When to Leave With What You Have

Runs in the adventure and collection categories end from collecting one more thing, almost never from leaving something behind. That is not a character flaw, it is a failure to do arithmetic in advance - and the arithmetic is simple enough to do before the run starts.

What an exit bonus is actually worth

Every game in these categories pays a bonus for leaving, and it scales with the resource you have left rather than being flat.

Treasure Hunt pays 300 per vault level plus eighteen points per second remaining on the clock. Dungeon Tunnel pays 400 per level plus twenty-five per second of lantern. Underwater Kingdom Quest pays 350 per level plus thirty per second of air. Monster Escape pays 320 per chamber plus fourteen per second under a minute.

Those per-second figures are the numbers that matter. Eighteen points a second means ten seconds of clock is worth 180 - which is between one and two gems. A detour that takes ten seconds and yields one gem is therefore a loss even if nothing goes wrong.

Most players have never done that calculation, and it changes behaviour immediately once they have.

Why banking compounds

The bonus is only half the value. The other half is what carries into the next level.

Dungeon Tunnel carries eighteen seconds of lantern down with you. Underwater Kingdom Quest carries your remaining air. Treasure Hunt tops the clock up on arrival, by less each time. Knight Quest restores a heart.

So leaving with a margin pays twice: once as points and once as a better start. That compounding is why a player who consistently exits with twenty seconds reaches level five comfortably while one who consistently scrapes out at two does not reach level three - even if they collect the same amount per level.

The corollary is that early levels matter more than they feel like they do. A clean first vault is worth more than a heroic third one.

Setting a rule you will actually follow

The decision has to be made while it is easy, which means before it is needed.

A rule like 'I leave at twelve seconds' or 'I leave once I have the quota plus three' can be made calmly at the start of a vault. The same decision at thirteen seconds with two chests visible is not a decision, it is an impulse, and impulses in this category are reliably wrong.

The rule should be numerical rather than conditional. 'I will leave when it gets dangerous' is not a rule, because danger is exactly the thing you will misjudge.

And it should account for the walk. In a large vault, twelve seconds of clock is not twelve seconds of collecting - it is perhaps six, plus the trip to the exit. Most players set their threshold as though the exit were free.

When staying is right

It is not always wrong, and the cases where it is correct are worth recognising.

When a chest is close. A chest is 250 points, which is roughly fourteen seconds of Treasure Hunt clock. If it is four seconds away and unguarded, take it.

When your combo is high. In games with a collection chain, the multiplier applies to everything you pick up, so a dense cluster at a five-times chain is worth far more than its face value.

When the next level is worse. In Underwater Kingdom Quest, courts get deeper with fewer vents. Sometimes the right play is to bank a bigger haul now rather than assume you will get a comparable chance later.

And when the resource is nearly full anyway. If your clock is topped up and the bonus is already near its ceiling, the marginal second is worth less than the marginal coin.

The specific trap: one more chest

The failure mode is consistent enough to name. You are at your leaving threshold, the exit is open, and there is one visible thing you did not get.

What happens next is that the detour costs more than expected - it always does, because you budgeted the distance and not the route - and you arrive at the exit with nothing left, or do not arrive at all.

The reason it is so consistent is that the visible item is concrete and the bonus is abstract. Two hundred and fifty points sitting on the floor feels like more than 'eighteen a second for the fourteen seconds this will cost', even when the numbers say otherwise.

The fix is to have done the arithmetic before you can see the chest.

The same decision in the collection games

It appears in a different shape in Dragon Treasure and Treasure Conveyor, and it is the same problem.

In Dragon Treasure the banking bonus grows faster than linearly with how much you carried at once, so one dangerous trip beats three safe ones - right up until being hit scatters everything. Carrying a full load slows you by forty per cent, which is when the decision stops being arithmetic and becomes nerve.

In Treasure Conveyor the carry limit is three, so the decision is per item rather than per trip: taking a third item when a delivery is one step away costs you the delivery.

In both, the rule that works is the same: decide the threshold before you are at it.

Frequently asked questions