Why Your Big Bazaar Buy Order Won't Fill | Hypixel Skyblock

A Commerce Journey — Episode 11. Real economics, measured against live Hypixel Skyblock Bazaar data.

Economies of scale say that making more of something lowers the cost of each one. Factories, minions, bulk orders. Episode 10 ended on the promise of covering it, so I went to test the buying half: does purchasing in bulk get you a better price on the Bazaar?

The answer is no, and the reason is not the one I expected. It is also not the one I published in the first version of this post, which I had to take apart and rebuild after finding a mistake. That part is at the end.

1️⃣ Walking the book

When you press instant buy, the game does not charge you one price. It works down a list. The cheapest seller goes first, and when their stock runs out it takes the next cheapest, then the next, until your order is full. The price you pay is the average of everything it took.

Enchanted diamond has 1,283,293 units offered for sale in total. The API shows me the cheapest 30 price levels of that, holding 17,628 units, and the cheapest is 1,281 coins with 408 units behind it. Buy fewer than that and you pay exactly the top price. Buy more and you start eating into the level above.

That upward drift is called slippage, and in real markets it is the main cost of trading in size. A fund that wants a million shares cannot have them at today's price, because buying them moves the price while they buy. So I measured it across every actively traded item on the Bazaar, 900 of them with real weekly volume and a usable book.

Enchanted diamond costs you 0.00% extra at a hundred units and 0.01% at a thousand. At ten thousand it is 0.37%, which is still small. At a hundred thousand the line stops, and that is a limit of what I can see rather than of the market: the visible levels hold 17,628 units, so my measurement runs out before the book does.

Across all 900 items the median slippage at 100 units is 0.00%. At 1,000 units it is 0.00%. At 10,000 units it reaches 0.23%, and at 100,000 it reaches 2.41% for the minority of items that can fill an order that large at all.

For any order a normal player places, size is close to free.

2️⃣ Where the cost actually shows up

Look at the two lines that stop early. Super Compactor and Recombobulator do not slope upward, they simply end. The book has no more sellers, at any price.

That is the finding, and it took me a while to see because I was looking for the wrong thing. I expected big orders to cost more. On the Bazaar, big orders mostly cost the same, right up to the point where they become impossible.

The orange bars are buying. 99% of items can fill an order of 100 units. At 1,000 units, 86% still can. At 10,000, 43%. At 100,000 units, only 18% of the Bazaar can serve you at all, and the rest have nothing to sell you no matter how much you are willing to pay.

Recombobulator shows the shape of it. Its cheapest 30 levels hold 147 units between them, against 5,263 on offer across the whole book, so almost everything for sale sits well above the visible prices. Super Compactor is the same: 535 units across the cheapest 30 levels, 9,465 in total. The number of levels tells you almost nothing; the amount column tells you everything.

3️⃣ The book is shallower than it looks

The Bazaar interface hides all of this. You see one number, you type an amount, you press a button. Nothing on the screen tells you whether there are twenty units behind that price or twenty million.

This measures how many units you can buy before the price rises by one percent. The median item gives you 1,982 units of room. The deepest tenth gives you 49,215.

The shallow end is where it gets uncomfortable. A tenth of actively traded items move one percent within 103 units. That is less than two stacks. If you are buying materials for a large craft on one of those items, you are moving the price against yourself before you have finished the first purchase.

Key characteristics of a thin book:

• Few units per level, often one seller holding most of the stock
• Large gaps between one level and the next
• Total stock across the whole book smaller than one serious order
• The displayed price looks identical to a deep item's price

4️⃣ Selling is the easier half

Everything above is about buying. I ran the same test on the other side, and it came out the opposite way to what I assumed.

The blue bars on the fillability chart are selling. At 10,000 units, 43% of items can supply you but 59% can absorb you. At 100,000 units it is 18% against 33%. Dumping is easier than accumulating, on most of the Bazaar, at every size I tested.

The price you give up is smaller too. Selling 100,000 units costs a median of 0.28% below the top buy order, against 2.41% above the cheapest offer to buy the same quantity.

The reason is in how the two sides are built. A buy order sits there until someone fills it, so lowball orders accumulate at every price level going down, and the book gets thicker over time. A sell offer gets eaten the moment it is the cheapest thing available. One side collects, the other side drains.

