Why checkpoints matter
When a player fails, the game has to decide what happens next. Does the player start the whole level again, retry only the last fight, or return to a spot saved long ago? That decision shapes how risky each moment feels and how willing players are to try something bold.
Checkpoints and save points both record progress, but they work differently. A checkpoint is usually automatic: the game quietly stores your position when you cross a threshold. A save point is typically something you interact with or choose to use, and it often stores more, such as inventory or story flags. Some games use only one kind, some use both, and some let players save anywhere.
The tension and frustration trade-off
Designers often describe checkpoint spacing as a dial between two feelings.
- Checkpoints far apart raise the stakes. Each section matters because failure costs real time. This can create strong tension and a satisfying sense of relief when you finally reach the next one.
- Checkpoints close together lower the stakes. Players can experiment and learn quickly, but the sense of danger can fade.
Neither extreme is wrong, but each suits a different kind of game. A game built around careful survival may want sparse saves so that every decision carries weight. A fast action game built around trial and error usually wants frequent checkpoints so players can retry in seconds. This choice connects to the broader difficulty curve, because the same encounter feels easier or harder depending on how much is lost by failing it.
A simple worked example
Here is a hypothetical case with made-up numbers, just to show the reasoning.
Suppose a section takes a skilled player about 3 minutes to clear, and a typical new player fails it twice before succeeding. The checkpoint sits at the start of the section.
- Each failed attempt costs, on average, 2 minutes of play before the failure, plus a short load or respawn, say 0.5 minutes.
- Two failures cost 2 x (2 + 0.5) = 5 minutes.
- The final successful attempt takes 3 minutes.
- Total time to get through: 5 + 3 = 8 minutes.
Now suppose the designer adds a checkpoint partway through, so a failure in the second half only costs about 1 minute of replay plus the same 0.5-minute respawn. If the player fails twice in the second half, the cost is 2 x (1 + 0.5) = 3 minutes instead of 5. These figures are illustrative, not measured, but they show the point: shifting a checkpoint changes the price of each mistake, and that price shapes how players feel about the whole section.
Common placement principles
Place them after effort, not before
A checkpoint that arrives just after a hard fight or a long climb tells the player their work is safe. Putting one right before a difficult moment also helps, so the retry begins at the challenge rather than a long walk away from it.
Avoid replaying the boring parts
One of the most common complaints is the “run back” problem, where a player must repeat a long, uneventful stretch before reaching the thing that killed them. If the difficult part is the fight, the checkpoint should usually be close to the fight. When a long walk is unavoidable, designers sometimes add shortcuts that open up after the first trip.
Respect the story and the pacing
Checkpoints often sit at moments when the action naturally pauses: the end of a cutscene, a room transition or the entrance to an arena. This keeps the structure tidy and avoids saving the player into a nearly lost position, for example with almost no health and enemies right behind them.
Save in a safe state
A technical but important point is that the saved state should be one the player can actually continue from. If a checkpoint triggers while the player is falling, burning or about to be hit, the reload can immediately cause another failure. Careful designers and programmers check that the stored state is stable, which depends on how the game handles save systems internally.
Different games, different answers
- Linear action games tend to use frequent automatic checkpoints, often just before and after encounters.
- Exploration games may let players save anywhere, since there is no single path and restarts would be punishing.
- Roguelike games often avoid traditional saves entirely, treating failure as part of the structure.
- Survival-style games sometimes tie saves to a limited resource or specific locations, making the decision to push on or return a core choice.
Whichever model is used, the game should make it clear when progress has been stored. A small icon, a sound or a short message helps, and good audio cues can reinforce it so the player does not have to guess.
What the evidence looks like
Designers do not have to rely on instinct. During playtest sessions they can note how many attempts players need at each point, how long a retry takes and where people quit. If a large share of testers stop at the same spot, that is a sign to look at both the challenge and the nearest checkpoint. A single player’s frustration is a story; the same pattern in many sessions is a signal.
Putting it together
Checkpoint placement is a way of pricing failure. Place them too sparsely and the price may feel unfair; place them too densely and nothing seems to matter. The best choices usually come from knowing what feeling the section is meant to produce and checking, through observation, whether players actually feel it. When the spacing fits the intended experience, most players never think about it at all, which is often the sign it worked.