Double-Compacting Your Minion Burns Coins | Hypixel Skyblock

A Commerce Journey — Episode 17. Real economics, measured against live Hypixel Skyblock Bazaar data.

You bought a Super Compactor. It cost 217,051 coins, it took a while to afford, and it sits in your minion turning raw material into enchanted material. Good purchase.

Then you add a second one, so the enchanted material becomes blocks. And this is where a great many players, including me for a long stretch, quietly set money on fire.

1️⃣ The idea

A sunk cost is money already spent that you cannot get back. The rule about it is short: it should have no influence on what you do next. Only future costs and future benefits belong in a decision.

Everybody agrees with this when it is stated abstractly and almost nobody follows it, because it feels like waste. You paid for the thing, so you should use the thing. That instinct is the fallacy.

It is worth being precise about why the instinct is wrong rather than merely inconvenient. The coins are gone in both futures. Whether you compress or not, your balance already reflects the purchase, so the payment cannot distinguish between the two options and therefore carries no information about which is better. Treating it as a reason is treating a constant as a variable.

The compactor is a clean example because it separates two decisions that players merge into one.

The first decision is whether to buy it. That is forward-looking and it is easy: Episode 12 measured the first compression at a median of about 2.3 times the value of the raw material, so the device pays for itself quickly on almost anything.

The second decision is whether to run the output through it again. That decision has nothing to do with the 217,051 coins, which are gone either way. It depends only on whether a block is worth more than the 160 enchanted items inside it.

2️⃣ Usually it is not

On coal, 160 enchanted coal are worth 211,064 coins and the block they make is worth 25,000. Every block costs its maker 186,064 coins. Redstone costs 71,738 a block.

Nothing physical is lost. The block contains every unit that went into it. What changed is that you moved the goods into a market where far fewer people are shopping, which was Episode 12's finding and remains the mechanism here.

The player doing this is not confused about the prices. They usually have not looked, because the question feels settled. They own a compactor, so compacting is what the compactor is for.

3️⃣ The size of the mistake

Weekly trading volume lets you put a number on how much of this is happening.

Across the 5 chains where the second compression loses money, the weekly block volume multiplied by the loss per block comes to 42.7 billion coins. Redstone alone accounts for 17.5 billion, coal for 17.0 billion.

That figure is large enough to be worth restating plainly. Somewhere between the minions and the Bazaar, players are destroying roughly 43 billion coins of value a week by performing a craft they already own the machine for. It would buy about 196,712 more compactors, which is the joke.

4️⃣ Measured in compactors

There is a way of stating this that made it land properly for me.

The machine cost 217,051 coins. One enchanted coal block loses 186,064 coins, so a single block throws away 86% of what the compactor cost. Make 1.2 coal blocks and you have destroyed more value than the device is worth. Redstone takes 3.0 blocks to do the same.

A minion running unattended produces blocks at that rate for as long as you leave it. The purchase was a one-off decision you thought about carefully. The compressing is a default that repeats every time the minion fills, and the second one is far more expensive than the first.

That asymmetry is the practical heart of sunk cost. People deliberate hard over the payment they can see and never revisit the recurring choice it set in motion, even though the recurring choice is where nearly all the money is.

5️⃣ The equipment makes the argument

Worth asking why this particular mistake is so common, because it is not simply carelessness.

A compactor in a minion is not a suggestion you evaluate each time. You install it once and it acts on every unit produced from then on, silently, without ever showing you a price. There is no moment at which the game asks whether today's numbers still justify it, and no screen anywhere that compares 160 enchanted items to the block they would become.

So the fallacy is not really being committed by the player each time. It was committed once, at installation, and then automated. Anything that turns a judgement into a default will do this, which is why the useful habit is not resolving to think clearly but putting a recurring check on the calendar.

6️⃣ Three chains where it is right

The chart has green bars, and they matter more than the red ones for anyone trying to learn a rule from this.

Quartz, diamond and gold all gain from the second compression. Quartz gains 52,619 coins a block and 8.5 billion coins a week across the server. Netting everything off, the eight chains together lose about 28.1 billion rather than 42.7.