You can see it on a single item. Enchanted diamond has 1,283,293 units offered for sale across its whole book against 19,647,089 units of standing buy orders, a ratio of about 15 to one. Enchanted cobblestone has 442,886 offered against 3,890,225 wanted. In both cases the demand side is the larger pile, and it is larger because it has been accumulating rather than being consumed.

This matters for anyone planning a large operation. The instinct is to worry about selling your output. The data says worry about sourcing your input.

5️⃣ Two items at opposite ends

Essence Crimson sits at 1,061 coins and has 3,311,015 units available before the price moves one percent. You could place an enormous order and barely disturb it.

Quickdraw Garden Chip has 1 unit of room. One. Buying two moves the market. Its price is displayed with exactly the same confidence as the first item's, and there is nothing in the interface to tell them apart.

6️⃣ The spread does not tell you this

Episode 10 spent a lot of time on the gap between the buy and sell price, and the obvious next thought is that a wide gap warns you about a thin book. It would be convenient, because the gap is on screen and depth is not. So I checked it across the 873 items where both numbers exist.

It does not work. The correlation between the gap and the depth of the book is +0.048, which is close enough to nothing. Sorting items into four groups by gap gives median depths of 3,489, 1,375, 1,666 and 2,659 units, which does not even move in one direction.

What does predict depth is weekly volume, at +0.606, and unit price, at -0.488. Busy items have deep books, expensive items have shallow ones, and both are more useful than the number the interface puts in front of you.

7️⃣ The mistake

The first version of this post said selling was the harder side. It was wrong, and the reason is worth more than the finding.

The Hypixel API returns two lists per item. One is called buy and one is called sell, and I assumed they meant what a player means by those words. They do not. They are named from the market's point of view, so the list called buy holds the buy orders, which are what you sell into, and the list called sell holds the sell offers, which are what you buy from. I had them backwards, so every number in the original post described the wrong side of the market.

What caught it was a sanity check I should have run first. The API also reports a single current buy price and sell price, and those two numbers have to match the top of the two books. I checked across 1,784 products. One direction matched to three decimal places. The other was off by sixty-seven percent.

Ten lines of code, and it would have saved the rewrite. Any time you pull data from a source you did not build, find two fields that must agree and check that they do before you build anything on top.

8️⃣ What it means when you place an order

Three practical consequences follow.

Stop worrying about slippage on ordinary purchases. If you are buying a few thousand of something with real volume, the price you see is the price you get. Splitting the order into chunks is a habit borrowed from real trading that does nothing here.

Check the book before any large purchase, and the number to check is not the price. It is the amount sitting at the first few levels. If the top level holds 200 units and you want 5,000, you are going to walk a long way down that list.

Plan your inputs harder than your outputs. Roughly a third of the Bazaar can absorb a sale of 100,000 units, and under a fifth can supply a purchase of the same size.

Real-world parallel: this is liquidity risk, and it is why large funds cannot trade small companies. A fund managing billions cannot take a meaningful position in a company whose shares barely trade, because getting in is slow and getting out is worse. The position is not expensive, it is unexitable, and every serious institution has size limits written specifically to avoid it.

9️⃣ What I could not measure

The order book is capped at thirty levels a side, and I badly underestimated how much that hides. The API separately reports whole-book totals, and comparing the two shows the visible levels are a small fraction of the real thing: enchanted diamond shows 17,628 units across its cheapest levels and holds 1,283,293 in total.

So every fillability figure above is a floor rather than a count. An item I have recorded as unable to supply 100,000 units may well be able to, at a price further up the book than the API will show me. The direction of the finding survives, because the comparison between buying and selling uses the same truncated window on both sides, but the levels themselves should be read as the share of items that can fill an order within the visible book.

Everything here is one snapshot, taken on 2026-09-05. Books refill. An item that cannot fill 10,000 units this minute may manage it an hour later when more sellers have posted. The snapshot shows the shape of the problem, not a permanent property of any single item.

Go and open the order book on something you buy regularly and look at the amount column rather than the price. It takes ten seconds and it decides whether your next big purchase works.

Coming next: Diminishing returns. Why the second compression step on a minion usually destroys value, and how to tell before you buy the upgrade.

Order books and prices from the Hypixel Bazaar API on 2026-09-05, covering 900 items with meaningful weekly volume. The API returns up to thirty price levels a side.