Uplifting Goat

There is a genre conversation happening right now that I think about a lot. It goes something like this: "The roguelite format is proven. It drives engagement. Players keep coming back. Why not make your shmup a roguelite?"

It is not a bad question. But I think the answer is more complicated than the engagement numbers suggest.

## What Roguelites Are Actually Doing

Roguelites work because they solve a specific problem: replayability without memorization burnout. When the run is different each time, players do not feel the pressure to learn a fixed sequence to completion. They drop in, have an experience shaped by the run's variance, and come back. The barrier to re-entry is low.

Dead Cells is probably the clearest modern example of this working at high quality. Each run has enough variance that it feels meaningfully different. But the *moment-to-moment* combat is deeply consistent. The core skills you build — reading enemy tells, positioning, timing parries — transfer fully from run to run. You are getting better at a stable game inside a variable structure.

This is the key. The variance is at the structural level. The skill is at the execution level. Those two things do not conflict.

## Where Shmup Roguelites Tend to Go Wrong

The tension comes from what shmups fundamentally are. Traditional shmup design is built around the assumption that players will play a stage many times. That assumption is not incidental — it is the design language. Bullet patterns, stage layouts, boss sequences — all of it is designed to be *learned*. The game reveals itself through repetition.

When you introduce roguelite variance at the macro level, you are removing the guarantee of that repetition context. Players who want to deeply learn a specific enemy encounter may never see that encounter in the same form twice. The thing that makes traditional shmups great — the depth of pattern mastery — is partially undermined by the thing that makes roguelites engaging.

This is not a reason not to make roguelite shmups. It is a reason to be clear with yourself about which game you are making.

Games like Nex Machina and Rolling Gunner sit in interesting territory here. They are not roguelites, but they are designed to be understood and replayed deeply. The patterns matter. The routes matter. You are supposed to learn them.

Contrast that with something like Crimzon Clover or recent indie shmup-adjacent games that blend genres. They make different promises and those promises change the design.

## The Hybrid Problem

Here is where I see most designers struggle: trying to be both. Deep, learnable patterns for mastery players AND roguelite variance for casual replayability. This sounds good in a pitch doc. In practice, it often means you have half-executed both visions and served neither audience fully.

The mastery players are frustrated because the variance means they cannot go deep on stage knowledge. The roguelite players are frustrated because the encounters are designed assuming pattern memorization, which they are never going to do.

You can bridge this. Vampire Survivors adjacent games do it by making individual encounters simple enough to grok quickly even without memorization. But the cost is that the individual encounters have shallow mastery ceilings. You are trading encounter depth for structural variety.

## IS2's Stance

Interstellar Sentinel 2 is a traditional shmup at heart. I want players to learn it. I want the twentieth run to feel different from the first because of what the player knows, not because the stage is different.

That is a choice with costs. The replayability argument for roguelites is real. But I think the specific satisfaction of shmup mastery — the moment you route through a wave that used to kill you, with calm and precision, because you genuinely understand it — is incompatible with heavy run variance. You cannot have that moment if the wave was different last time.

So IS2 will not be a roguelite. The stages are fixed. The patterns are learnable. The game rewards coming back because you get better, not because the game reshuffles.

If you want to taste where the skill ceiling is, the demo is a fixed slice of that experience: [Play the IS2 demo on Steam](https://store.steampowered.com/app/3486180/Interstellar_Sentinel_2/)

---
*Previous: Post 5 — Randomness in Enemy Design: Spawning, Behavior, and Fair Surprises*
*Next in series: Post 7 — Perceived Randomness: What Players Think They're Fighting*