EldenRing:CharacterCreationRedesign

For new players, the volume of choices creates a kind of paralysis before the game has even started. Ten origins, eight stats, an optional keepsake, and no explanation for any of it.

The goal: reduce cognitive overload without reducing depth. The complexity has value — the problem is how the information is organized and surfaced.

Solo project. Unofficial fan concept, not affiliated with FromSoftware or Bandai Namco.

Scoped to PC — mouse and keyboard. That decision drives the interaction model throughout: hover-triggered contextual help, cursor-based selection, keyboard prompts. A console version would need a different help layer, since focus replaces hover and the readable distance changes. Out of scope here, on purpose.

I made three structural changes. Everything else follows from them.

Implemented in UE5 with UMG — a partial prototype, built to test whether the design survives contact with an engine instead of a Figma canvas.

Three things I built for it rather than around it:

I mapped friction across the early-game flow — character creation, opening cinematic, first combat encounter. I focused on moments of hesitation: where players slowed down, reread, or gave up and randomized a choice.

Two problems accounted for most of it.

I prototyped a new flow from the friction map, then built V1 around the three structural changes.

What the testing was, and what it wasn't

I ran formative testing with 5 participants — a mix of new and returning Elden Ring players. Five is enough to surface comprehension problems. It is not enough to prove a success rate, and I didn't run a baseline pass on the original screen, so nothing here is a comparative claim.

The split mattered more than the total. Returning players already know what a keepsake is, so their sessions validated the layout but told me nothing about comprehension. The new players are who the redesign is for. That's where the findings came from.

The second flag stung, because action prominence was one of my three structural changes. V1 fixed the layout and didn't fix the noticing — the buttons were present, labeled, positioned in the scan path, and still missed. Presence isn't prominence. V2 is where that got addressed.

The visual direction had to walk a specific line: different enough to feel like a genuine proposal, close enough to the Elden Ring universe that it didn't feel like a different game.

Modernizing it would have been easier and wrong. The opacity is a structural problem, not a stylistic one — solving it by flattening the game's visual language would fix the wrong thing.

I ran the palette through a color blindness pass — protanopia, deuteranopia, tritanopia — to check that nothing in the interface depended on hue alone to be readable.

Shared with the same testers after the visual pass:

"Character creation is much more intuitive and easier."

"I understand more easily the differences between origins."

"It is clear to understand what a Keepsake is."

"I like the new design, it's a new interesting style."