Skyblock Craft Flipping: the Same Craft, -42% or +63%

A Commerce Journey — Episode 10. Real economics, measured against live Hypixel Skyblock Bazaar data.

Episode 9 was about who controls prices. This one is about what it costs to make the thing being priced, which is where most economics courses start and where I probably should have started too.

I was going to write the standard version of this post. Fixed costs, variable costs, a worked example, done. Then I pulled the Bazaar API on 2026-09-04 to get real numbers for the example, and the example turned out to be more interesting than the lesson I had planned around it.

Here is the test. Take the simplest craft in Skyblock: 160 raw blocks into one enchanted block. Price it two ways. First as an impatient player, using instant buy for the materials and instant sell for the result. Then as a patient one, placing a buy order for the materials and a sell offer for the result.

Same recipe. Same items. Same minute.

The red bars are the impatient version. Buying 160 diamonds with instant buy costs 1,969 coins. Instant selling the enchanted diamond you make from them returns 1,141. You have destroyed 42% of your money by pressing two buttons. Cobblestone loses 50%, redstone 39%.

The green bars are the patient version. Buy order for the diamonds at 788, sell offer for the enchanted diamond at 1,286. The same craft now returns +63%. Cobblestone goes to +278%.

Nothing about the recipe changed between those two lines. What changed was whether you were willing to wait.

1️⃣ What instant buy actually does

Most players treat instant buy as a button that gets you an item. It helps to know what it does underneath.

The Bazaar is two queues. On one side, people who want to buy have placed orders saying how much they will pay. On the other, people who want to sell have placed offers saying what they will accept. The highest buy order and the lowest sell offer are the two numbers you see.

Those two numbers are never the same. There is always a gap. When you press instant buy, you are not creating a new price, you are taking the lowest sell offer that already existed. When you press instant sell, you are taking the highest buy order that already existed.

So both times you cross the gap, and both times the person on the other side keeps it. Do it twice in one craft and you pay the gap twice.

Key characteristics of the two sides:

• Instant buy and instant sell get you filled now, at a worse price
• Orders and offers get you a better price, at an unknown time
• The gap between the two is the fee for going first
• Nobody charges that fee. It is just what the other side gets for waiting

2️⃣ Three costs, and only one of them is on the screen

The textbook version of this episode splits costs into two kinds. Variable costs scale with how much you make, so the 160 diamonds are a variable cost. Fixed costs are paid once no matter how much you produce, so minion slots, the Super Compactor and the island upgrade are fixed costs.

There is a third one that does not appear on any screen in the game. Opportunity cost is what you gave up by doing this instead of the next best thing. If crafting made you 50,000 coins in an hour and a dungeon run would have made you 400,000, the craft cost you 350,000 coins even though it showed a profit.

Accounting profit subtracts the first two. Economic profit subtracts all three. Only the second one tells you whether the hour was worth spending, and it is the number almost nobody calculates.

That +63% looks strong until you ask how long the orders took to fill. If it was four minutes, good. If it was four hours, you made 497 coins in four hours and a fishing session would have beaten it. Episode 13 is entirely about this, so I will leave it there.

3️⃣ How common is a gap that big?

Three items is not evidence. So I took every product on the Bazaar with real weekly volume — 1,245 of them — and measured the gap on all of them.

The median is 33%. Half of everything traded on the Bazaar has a gap wider than a third of its own price.

For comparison, a share in a large company on a real stock exchange typically trades with a gap of well under one percent, often a few hundredths of a percent. The Bazaar is running spreads roughly a hundred times wider than that, all day, on thousands of items, and almost nobody notices because the buttons are right next to each other.

4️⃣ I was wrong about why

When I first saw the six-item version of this, I assumed the gap tracked how heavily an item was traded. Busy item, lots of competing traders, tight gap. Quiet item, nobody competing, wide gap. That is the standard explanation and it sounds right.

It is not what the data says. Across all 1,245 products, the correlation between the gap and weekly trading volume is +0.011, which is another way of saying there is no relationship at all. Sorting items into ten groups by volume gives a median gap between 25% and 45% in every single group, with the busiest group no better than the quietest.

What does predict it is price. The correlation between the gap and unit price is -0.421.

