happens.

Let's say I spend about two hours a day (akf time and all that included) on AH related stuff and I earn about 40k a week (If anyone's wondering, I think this is fairly accurate). So that's 14hrs a week (yikes!) for about 2,857 GPH. If I spend 30min on Sinshroud's route, I could increase my gold per week by 1.6k. So that's 41.6k a week for 14.5 hrs invested, or 2,869 GPH. That's only a 12 GPH increase. Even using Sin's original 15min for the route, I'm still only increasing my GPH by 62 gold. And to be honest, that's really not a whole lot.

The main argument against what I'm saying seems to be that doing the Outlands run still puts an extra 1.6k in my pocket and even if that doesn't seem like much now, it will accrue over time. E.g. ten weeks later I'll have 16k more from doing this than I would from not. To which my response is, ten weeks later doing what I normally do nets me 400k. That extra 16k seems rather irrelevant compared to 400k. And that's exactly the point of this article.

When I'm already making 40k a week, going out of my way even once a week for an extra 1.6k just doesn't seem worth it to me. That's my "point of diminishing return." This is not to belittle Sin's rather excellent guide, as for people just starting out on their gold making journey, an extra 1.6k a week is phenomenal! If you're used to making less than that with more time invested, this route is an amazing way to start increasing your capital. But for those already set in their ways with a steady income, this is simply one more thing to deal with, and for me, the extra 1.6k just is not worth it.

A secondary argument could be something along the lines of my interpretation of GPH. Sin's guide shows ~1.6k profit for ~15min of work. So that's 6.4k GPH! Well, no, it's not. It's 1.6k profit for 15min or work. If you do the route three more times that day you're not going to make that 6.4k. You're going to be capped by what people are willing to buy. It could take a whole week for those items to sell.

To further illustrate, say it takes me one minute to craft a gem and sell it on the AH for 50g profit. Is that 3k GPH? No, that's 50 gold per minute. It'll only be 3k GPH if I actually spend an hour crafting and selling all the gems. Hence why I'm looking at the week average of time spent and gold earned. You can't take what you've earned in isolation and extrapolate your GPH from there; you have to add the total you earned and divide that by the time you actually expended to earn it. And that's not just crafting time. You should probably factor in AH scans, mailbox runs, and cancel/reposts as well, at the bare minimum.

Bringing this post back home, even gold making will have that point of diminishing returns, where the extra time invested doesn't really justify the extra gold earned. That extra gold just really isn't making that big of a difference. Again, what that point is is going to be subjective. For some, that extra 1.6k gold is a 50% boost to their weekly income. For others, it may barely be a 2% increase. And for some, maybe that 50% increase isn't worth it, while that 2% totally is.

P.S. I would like to reiterate that I am in no way belittling Sinshroud's guide (though I still think the total time spent is a bit off). I'm only using it for illustrative purposes as flipping vendor items is something I rarely do (for reasons I hope have become apparent in this post), and I do not believe there is any other guide out there that so thoroughly compares time spent with gold earned on flipping vendor items. Please, if you are having trouble making gold, go read this guide because it could very well help you on your path to riches.

As always, feel free to comment and disagree or agree with me on anything said. I'm sure there are points in this post where I'm unclear, misspoke, or assume too much, and hopefully they will come out in the comments so that I might clarify.


  1. Going to college is opportunity cost in my mind. Just like leveling a toon. Buying a warehouse or Guild banks and bags for alt banks.

    So when does the time to be factored into gold making begin, the moment you in ernest begin an activity exclusively for gold making... or the moment your level 1 toon is ready in capitol city of choice. When your level 1 opens the mail for the first time, i hope it is to get about 11 bags and a pile of gold to become effective. With gold in purse, bags in bank, bags on toon, NOW imo that toon is a viable money maker AH cancel posting beast.

    Likewise a crafter, untill they are what ever level you need for your crafting goals... when you stop doing that activity and begin only crafting posting... then you are making gold... leveling is not gold making, nor are dailies IMO. They are incidental income sources, but I do not figure them into and GPH calculation... thats mey thoughts on that...

    as to the rest... I feel you, there are very few vendor recipeies worth it to go so far out of the way to get. It is just not worth the flight time... i can not craft in flight, i can not mill or prospect in flight. I can AFK in flight, but if I go AFK why did I log on in the first place... (not a multi tasker type) that and i tried a round of items... and found them to be anoying on my server to try and sell. constant flipping and cancel posting. It would be better now with Zero Auctions, (did not have it at the time) but still... Yeah, great starter cash in this, and if you like THAT game, its a great continuing income source. A great income source with very little room to grow, or move or learn markets and be ready to play some AH PVP when you get three people on your server playing at it and undercutting values below vendor purchase prices to chase you out of the market.

  2. I'd like to throw another thing into the room before I'm gone again:

    When I decide to do an Outlands Limited Supply Run, I always use my Paladin, because he's got 310% flying speed + Crusader Aura, and he's also a gatherer.

    That way, it's close to the 15 min if I'm in a hurry.
    But the main point is that in 30 minutes, I'll have those limited supply recipes plus lots of ores and herbs in my back, which is much more profitable than the run without gathering professions.

    So, optimising the route can make the whole process much more profitable (or faster, respectively). Other than that, I agree with you.

  3. Except that your original example is incorrect. Two workers producing a widget in half the time is *NOT* diminishing returns. The widgets per hour of the second worker is the same as the first worker; nothing is diminishing.

    Your later examples are more the way I have seen diminishing returns being used. 1000g tends to mean a lot more when you have 1g on the realm than when you have 1000000g. Which is why spending 5 minutes to make a 5g profit on flipping a pet/recipe is bad for someone who is gold capped and great for someone with a net worth of 3g.

    And just like the business school concept of "cost of capital", the gold per minute does matter. If your goal is 2400g/hour then spending a minute making 25g hurts your progress toward 2400g/hr. Making 50g in a minute helps.

  4. Good article.

    Personally, I think any GPH analysis is mostly intellectual masturbation unless you consistently sell exactly X good Y times a week for Z profit. Two days ago I snagged BoE epic tanking shoulders from a guy in Trade Chat for 7,000g, then sold them on the AH for 21,000g in less than two hours. Was that 7,000g/hour (length of auction)? Was it 280,000g/hour (length of game interaction)? Should this (so far) one-time event impact any GPH calculation that incorporates my daily routine? What if the epic drop was from a mob instead?

    GPH only really works for daily quests, and is 100% irrelevant anywhere else. The sort of concept that GPH flirts around is real, e.g. using one's time efficiently, but it is a poor fit in practice.

  5. @Azuriel:

    I heartily disagree. GPH is a measure of efficiency and can generally be gauged in one way or another. Your analysis of the trade chat snag is flawed. Indeed, you should incorporate ALL of the time you spent in trade (assuming all you did was watch /2) within your GPH calculations. If you snag, on average, 1 item every 5 hours and profit 3k on average, your GPH from watching trade is 600.

    Applying the same flawed logic to dailies produces strange results: it took me 12 seconds to turn in a quest worth 27g. Does this mean my GPH is 8100? Of course not, considering that I had to kill mobs for 20 minutes, my GPH is 81 (plus gold from mobs, etc).

    The trade chat example is still ambiguous, I agree. But the purpose of GPH is to compare earning efficiency and allocate time in a wise fashion. This has been demonstrated on multiple occasions; shuffling herbs vs shuffling ore vs farming spots etc. I think GPH is not so much of an intangible concept as you make it out to be.