An onboarding overlay can explain where a control is. It cannot guarantee that the same control will make sense after the overlay disappears. A more durable explanation lives in the navigation itself: names that stay consistent, visible relationships between views, and a way back that does what the player expects.
Our drawing starts with a small library of imaginary games. There are no real titles or hidden recommendation scores. The player opens a detail from a collection called “Small worlds.” The design question is whether they can understand where they are and return to the same collection without reconstructing the route from memory.
Let each move leave a readable trace
The parent view has a name, and the detail view keeps a return control. These are small pieces of information, but together they create a relationship. The player can distinguish opening an item from replacing the entire library. That distinction should survive a smaller screen even if both panels can no longer be shown at once.
On desktop, the illustration places the parent and detail next to each other. On mobile, a full-screen detail would be more plausible. In that version, a clear “Back to Small worlds” link can carry the relationship that the desktop layout communicates spatially. Copy and layout solve different parts of the same problem.
The return should preserve the chosen filter and, where practical, the player’s position. Otherwise, inspecting one item becomes a destructive act: the person has to rebuild their shortlist before inspecting another. Our drawing does not implement browser history or scroll restoration, but it makes those requirements available for discussion before code hides them.
Use names that explain the decision
“Explore” may be an inviting label, but it tells us little about whether the next view is a catalog, a search field or an editorial feature. A concise noun can be more useful when the player is trying to locate a known object. An expressive label still needs a stable meaning.
In this example, “Library,” “Small worlds” and the selected detail describe three levels. They should not suddenly become “Home,” “Discover” and “Journey” on a different device. A mobile layout can reorganize space without changing the vocabulary that a reader has already learned.
Tags are another form of navigation language. Steam’s documentation offers a real example of a platform using descriptive tags. Our concern here is not how to gain visibility on that platform. It is whether a tag helps someone make a decision. A label that sounds attractive but cannot distinguish two items adds clutter rather than direction.
Teach through consequences, not surprises
The first click should give the player evidence about how the system works. If selecting a title opens a detail, the next title should follow the same convention. If a control starts a game immediately, it should look and read differently from one that merely opens information.
This is especially relevant to focus. Keyboard users need to know which control is active before triggering it. A design that reveals its selected state only on hover leaves that explanation unavailable to another input method. Visible focus belongs to the navigation language; it is not a debugging decoration.
A gesture can be a useful shortcut, but a return path should not require guessing that a swipe exists. The quiet return pattern offers a named action instead. That takes space, and the trade-off is worth discussing. Hiding the action also has a cost: the interface must then teach and support an invisible convention.
Walk the difficult path first
For a critique, begin with a person who opens the wrong item. Ask them to return, change a filter, inspect another item and undo that filter. Then imagine they leave and come back through a copied link. Which pieces of context are still meaningful? Which are local to the previous visit?
An empty result is another useful test. It should explain that the current combination has no matches and offer a reset. A blank space can instead suggest that the whole catalog is missing. Good navigation teaches the scope of an action as well as its location.
Write down the expected return destination for each route. Avoid a generic “Back” rule that sometimes means browser history and sometimes means a fixed parent. Either behavior can be appropriate, but inconsistency makes the interface harder to learn. Name the destination when the surrounding context cannot make it obvious.
The limits of a tidy map
This study covers a small hierarchy. It does not resolve complex multiplayer menus, deep inventories, localized text expansion or every controller convention. Those contexts may need different navigation structures. We have not observed players completing tasks in this prototype.
What the diagram contributes is a testable question: can a player explain where an action will take them, what changed, and how to return? A navigation system that answers those questions through use has already taught something valuable. It gives the player a model they can carry into the next decision, without asking them to remember a tour.
Sources & limits
- W3C: WCAG 2.2, Consistent Navigation
A reference for keeping repeated navigation mechanisms in a consistent order. - Microsoft XAG 107: Input
Supports alternatives to input demands that can exclude players, including gesture-only interactions. - Valve: Steam Tags
A real platform example of descriptive tags; our synthetic taxonomy does not reproduce its ranking system.
Sources consulted 16 November 2026. Diagrams and interpretations are original editorial work. These sources support the stated principles; they do not validate our example through player testing.