The cheapest tenth of items, median price 24 coins, runs a 76% gap. The most expensive tenth, median price 5,039,764 coins, runs 11%. The middle is noisy and does not fall in a neat line, so I am not going to pretend it does, but the two ends are far apart and the direction is consistent.

I also checked whether this was volume in disguise, since expensive items might just be traded more. It is not. Looking only at the busier half of the Bazaar, price still predicts the gap at -0.490 while volume sits at -0.042.

Why price and not volume is a question this data cannot fully answer. My best guess is that a trader needs a minimum absolute profit per unit to bother, and on a 24-coin item that minimum is most of the price, while on a five-million-coin item it disappears into rounding. That fits the shape, but I have not proved it, and I would rather say so than dress a guess up as a finding.

5️⃣ The same craft, one hundred times

Percentages hide how quickly this compounds, so here is the arithmetic on a run of a hundred diamond crafts. Nothing complicated, just the two numbers from the chart multiplied out.

Impatient, each craft costs 1,969 and returns 1,141, losing 828 coins. Run it a hundred times and you are down 82,757 coins. You have spent an evening farming, crafted a hundred items, and finished poorer than when you started, which is the outcome most players mean when they say crafting is dead.

Patient, each craft costs 788 and returns 1,286, making 497. A hundred of those is 49,748 coins. The distance between the two runs is 132,505 coins, and every coin of it came from order placement rather than anything you did in the world.

There is a catch in the patient column that the percentage does not show. Running a hundred crafts on orders means having 78,814 coins sitting in buy orders at once, unavailable for anything else until they fill. The impatient player needs the coins for about four seconds. The patient one needs them for as long as the market takes.

That is the real trade being made. Not effort against reward, but liquidity against margin. A player with 50,000 coins and a player with 5,000,000 coins are playing different games at the same Bazaar, and the second one can afford to be patient in a way the first cannot.

6️⃣ What patience actually costs

The green bars are not free money, and the reason is that an order is a promise to trade, not a trade.

When you place a buy order below the current price, you are betting the price comes down to you. Often it does. Sometimes the item runs away upward and your order sits unfilled for a day while you watch the thing you wanted get more expensive. Meanwhile your coins are locked in that order and cannot be used for anything else.

That is a real cost, and it is why the gap exists in the first place. The patient trader is being paid for taking on two risks the impatient one avoids: the risk of never filling, and the risk of holding an item whose price moves while they hold it.

Real-world parallel: this is one of the oldest ideas in market economics. Harold Demsetz set it out in The Cost of Transacting, published in the Quarterly Journal of Economics in 1968, and his argument was that the bid-ask spread is not a fee or a tax but the price of immediacy — what you pay someone else for being ready to trade the moment you want to. Every market maker on every exchange since has been doing the same job as a Skyblock flipper sitting on a buy order.

7️⃣ A number I threw out

Lapis was in my first version of the craft chart. The API said 160 lapis cost 6,365 coins at instant buy while enchanted lapis sold for 798. That is an 87% loss, far worse than anything else, and it would have made a dramatic bar.

I cut it, because the number was not real. Raw lapis had 9 buy orders on the entire Bazaar at that moment. Diamond had 249. When a price is built from nine orders, one person quoting far away from everyone else moves the whole number, and what you are measuring is that person rather than the market.

This matters beyond lapis. Any time you look up a price for a quiet item, you are reading a number that a handful of people decided. The Bazaar shows it with the same confidence it shows the diamond price, and there is nothing on screen to tell you which is which. Checking the order count first is a habit worth building, and Episode 23 goes into it properly.

What to do with this

Three things follow from the numbers above.

• Price every craft both ways before you run it. The recipe does not tell you whether it is profitable, the two order books do
• Assume the gap is the cost of the trade, not a display quirk. On a median Bazaar item you are paying 33% for being in a hurry
• On cheap items, the gap is large enough that instant buying materials will usually destroy the whole margin before you have crafted anything

Pick one craft you actually run and price it both ways tonight. It takes about two minutes and it is more convincing than this post.

Coming next: Economies of scale, tested on the buying side. Whether purchasing in bulk gets you a better price, and what the order book will and will not give you.

Every number here came from the Hypixel Bazaar API on 2026-09-04, covering 1,245 products with meaningful weekly volume. Prices move, so the coin values will be stale by the time you read this. The percentages and the shape are the part that holds.