How these numbers are calculated

For every tracked Path of Exile boss we answer two separate questions. What does it pay on average? — profit per run and per hour. And what does it take to actually get there? — how much of that profit depends on a rare drop, how much currency you need to survive a bad streak, and how many runs before you are reliably ahead.

The second question is the one most profit tables skip, and it is where people lose currency: a boss can show a healthy profit per hour and still bankrupt a small stack.

Each drop contributes its current market price multiplied by its drop chance and quantity. Summing those gives the expected drop value of one run. Subtracting the market cost of every entry requirement — fragments, invitations, keys — gives profit per run. Profit per hour scales that by the boss's average run time.

All of this is computed in the database from the current price of every item, so the whole table moves together when the market does.

A drop qualifies as a lucky drop only when it is both:

• Rare — under a 5% chance per run, and
• Material — worth at least 15% of everything the boss is expected to drop.

Both conditions matter. Rarity alone is badly misleading: The King in the Mists has more than twenty drops under 5%, but almost all of them are beasts worth a few chaos. Counting those as jackpots would suggest a payoff every couple of dozen runs when the real figure is in the hundreds — and would understate the budget by an order of magnitude.

Floor profit is what a run earns if the lucky drop never lands. The tier follows from it:

• Low — floor profit is positive. The reliable drops already cover the entry cost; the lucky drop is pure upside.
• Medium — the boss loses money without its lucky drop, but either the shortfall is shallow (at most 25% of the entry cost per run) or the lucky drop lands often enough that dry streaks stay short.
• High — a deep shortfall and a rare payoff. These are the bosses that need a real budget behind them.

This one is simulated, not derived from a formula. We run 10,000 simulated farmers, each fighting the boss up to 1,000 times, using its real drop table — including drop pools, where only one item from a group can drop per kill, which no closed-form estimate can represent.

The recommended budget is the 90th percentile of what those farmers needed: enough that 90 out of 100 never go bust from their starting stack. One in ten will still hit a worse streak.

What it is not: this is not peak-to-trough drawdown. It measures how far below your starting point you fall, not how much of your winnings you give back after a good run — the latter is a much larger number and would badly overstate what you need to begin.

We used to estimate this with a closed-form approximation. It landed anywhere from 0.6× to 43× the simulated value, worst on cheap bosses whose variance comes from a rare upside drop, so it was replaced outright.

From the same simulation: the first run count at which 90% of farmers are genuinely ahead. It is evaluated at every run, so a boss that turns reliable after four runs reports four.

Some bosses never reach it inside 1,000runs. Those are shown as “>1,000” — profitable on average, but too dependent on a rare drop to be dependable over a normal league. A boss that loses currency on average shows “—” instead, because no number of runs fixes it.

The share of individual runs whose loot is worth less than the entry cost. High values are normal and not a warning by themselves — a boss can lose currency on most runs and still be strongly profitable, provided you can afford to reach the runs that pay.

Item and currency prices come from poe.ninja and are scraped daily. Every profit, risk and budget figure on the site is recalculated from those prices, and the simulated figures are recomputed immediately afterwards, so the risk numbers always describe the same prices you see in the table.

Prices are league-specific. Drop rates and entry requirements are maintained by hand against community data and patch notes.

• Drop rates are community estimates. Grinding Gear Games does not publish them. Where the community is uncertain, so are we — and a wrong drop chance moves the profit and the risk tier together.
• Prices lag the market. Values reflect the most recent scrape, so a sharp move can leave a figure stale for up to a day. Thin-margin bosses are the most sensitive: a small price move can swing the recommended budget by an order of magnitude.
• Averages describe populations, not your evening. Every figure here is what happens across thousands of runs. None of it predicts your next one.
• Run times are averages. Profit per hour assumes the listed run time; your own clear speed will differ, and each boss page lets you set your own.

Expected drop value

The average value of a single run’s loot: every drop’s market price multiplied by its drop chance and quantity, summed. It is a long-run average, not what any single run pays.

Profit per run

Expected drop value minus the cost of everything needed to open the fight — fragments, invitations and keys, priced at current market value.

Lucky drop

A drop that is both rare (under a 5% chance per run) and material (worth at least 15% of everything the boss is expected to drop). Rarity alone does not qualify.

Floor profit

Profit per run if the lucky drop never lands. A boss with positive floor profit pays for itself from its reliable drops alone.

Risk tier

Low, Medium or High, describing how much of the boss’s profit depends on a rare drop rather than on reliable income.

Recommended starting budget

The currency to have in hand before starting, so a cold streak does not leave you unable to open the next fight. Taken from the 90th percentile of simulated outcomes.

Runs to reliable profit

The number of runs after which 90% of simulated farmers are actually ahead — as opposed to being ahead only on paper, on average.

Runs at a loss

The share of individual runs whose loot is worth less than the entry cost. High values are normal for bosses whose value sits in rare drops.