How the data was collected.

Think Aloud Protocol is the cleanest way to peek inside live cognition. Players verbalise everything that crosses their mind while they perform the task; the recording is a running narration of what they're paying attention to in the moment. Only the live competitive-match audio fed the analysis. No post-game interview was recorded or coded; every code in this dataset was applied to in-the-moment verbalisation, not to anything participants said retrospectively about how they had been thinking.

Recordings are then transcribed, sliced into short segments, and tagged with what each segment is actually about. For this study, eight elite Valorant players (four Radiant, four Immortal) played a competitive ranked match while talking through their moment-to-moment thinking. The pipeline produced 1,909 coded segments across 162 live-play rounds.

Every coded segment ends up under one of four parent categories. They appear everywhere on the site (streamgraphs, donuts, cards), using the same colour throughout.

• Attentional Metacognitions

  Stepping outside the play to monitor your own thinking: recall, reflection, doubt, post-hoc appraisal.

• Attentional Control · Internal

  Attention turned inward to your own actions or state: feedback, plans, technique under conscious control.

• Attentional Control · External

  Attention on the world outside you: fight planning, communication, reading the opponent's behaviour.

• Game Situation

  Background awareness of the game state: positions, abilities, economy, who's still alive.

A round in Valorant has a rough internal rhythm: buy phase, move out, approach a fight, engage, aftermath, then the round-end echo. Combining the round phase with what's happening combat-wise gives us seven analytically meaningful cells. Every segment belongs to exactly one.

The fight-related cells (Approach, Engage, Aftermath) are recursive: rounds often contain multiple fights, especially when an engagement doesn't end in a death (a poke, a trade attempt, a re-peek). The cohort streamgraph collapses these into a single averaged sequence per round; the round-by-round per-participant view keeps each fight cycle distinct so you can see them stack up.

The cohort streamgraph and per-participant average rounds both use this seven-cell axis, so cohort-vs-individual comparisons stay structurally honest.

Players consented to public release of the analysed data, but kept their identifying details private. Display IDs (P01–P08) are the only identity available anywhere on this site or in the underlying CSV. Original recruitment IDs were remapped at export time and never make it past the anonymisation boundary.

If you reference this work in a piece of media, please cite the underlying study:

Langfield, D. C. M. (2026). Meta-attentional flow in elite Valorant
esports: A phase-structured think aloud investigation (Currently
unpublished BSc Empirical Project, Glasgow Caledonian University).

The full empirical report (methods, results, discussion, references) is available on request. If the report is later adapted and published in an academic journal, the citation above will be updated to point to the published version. Request the full report

* The underlying analysed data is held as CSV; the JSON files this site reads are an interactive representation of the same data, structured for the visualisations on this site.

The eight-player cohort on this site is the elite slice. A follow-up extending the same protocol to novice and sub-elite players is currently in data analysis, with first results expected mid-May 2026. The comparison itself opens a question the elite slice cannot answer alone: do the same temporal-cell patterns hold at lower ranks, or does cognition look fundamentally different further down the ladder?

Once the data lands, a comparison view will unlock on the cohort page automatically.

The next phase opens the protocol up to a much wider pool of players, with a self-recording tool to make participation lighter. Register interest →

This study was designed, conducted, coded, and written up by D. C. M. Langfield (also known as ‘Langfelagh’ and ‘Noutan’), a soon-to-graduate BSc Applied Psychology student at Glasgow Caledonian University, supervised by Dr. Alex Oliver. The formal dissertation is the empirical write-up; this interactive site is a personal extension of that project, built outside the submitted scope to make the data legible to a non-academic audience. Public dissemination of the analysed data was covered by the original ethics approval, with consent collected from every participant for that purpose.

Langfield's research interests sit at the intersection of psychology, neuroscience, and esports: how high-tempo competitive play can be used to enhance performance and to build cognitive reserve that protects long-term mental function across the lifespan.

Alongside the dissertation, Langfield works on tools in the same space: a planned self-think-aloud recording app for players (see /study), and other projects and research to come under @langfelagh on Instagram. Direct messages welcome on Discord at noutan_ctrl.

For academic correspondence, email dlangf200@caledonian.ac.uk. (This university address will be retired later in 2026 on graduation; check back here for an updated contact closer to then.)

Thank you to every participant for being lovely and generous with their time.

Thanks also to the following for their help with the recruitment drive:

• Sini and Grissom, from Voltaic
• Kovaak, owner of the Kovaaks aim trainer
• Tammas
• Elise “Soup” Cunningham, from the GCU Esports Society / Caledonian Chargers

And to:

• Dr. Alex Oliver, for his supervision and guidance throughout the project
• Dr. Jane Guiller, for the encouragement to push through