Roguelikes, Persistency, and Progression

One of the most fascinating genres in games is the roguelike. The term originally described games that closely mimicked the 1980 dungeon crawler Rogue, but it has since been stretched across a wide range of indie games.

Most of these games share two important traits. The levels are randomly generated each time you play, and death resets progress back to zero.

But that second trait is not always what it seems. In Spelunky, permadeath means permadeath. When the player dies, progress through the game is wiped, and the next run begins with the same starting health, bombs, and money as the very first attempt.

In Rogue Legacy, death is less severe. The money found during a run can be spent on upgrades that improve health, magic, attack power, and other stats. Those upgrades persist across future characters.

Some players call games without persistent upgrades roguelikes, and games with permanent upgrades roguelites. The distinction matters because each approach creates a different relationship between failure, progress, and skill.

Start with roguelikes that do not have persistent upgrades, such as Spelunky and Enter the Gungeon.

The key advantage is that they almost exclusively reward player skill. Because the game does not change from run to run, the broad difficulty level across multiple attempts stays flat.

The only way to break through that barrier is to improve: learn the rules, practice the controls, understand enemy behavior, and become more familiar with the game's systems and world.

Nothing stands between the player and the final boss except ability, knowledge, and the luck or cruelty of a random roll. A highly skilled player has no artificial gate stopping them from finishing the game immediately.

The disadvantage is obvious: a less skilled player may never finish. A roguelike asks the player to survive a difficult game in one go, with no permanent way to lower the difficulty over time.

If the player cannot meet that demand, they may simply never succeed. That can be thrilling for one player and exhausting for another.

There is another drawback. Because progress is wiped clean every time the player dies, failed runs can feel like a waste of time. The player may be improving, but there is nothing tangible to show for it.

Skill growth is real, but it does not always produce the satisfying feeling of numbers going up, upgrades unlocking, or a visible collection expanding.

Roguelites take the opposite approach. Games such as Rogue Legacy and Dead Cells include permanent upgrades that carry from run to run.

The advantage is that almost every attempt gains meaning. When a player dies in Rogue Legacy, the run still matters if it earned enough money to buy a new upgrade and improve future chances.

Over time, every attempt helps bring the difficulty down. That also means many more players can eventually finish the game. If they keep upgrading, the character becomes stronger and the game becomes easier to overcome.

For players who would otherwise bounce off a pure roguelike, that structure can be welcoming. Failure is still failure, but it is rarely empty.

The drawback is that this creates a strange difficulty curve. The game is often hardest at the beginning, then gradually becomes easier as permanent upgrades accumulate.

That is the reverse of most games, where difficulty rises as player skill improves. Here, the player's skill may be rising while the game's resistance is falling.

One pernicious result is that success can feel more dependent on character level than player skill. In Dead Cells, it can seem as though the real chance to win only arrives after buying more health potions or other upgrades.

That creates an artificial barrier to grind through before the player can focus on winning. And when the victory finally comes, the result can feel ambiguous. Did the player get better, or did the game simply get easier?

Both approaches have advantages and disadvantages, but designers have found ways to soften the worst problems.

Pure roguelikes are often accused of offering no sense of progression. One answer is to unlock things that are fun without meaningfully increasing the player's chance of finishing the game.

Enter the Gungeon gives the player a special currency for killing bosses. That currency survives death and can be spent in a hub shop, but it does not buy extra health or bigger ammo crates. It adds more weapons to the pool of possible random drops.

Because the new guns are not strictly more powerful than the starting pool, the game's balance is not dramatically changed. Future runs gain variety rather than a permanent difficulty reduction.

Nuclear Throne does something similar with unlockable characters. They are not simply stronger than the defaults. Each has advantages and disadvantages, creating different ways to play rather than a ladder of power.

Cosmetic unlocks offer another path. Downwell rewards long-term gem collection with new color options that change presentation without changing balance.

Progression does not have to be mechanical. Hades includes persistent upgrades, but one of its most interesting forms of persistence is narrative.

Each return to the hub can reveal more about characters, relationships, and backstories. Items found during runs can be given to characters to unlock more story content.

That gives the player something to hold onto even while failing repeatedly at the action game. The run may end, but the world, characters, and conversations continue to move forward.

This kind of persistence can make repeated failure feel less empty without relying entirely on permanent stat growth.

Pure roguelikes can also seem impenetrable because there is no way to permanently reduce the difficulty. But they can still help players without permanently changing the game's balance.

Spelunky uses shortcuts. By helping the tunnel man, the player can unlock a door that starts a run in a later area, such as the jungle, skipping the mines entirely.

That makes practice easier and lets new players learn later areas. But Spelunky clearly signals that a proper completion still starts from the beginning. Leaderboard times do not count when shortcuts are used, and the true ending requires a chain of items hidden across every area.

Temporary boosts are another option. Into the Breach lets the player carry one leveled soldier into the next attempt. That provides a useful leg-up and a strong reason for one more try, but it does not permanently lower the difficulty forever.

Roguelites face the opposite problem: the suspicion that success depends more on time invested than on skill. Designers often add rules that force skill to remain relevant.

Rogue Legacy uses Charon, the doorkeeper who takes all unspent money before the next run begins. The player cannot save up small amounts across many failed runs and then buy a huge upgrade all at once.

To afford expensive upgrades, the player has to earn a lot of money in a single run. That means staying alive long enough to prove some level of mastery.

Dead Cells uses a different rule. Money is not kept after death, so the player has to reach upgrade stations and bank resources before losing them. Progress still exists, but it must be secured through performance.

A roguelite can also make the relationship between skill and upgrades more explicit. Most are designed so that a good enough player can finish without buying a single upgrade.

That possibility can become a meta challenge: how few upgrades are needed before the player can win? The upgrade system becomes a support structure rather than the only path to victory.

The pure roguelike has a powerful appeal because random generation and permadeath prevent memorized routes and slow power accumulation from carrying the player. Success is about skill, knowledge, adaptation, and nerve.

Roguelites offer a different pleasure: the slow accumulation of power and the reassurance that failed runs still count. But that can create a backwards difficulty curve and a lingering question about whether victory came from better play or a softer game.

Neither answer is automatically better. The important design question is what failure should mean, and whether the game's progression system preserves the value of getting better.