So the correct instruction is not to stop compressing. It is to make the decision per item, on current prices, every time. A blanket rule in either direction is wrong on at least three of these eight chains, and the rule that happens to be wrong most often is the one your equipment encourages.

Key characteristics of a sunk cost trap:

• A payment already made that cannot be recovered
• A follow-on choice that feels like it honours or wastes that payment
• A forward-looking calculation that nobody runs because the matter feels settled
• Equipment, habit or identity that makes one answer the default

7️⃣ A prediction I got wrong

While working on this I expected to find a related effect and went looking for it.

The reasoning was that players who bought an item at a high price would refuse to sell it after the price fell, because selling makes the loss real. If that were happening at any scale, items whose price had dropped over the past week should show unusually few units offered for sale relative to units wanted.

There is no such pattern. Across 535 items with reliable price history, the correlation between the week's price change and the ratio of offers to orders is +0.07, which is nothing at all.

I am reporting it because a null result is still a result, and because 535 items is a small sample that could easily hide a real effect. What it does rule out is anything large and obvious. If holders are clinging to their losers, they are not doing it hard enough to bend the order book.

One thing the chart does show, incidentally, is that the ratio sits well below one almost everywhere. The median item has 0.31 units offered for every unit wanted, which is Episode 11's asymmetry turning up again in a different measurement.

8️⃣ The opposite error

Sunk cost has a mirror image that gets less attention and costs real money too.

Having learned to ignore what you paid, it is easy to overcorrect and abandon something that is still worth running. A minion that cost a fortune and now produces a cheap item is not automatically a mistake to keep. If the coins it earns exceed what you would earn using that slot for something else, keep it, and the purchase price is as irrelevant to that decision as it was to the first one.

The rule cuts both ways. Ignore the past, price the future, and accept that sometimes the future says carry on.

The two errors also feel completely different from the inside, which is what makes them hard to catch. Carrying on with something expensive feels like commitment. Abandoning something expensive feels like discipline. Neither feeling contains any information about whether the next hour of that minion pays, and both arrive well before you have looked at a price.

9️⃣ The check itself

The whole calculation takes about twenty seconds per item, so there is no real excuse for not running it.

Look up your enchanted item, multiply its instant-sell price by 160, and compare that to the instant-sell price of the block. If the block is lower, take the compactor out. If it is higher, leave it in. That is the entire decision, and the price you paid for the machine appears nowhere in it.

Do it monthly rather than once. Every number in this episode is a snapshot, and three of these eight chains sit close enough to the line that an ordinary week of price movement could flip them. A rule you worked out carefully last season is a sunk cost of its own kind.

Real-world parallel: Hal Arkes and Catherine Blumer demonstrated this in 1985 in The Psychology of Sunk Cost, including a famous study where people who had paid more for a theatre subscription attended more performances in bad weather, having got no extra value from doing so. The largest version is usually called the Concorde fallacy, after the British and French governments continued funding a supersonic airliner for years past the point where anyone expected it to return its costs, on the grounds that too much had already been spent to stop.

10️⃣ What I could not measure

The weekly figures use the volume of blocks traded, which is not the same as the volume produced. Some of those blocks change hands more than once, and a player who crafts a block and keeps it never appears in the number at all. The direction is right and the magnitude should be treated as an estimate rather than a count.

I also cannot see intent. A player compressing coal into blocks may be doing it for storage rather than sale, since a block takes one inventory slot and 25,600 coal do not. That is a genuine reason and it is invisible in trading data, so some share of what I have called a mistake is a considered choice about space.

Open the compactor decision for your own main item tonight. Multiply the enchanted price by 160, compare it to the block price, and do it on today's numbers rather than the ones that were true when you bought the machine.

Coming next: Fixed cost and minimum viable scale. Why some crafts only work in bulk, and how to find the volume below which a recipe is not worth starting.

Prices and volumes from the Hypixel Bazaar API on 2026-09-05, with price history from Coflnet. Weekly loss figures use traded block volume as a proxy for production, which is discussed in the last section.