Resource Management in Dungeon Games

In a survival game, health keeps you alive. In a dungeon game it is currency, and the run is decided by what you spend it on. This guide covers the arithmetic that makes that framing correct, and how to apply it while actually playing.

Health does not refill, and that is the design

Dungeon Blade gives you six health and exactly three caches, each worth one heart. Pixel Sword Quest gives you five and restores one between levels. Sword Maze gives you four and no healing at all.

That scarcity is what makes every fight a purchase rather than an obstacle. A fight you win at the cost of two hearts is a fight that may have cost you the run three rooms later, even though you won it.

The question that replaces 'can I win this' is 'what does winning this cost, and is that what I want to spend on'.

The arithmetic

Dungeon Blade makes it explicit. Escaping pays 3,000 plus 500 per point of health you finish with. Clearing a room pays 300 plus 60 per room already explored.

So finishing with six health is 3,000 points. A room cleared mid-run is worth perhaps 600. Six health is therefore worth roughly five cleared rooms - and clearing five rooms costs considerably more than six health at any reasonable difficulty.

The conclusion is uncomfortable and correct: the highest-scoring run is usually the one that fought least. A player who walks a direct line to the key and out again routinely beats one who cleared all twelve rooms.

Which fights are mandatory

The room containing the key, obviously. Doors bar while anything hostile is alive, so a room you enter is a room you finish.

Any room on the shortest route to it. In Dungeon Blade the key is always in the far column, so at minimum you are crossing three rooms to reach it and three to return.

That is six rooms out of twelve as a floor. Everything above six is a purchase, and the only ones usually worth making are rooms you have reason to believe contain a health cache.

Sword Maze: time instead of health

Sword Maze runs the same idea with a different currency. The torch is 150 seconds, each kill adds four and each hit costs eight - a twelve-second swing per encounter.

That converts the same decision into a time budget. Six clean kills is 24 seconds of torch, which is most of a corridor; six hits taken is 48 seconds lost, which is usually the run.

The difference from Dungeon Blade is that fighting well is actively profitable here rather than merely cheap. A player who kills efficiently gains time, which means combat competence and route efficiency point in the same direction rather than competing.

Practical rules

Applied to any of the three exploration games.

• Ask what a fight is for before entering the room. If the answer is nothing, do not enter.
• Count your health as points, not as lives. In Dungeon Blade each point is literally 500.
• Detour only for health caches and only when you know roughly where they are.
• In Sword Maze, kill things on your route and nothing else - each is four seconds either way.
• Take the direct line to the objective first, and only explore afterwards if you have health to spare.

Planning around the caches

Three health caches in Dungeon Blade, each worth one heart, and they are the only thing that makes a detour arithmetically defensible. Everything else you might find is worth less than the health it costs to reach.

A cache is worth 500 points at the exit plus whatever it lets you survive, which means a detour costing one heart to gain one heart is roughly break-even on score and clearly positive on safety - you converted a certain cost into a restored buffer. A detour costing two hearts to gain one is a straightforward loss.

The practical rule is to detour for a cache only when you can reach it through rooms you expect to survive cleanly, which in practice means rooms nearer the entrance. That is the same conclusion the difficulty gradient produces from the other direction, and the two together are most of the route planning the game needs.

What a bad run looks like from the inside

Dungeon runs fail in a recognisable pattern, and it is worth naming because it feels like competence while it is happening.

The player clears the first two rooms comfortably, which establishes that fighting is affordable. They clear two more on that basis, spending a heart or two. Now they are deeper, the rooms are harder, and they have less to spend - but the habit of clearing is set, and each subsequent room is entered on the assumption established at the start.

The tell is entering a room without deciding to. If you find yourself inside something you did not choose, you are running on a policy set several rooms ago, under conditions that no longer apply. The counter is the half-second pause at every doorway - not to decide whether you can win, but to notice that you are making a decision at all.

Reading a room before entering it

You cannot, and that is deliberate - the doors bar behind you, so peeking is not free. What you can do is reason about position.

Enemy difficulty in Dungeon Blade scales with how far into the dungeon a room sits, measured from the entrance. A room in the far column with a row offset is the hardest kind in the game; one adjacent to the entrance is the easiest.

That means the cheap rooms are the ones near where you started, and the expensive ones are exactly the ones on the route to the key. Detouring back toward the entrance to farm easy rooms is a real strategy, and it is why the minimap matters.

Frequently asked questions