# Game design and UX/art-direction review

The player's goal is to build a defensible explanation of how a political advantage changed another actor's options, withstand source questions, and test that explanation against another historical case. The deliverable is a reasoned hearing record, not a changed election result. What makes this a game is the combination of an explicit role, a three-piece presentation constraint, evidence selection, challenge, revision, and transfer. Choosing which two observations support a claim and which complication deserves scarce public attention creates a meaningful decision. Defending, narrowing, or replacing that case can give those decisions consequences. Without a visible connection between those actions and the evolving case, the experience risks feeling like a guided research form.

This is an AI role simulation, not human user evidence. I reviewed only the supplied Player_Screens text/HTML packet, How_to_Play text, and permitted entry image. The HTML identifies disclosures, and the packet expands their content; I have not assumed all of it appears together. The image supports conclusions about the entry screen only. I did not test clicking, feedback, navigation, responsiveness, persistence, modes, or exports. Illustrative board selections and blank fields are not learner behavior or evidence of malfunction.

The design's strongest asset is intellectual agency. “A hearing, not a hidden answer,” an own-question route, all six cards remaining readable, and permission to keep a supported explanation resist answer hunting. Distinguishing observation from inference, complications from wholesale refutations, and shared documents from independent confirmation makes the constraint educationally productive. The library example and optional starters reduce preparation demands. Pair roles offer a useful social loop: propose, challenge, swap, brief. The suggested pace explicitly avoids a clock.

The load is nevertheless substantial: names and institutions, electoral units, source provenance, an actor/action/options connection, a tightly bounded paragraph, challenges, rival explanations, distinguishing evidence, and comparison. Disclosures reduce simultaneous clutter; they do not eliminate the need to remember how these pieces fit. Six numbered stages and “Next useful move” provide orientation. Whether that orientation survives movement between the board, sources, and writing remains untested.

## Eight concrete changes

### 1. Give case choice a research hook — High

**Screen/quote:** Entry, “Choose your main case,” currently emphasizes offices, dates, and “Known outcomes”; the headline asks “Who holds the advantage?”

Keep these facts, but add one short investigative hook to each case choice: Russia, “Reported backing and restricted challenge”; Malaysia, “An opposition victory despite incumbent advantages”; Malawi, “A reopened contest and a different winner”; Serbia, “Active alternatives and an uneven contest.” Make clear that these are questions to investigate. Currently a low-preparation learner can choose a country but has little basis for choosing a kind of problem until entering its brief. The hooks would let curiosity drive the first decision and clarify that advantage is not a single score. Preserve the different offices, periods, and measures; do not imply a country ranking or causal answer.

### 2. Make the attention constraint felt where cards compete — Medium

**Screen/quote:** Entry says “3 board spaces”; How to Play explains, “The limit is public attention.” Evidence says, “Choose what goes on your board.”

Bring the public-attention rationale onto the evidence screen and board: “You can present three pieces. What earns each space?” When replacing a card, show the outgoing and incoming titles and the player's existing notes together before the replacement is committed. This makes the opportunity cost tangible without restricting reading or demanding another essay. A full board should feel like a chosen presentation, not three required containers. Keep all six cards accessible, retain earlier choices as promised, and avoid claiming that replacement improves the explanation. The current packet offers replacement/history content; its actual transition feedback was not tested.

### 3. Direct source-review effort toward the chosen case — High

**Screen/quote:** Evidence, “Full cards opened: 0 of 6 / Source records opened: 0 of 6,” followed by “Opening a record does not prove you have read it.”

The qualification is sound, but aggregate counters can make opening disclosures look like collectible progress. Put review status beside each selected board card instead: full card opened, source record opened, and a route back to that record. Keep the complete-case counters secondary. At the point of selection, direct attention to one question: “Which step does this observation support or complicate?” Preserve all current source details, limitations, and the human reading confirmation. This changes the prominence of administrative feedback, not the evidence standard. The aim is to reward purposeful investigation through relevance and visibility, without adding points or pretending the system can verify comprehension.

