Frequently asked questions

One player wants something from the in-game shop but cannot, or would rather not, pay for it. Another player buys it for them with real money — an in-app purchase, IAP in most trade posts — and gets hearts in return, sent day after day.

The hearts are the ordinary ones players give each other as friends. Not the Seasonal Hearts, which are bought from the seasonal spirits with seasonal candles and need a Season Pass to reach at all: those are spent with the seasonal guide, and they are not what gets sent back here.

The item arrives at once. The hearts take weeks, sometimes months. That gap is the whole difficulty of this kind of deal, and it is what this site exists to price.

No. It works out a figure, and nothing more. It sells nothing, arranges no trade, introduces nobody and takes no commission.

It exists to prevent two things: one player farming for months on a lopsided trade, and the other fronting money they never see again. Whatever happens after the number appears happens between two players, without it.

No. Unofficial fan project, unconnected to thatgamecompany.

Sky: Children of the Light and thatgamecompany are trademarks of thatgamecompany, Inc., used here only to identify the game; this site is not endorsed by or associated with them.

No game material is reused here; the flags come from a third-party icon set, credited in the repository.

In two steps. First a base rate in hearts per euro, which depends on the amount alone. Then that rate is adjusted by whatever changes the risk and the delay.

The base rate falls smoothly: there are no tiers, so one more cent never makes the price jump. The result is kept inside a band, and the braking is gradual rather than a wall: no lever ever stops counting in the calculation. The price is a whole number of hearts, though, so hard against an edge one more vouch can be worth less than the heart it would take to show.

The coefficients are not published. What is, and what every result shows, is the direction and strength of each lever.

Because that is where the risk sits. Whoever fronts the money has no recourse: no escrow, no guarantee, no way to compel anyone. If the other player stops after three weeks, the money is gone.

A first time together therefore asks for more hearts than a familiar face, and a familiar face more than a close friend; vouches you have verified ask for fewer still. It is not a judgement on the other player: it is the price of uncertainty, and whoever fronts the money is the one carrying it.

Because it decides the duration, and the duration carries the risk: the longer a trade runs, the more chances it has to break.

The calculation uses the true average per calendar day rather than a good day: six hearts three days a week is slower than three hearts every day. Past a certain pace, going faster barely moves the price — the risk has already stopped dropping.

Two things pull against each other here. Setup effort does not scale with the amount — a sixty-euro trade is not six times the work of a ten-euro one, it is the same conversation, the same purchase, the same follow-up. But a larger amount implies a longer trade, and duration carries the risk.

The effort term wins at small amounts and then fades, which is why the improvement flattens out and barely moves at all on the largest orders. Duration keeps acting through the sending pace.

The custom is that the buyer pays first and the hearts follow. This setting reverses the order: the other player sends for a number of days before the sum is actually spent. That reduces what is exposed if the arrangement stops, and the price follows.

For pricing it is capped at the length of the trade: asking for more days than the trade runs does not lower the figure any further, and the result says so when that happens. The shared link and the preview card state the number that was asked for, which can be longer. When it is not zero, the hearts being sent cover a purchase that has not happened yet.

About 70 hearts for 10 €. That is the anchor the whole site is calibrated on, and it comes from the Season Pass: it is the most common purchase in this kind of trade, and the community has largely settled around that figure. The anchor follows the amount rather than the item, since what a Pass costs differs a little from one regional store to the next.

It holds for that amount only: the per-heart rate improves on larger orders and worsens on very small ones, and the rest of the calculation moves it further according to risk and delay. Talking about a single rate for the site would mean nothing.

The anchor is a convention, not a result. What the calculation adds is consistency and direction — the same adjustments for everyone, in the same sense. It does not demonstrate that this level is the right one. Anyone who thinks the convention is wrong is arguing with the community, not with the arithmetic.

The amounts are in euros, and so is the anchor. Nothing is converted for you.

