Lives are a second checkpoint

Platformers dropped lives, and the eulogy wrote itself: lives are an arcade tax, a way to eat quarters, a punishment modern games have outgrown. Look at the rule instead of the costume. A life is a checkpoint of checkpoints. Miss a jump and you restart the room. Miss the room too often and you restart the set. Two layers, two prices.

The layer is useful when the set is short enough to hold in your head, the way an arcade board or a single Cuphead boss is short. Spread across a forty-hour adventure, the same counter just deletes an evening. Scope it to one challenge, or throw it out. Better, sometimes, to let the player spend a resource to place their own save, so the harshness is a decision. The outdated part was never the idea of a cost. It was sending the player back further than the lesson required.

A life counter is a second layer of checkpointing sitting on top of the first, and it reads much more clearly once you look at it that way.

The bottom layer is the permanent save, which banks progress for good. On top of it sits a temporary checkpoint that shortens the retry loop inside the stretch you have already banked. Die, and you spend a life and resume at the temporary point. Spend the last one, and the temporary points evaporate and you drop back to the permanent save.

Where a game draws that boundary varies a lot. The oldest games put the permanent line at the start of the whole run. Later platformers moved it to a level entrance or a point on an overworld map. Crash Bandicoot 4's Retro mode treats a level as the unit of permanent progress and the checkpoint boxes inside it as the disposable layer.

Recent design has mostly collapsed the two layers into one. When every checkpoint is permanent, a death counter may still tick upward, but nothing is actually taken from you for grinding two hundred attempts at the same jump.

Lives fell out of favour because running out usually means replaying content you have already proved you can beat.

Anyone still learning the game feels that most. One awkward section becomes a long stretch of redone work, and being kicked back to a level entrance turns a single obstacle into a chore.

They are also awkward to tune. Hand out extras freely and the counter stops meaning anything. Hand out too few and the curve spikes somewhere that reads as punishment rather than as a skill to learn.

Dropping the counter altogether cleared that up for a lot of platformers. With no lives in the mix, difficulty gets tuned against the thing in front of the player. A room, a jump, an enemy pattern, a whole level. Not against the distance to a far-off save point.

The strongest defense of lives is stakes. When emptying the counter costs you progress that mattered, every individual mistake starts to weigh something.

Players behave differently under that weight. Attrition stops being a viable strategy, so they slow down, actually read the patterns, and get a much bigger release when the next safe point arrives.

Stakes also give rewards something to attach to. A 1-UP mushroom, a Mega Man face, a coin total, a hidden room or a bonus game is worth chasing precisely because it buys attempts you would otherwise not have.

For a designer that is leverage. Exploring, gambling on a risky route, picking up an optional item, or memorising a secret path all become worth doing, because the payout is attached to a consequence the player can feel.

Lives also chain challenges together, which a single permanent checkpoint cannot do.

You are no longer asking whether the player can eventually clear one room. You are asking whether they can clear a run of rooms and still arrive at the end with something left in reserve.

Mega Man shows this plainly. Getting to a Robot Master is not just a matter of reaching the boss door. You want the earlier stretch played cleanly so you walk in with health, lives, ammunition and nerve intact.

That is a satisfying shape because good play compounds. A tidy opening stacks the odds in your favour later on, and a sloppy one raises the tension without necessarily ending the run on the spot.

Often the fault is not the life counter but the fact that it floats freely across an entire game.

Pinning it to one level, boss or set-piece fixes that. Furi hands you exactly three lives at the start of every boss fight, so you never turn up with a hoard or with nothing.

The designer gets a fixed target to balance against. Everyone meets the fight on roughly the same terms, and the counter becomes part of that encounter's design rather than a running total to keep track of.

It also heads off the classic failure mode, where the strong player banks so many lives that risk quietly disappears while the struggling player reaches the hard part with no room left to learn in.

The other fix is to keep lives and take the sting out of losing them all.

Kero Blaster still sends you backwards on a game over, but your cash and upgrades come with you. Redoing an area stings less when some of what you earned is permanent.

It goes further and makes the failure itself productive. You come round in a hospital, and the hospital has a shop selling things worth having, including extra health and extra lives.

Other games attack the same problem from other angles. Sonic levels can open alternate routes on a second attempt. Roguelikes take it furthest of all, rebuilding layouts, items, events and situations so that what you lost is not what you have to redo.

A third route drops lives but keeps risk spread across a chain of challenges, by putting the checkpoint itself in the player's hands.

Shovel Knight lets you smash a checkpoint for gems. You bank the reward immediately and you know the bill: die after that and you restart much further back.

Panzer Paladin runs a variation on it. Activating a checkpoint costs you a weapon, so safety is traded directly against the state of your arsenal.

Ori and the Blind Forest goes furthest, letting you place your own checkpoints in exchange for a resource. The question stops being how many lives are left and becomes where progress is worth saving and whether you can spare the cost.

The sharpest move may be to invert the penalty entirely.

Rather than taking something away when the retries pile up, hand something out when they do not. The mastery challenge survives intact, and nobody new has to replay half a level to get another shot at it.

Bullet hell shooters have run this way for years. Continues are unlimited, yet the one credit clear is still the goal that earns respect, because it means the whole game without a single continue. Sonic Forces folds retries into the final score, so a top rank demands a clean run.

Crash Bandicoot 4's Modern mode does the same thing in a platformer. Deaths are still counted, they just no longer throw you out of the level, and one gem is reserved for finishing with fewer than three retries.

Rewarding clean play instead of punishing messy play also settles an audience problem that lives usually make worse.

Someone learning the game can keep going without a hard section threatening to wipe out an hour. They get the space to read the level, get comfortable with the controls and move on.

The expert still gets the discipline of a life counter, only now they choose to take it on. Clean clears, higher ranks and gems are there for anyone who wants the pressure.

One system, two experiences: progress that keeps flowing for players who need it, and stakes worth sweating over for players who came for exactly that.

So are lives outdated? Not inherently. It comes down to how they are built.

Balance decides a lot of it. So does the gap between permanent saves, the value of an extra life, the cost of replaying a stretch, everything the counter touches, and how legibly the game communicates the pressure it is applying.

The harder question sits underneath all of that: why is the counter there at all? Plenty of games inherit lives because the platformers they grew up on had them, or because a life icon reads as retro at a glance.

That is not a reason. Lives should be in a game because the game needs the tension, the reward structure, the mastery ceiling, the pacing or the resource planning that they produce. You find out whether they do by playing your own build and being honest about what the counter changed.

Lives are not obsolete. They are just no longer the default.

If your game wants sustained stakes across a run of challenges, a counter may be exactly right. If it wants fast retries, accessibility, or mastery measured one room at a time, it is probably better off without one. If it wants both, clean play can be an optional prize rather than a tax on survival.

The counter is one instrument among many for controlling pressure. Permanent checkpoints, checkpoints you gamble with, retries scoped to a single level, upgrades that stick through failure, score penalties and mastery rewards all reach for nearby problems.

And the point generalises past lives. Good games do not come from inheriting the furniture of a genre. Every system in the build should be there because you decided it belonged there.

In the Flockbay app, build one short run of three rooms and give it a stock of tries. Then build the same three rooms with a restart that only sends you to the room. Play both. Keep the version where a failure teaches the jump instead of taxing the afternoon.

A precision platformer template makes the cost obvious. The AI platformer maker can build the two versions side by side.