Independent Game UX Study · RDR2
Investigating how players anticipate and recognize who will receive an interaction when several characters or animals are nearby.
context and role
After returning to RDR2 following a long break, I recorded situations where I knew what I wanted to do but had to work out who would receive the action. I conducted the investigation, player research, HUD exploration and Figma prototyping.
How might I help players anticipate and recognize the target of an interaction while preserving RDR2's pace and their connection to its world?
This is an independent portfolio project with no affiliation with Rockstar Games and no access to the game code.
research scope
Seven players explored the map and completed a mission while verbalizing their intentions. I observed intention, action and recovery.
Seven players compared ways to indicate the recipient. Two had also participated in the first stage.
The same seven from the visual evaluation tested identification and target switching with a PS5 controller.
problem definition
I documented issues involving recalled controls, contextual actions, selected objects and system feedback. Target selection became the focus because it appeared in my recordings and in sessions with other players, and could be investigated through a specific change to feedback.
Some participants prepared an action with the horse in mind while horse options remained visible, then interacted with a nearby person. The action was available, but the recipient did not match their intention.
Near a carriage, one participant wanted to address the driver but the interaction went to the passenger. The difference changed whether the action could achieve the intended outcome.
Participants knew how to move the camera or switch targets, yet hesitated while identifying who was selected. Ease of recovery did not remove the mismatch between intention and feedback.
A cue closer to the recipient could help players associate the command with the character. An explicit switching instruction could also expose how to correct the selection.
concept exploration
I applied each direction to a horse with nearby NPCs and to a wagon with two occupants. The concepts addressed different moments: anticipating a target before selection and recognizing it after the menu opened.
Direction A
The portrait avoided adding another marker over the scene, but participants still had to compare a small image in the HUD with the characters in the world. I observed more hesitation and did not pursue it as the primary direction.
Direction B
The marker linked selection directly to the NPC while preserving the action list in its existing corner. It reduced the need to compare the scene with a separate representation.
Direction C
Placing L2 and “Interact” over the potential target addressed the earlier moment: helping players anticipate the recipient before opening the action menu. All seven participants associated L2 with the indicated NPC in the visual evaluation.
chosen solution
I combined the prompt from Direction C with the proximity of Direction B. Before L2, the prompt identifies the potential target. During selection, a marker follows the selected NPC, the actions stay in the lower-right HUD, and “Switch Target” remains visible beside them.
Anticipate who will receive the interaction.
Connect the recipient to the available actions.
Make the change immediately perceptible.
prototype evaluation
I replaced generic screen clicks with buttons from a PS5 controller connected to the computer. The same seven participants from the visual round then tested target recognition and switching.
All associated the L2 prompt with the indicated NPC.
All completed target switching without observed difficulty in the tested flow.
None had difficulty identifying the target when two characters partially overlapped.
The portrait remained understandable but added comparison between two screen areas.
The final participants already knew the proposal and the scene. The prototype tested comprehension and execution, not first exposure, moving targets, combat or every contextual rule in the game.
what I learned
Players could know the correct input and still hesitate because the selected recipient was unclear. Adding information to the HUD could preserve that comparison effort. The strongest direction linked feedback directly to the relevant character while keeping the action list in a familiar location.
A playable implementation should next test moving targets, entering and leaving interaction range, denser NPC groups, different viewing distances and players with no prior exposure to the proposal.
secondary research
These sources supported hypothesis development and opposing perspectives on friction, pacing and access to information. Public comments were treated as exploratory evidence rather than prevalence data.
next case