### 4. Keep the player's connection visible while writing — Medium

**Screen/quote:** Connection asks “Who acted or held the advantage?”, “What did they do or hold?”, and “Whose options changed, and how?”, then introduces “Your explanation 60–85 words by the end.”

Display the player's three short answers as a compact connection strip immediately beside or above their current paragraph, with the optional inference visibly labeled as inference. Carry this strip into Hearing and Compare. Do not generate prose or fill missing links. Optional evidence attachments can remain optional, but attached card IDs should stay visible with the relevant step. This gives learners an object they can inspect and change, and makes the jump from separate fields to a concise explanation less memory dependent. It preserves authorship and the distinction between a recorded action, a changed option, and a possible broader effect.

### 5. Make revision the central payoff — High

**Screen/quote:** Hearing offers “Keep it, with a reason,” “Make a step more precise,” and “Change supporting evidence”; later it shows “Recorded hearing decisions.” The opening explanation “stays in the record.”

After recording a decision, show a restrained before/current view: first idea, current board, current connection or paragraph, and the player's recorded reason. Explicitly identify unchanged content as unchanged. A small notice such as “Supporting card changed; review whether your paragraph still fits” should connect a material action to the existing self-check. This is state feedback, not semantic judgment. At Finish, the same view can support “What changed from your first explanation?” without another required response. The text promises history and repeated current writing, but does not demonstrate this comparative payoff. Visible intellectual movement would strengthen the game's arc more than celebratory completion effects.

### 6. Challenge the closest plausible overreach — Medium

**Screen/quote:** Serbia Hearing asks whether “Vucic won presidential reelection here” and whether the government's rationale “independently verifies equal campaign conditions.”

That supplied claim combines an office error with a source-independence error; a player may reject it through elementary recall without confronting their explanation. Keep the short practice format, but give each question one consequential distinction. For RS3, for example: “Do separate campaign slots resolve the ordinary-news imbalance?” Ask which observation bears on that question and which missing link remains. Use the chosen card IDs to explain relevance, while retaining the honest statement that software did not discover this claim in the paragraph. Preserve demanding denominator and provenance checks. Stronger hearing tension comes from a credible question about the evidence, not from making an obviously excessive claim easier to dismiss.

### 7. Turn comparison selection into a testable intention — High

**Screen/quote:** Compare offers “Test a similar advantage across cases” and “Compare routes to a result”; the final comparison must be “30–45 words” and include period, action, affected option, difference, and an ID.

Before the second dossier, ask the player to choose the step they intend to compare: ballot access, organizing, a legal challenge, implementation, or conversion into office, with an own-step option. Show the main-case connection beside that intention. Then provide a compact planning row for second-case action, affected option, and a difference that matters. These can be short notes reused on screen, not extra polished writing. Keep the final word limit and full second-case evidence. This makes comparison an investigation with a purpose instead of another reading assignment followed by compressed summary. Do not automatically pronounce cases equivalent or infer a shared cause from similar outcomes.

### 8. Let the hearing artifact carry the atmosphere — Medium

**Screen/quote/image:** The entry shows a large document/courthouse collage, “Build a case. Face a serious question,” and a prominent dark panel, “Your hearing is here,” with resume controls.

The cream ground, dark green, restrained red, serif headline, and physical-paper imagery establish seriousness and should remain. In this saved-session image, the large hero precedes the useful resume panel; place a compact continuation panel immediately below the header for returning players. For immersion deeper in play, use a consistent, clearly fictional hearing agenda and an evolving presentation-board identity rather than adding decorative institutional imagery. Source labels, dates, and the player's own revised notes can supply the visual texture. Surface the How to Play role-swap cue at the relevant pair hearing moment if it is not already present in that untested mode. Preserve calm pacing, legibility, evidence access, and honest limits: no simulated electoral effects, countdown pressure, or rewards for adopting particular conclusions.