A shop price is not the same everywhere: it depends on the region of the account making the purchase, and on the platform it is bought from. Someone counting in another currency should convert the shop price of their own region into euros and enter that. Converting the number of hearts instead would be converting the wrong side — a heart is not money, and its rate here is a ratio, not an exchange rate.

That calculation comes up often in discussions, and it is correct. But a shop price is not a value: it is an asking price. That rate has no observed market — nothing is known to change hands at a euro a heart — so it is an asking price rather than a settled one.

The figure works as a ceiling: above it, whoever is paying has no reason to trade at all. A ceiling does not say where the price should sit underneath it.

Above all, matching it would mean that whoever fronts the money gains nothing by doing so. They pay in full, once, irreversibly, then wait months for hearts from someone nothing obliges to finish. A price only one side has a reason to accept is not a price two people would agree on.

The ratio between a sum and a number of hearts does not describe what a heart is worth. It describes a delay and a risk.

The shape is deferred payment: the item arrives at once, and is paid for over several months. Nobody says paying in instalments devalues the money of whoever gets the goods; it comes to more than cash because someone is fronting and waiting. The gap is measured here against the anchor, not against the shop price.

And a heart is not free, it is slow. A hundred of them cannot be had in a day, however long anyone plays. What is being paid for here is that calendar.

No, and presenting it that way would be dishonest. It is a defensible starting point, worked out the same way for everyone.

Two players remain free to agree on something else. The number is mostly there so a conversation starts from common ground rather than from an impression.

The money goes first and does not come back. The sending runs over days, often weeks, sometimes months. Nothing guarantees it is seen through, and there is no recourse if the other player stops partway: no refund, no arbitration, no way to take back an item already given.

The longer it runs, the greater the risk. A trade spread over five months assumes the other player stays present, willing and reachable for five months.

None of this is legal, financial or tax advice. The figure is produced as is, with no warranty, and nobody is answerable for a trade made on the strength of it.

The thatgamecompany terms govern exchanges of in-game items outside the game. A deal of this kind may go against them and put an account at risk, on both sides.

This site encourages nobody to make one. It works out a price for those who have already decided to.

Sharing the result sends the figures a conversation needs: the amount, the number of hearts, the duration, the sending pace, any days of hearts sent before payment, and the day the price was worked out.

Sharing the form also sends how well the sharer says they know the other player, and the vouches counted. That is a personal appraisal, and it does not necessarily belong in the hands of the person it describes.

In the first case those two settings are absent from the link rather than merely hidden: they are not written into it. The sending pace is, because it is one of the terms — so the other player can work out roughly how many hearts a day were assumed of them.

Enough to show the result again, and nothing else. No account, no nickname, no address, nothing that identifies a person.

A link that has been truncated, or mangled in transit, is rejected rather than guessed at: the page opens on its defaults. The link carries no signature, though: it proves nothing about who built it, or that it has not been altered. The figures in it read like any other received message, and are worth re-entering to check.

A shared result keeps the price of the day it was made, which is why that day appears on the preview. Moving any setting recalculates it.

There is no account, no sign-up and no identifier. The settings are sent to the pricing server to obtain a price, and nothing that leaves identifies a person. The server keeps no record of the request.

They do travel further in one case: with "share the form" the settings are written into the link itself, so the server that builds the preview reads them, and the picture it draws is cached for about a month.

Traffic is counted by Cloudflare Web Analytics, which sets no cookie and stores nothing on the device; like any web request, it discloses the visitor IP address to that service. The one thing the site keeps locally is the chosen language.

That the calculation may still change. A price obtained today is not guaranteed to be the same in six months, and a shared link keeps the price of its own day.

Nothing is promised about availability, or about keeping anything.

The code and the content are under CC BY-NC-ND 4.0: redistribution with attribution, no commercial use and no derivatives.

The flags come from a third-party icon set under the MIT licence, credited in the repository. Suggestions go through an issue on GitHub.