Release pass: wording, colour-blind-safe colours, map visibility, phone layout

Walked every tab on desktop and phone, clicked every link and button:
- Colours: new map inks and page colours, checked with simulated
  protanopia/deuteranopia/tritanopia (worst pair ~12 CIEDE2000, was 1.4 for
  Red/Green on the map and 9.7 for Orange/Brown on the page). Map label text
  is always dark ink; dragon colour goes on borders. docs/COLOURS.md and
  research/python/check_colours.py.
- Valley map: dragons get a paper halo and a visibility floor (faded ones never
  drop below 0.3; bonding dragons start at 0.5); "no shadow" roll labelled
  correctly.
- Wording: methods updated (strata, the newer form's questions, missing
  answers, a new section on the solver, The Odds cards and genuine attempts);
  findings use readable factor names and no longer claim a date adjustment;
  'Load a newer route export above'; code links include solver.js and
  explorer.js; The Odds shows 93% instead of '1 in 1'.
- Deep Dive route map treats all six dragons alike ('Bond a dragon' row,
  default 'Bond any dragon'; rare sightings separate).
- Phone: Deep Dive no longer overflows (sections and selects can shrink).
- scripts/qa/crawl.js and mapcheck.js for future release checks.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UCzTd31GphpMXDhQqWrCMQ
This commit is contained in:
ekcSchiano
2026-11-02 17:57:51 +00:00
co-authored by Claude Opus 5.5
parent 0ac5dfd971
commit 7366af2f7d
11 changed files with 236 additions and 27 deletions
+9 -8
View File
@@ -9,7 +9,7 @@
const state = {
raw: null, bundled: null, allRuns: [], runs: [], res: null, missing: [],
filters: { source: "all", goal: "all", excludeFlagged: false },
target: "bond_Blue", widthMode: "traffic", minN: 10, conds: {},
target: "bond_any", widthMode: "traffic", minN: 10, conds: {},
fxOutcome: "blue_meadow", fxFactor: "resolution", rtTarget: "see_Blue", rtMin: 10,
};
@@ -148,7 +148,7 @@
const biasShown = chaseB && bulkB && chaseB.n >= 20 && bulkB.n >= 20 && chaseB.lo > bulkB.hi;
let t;
if (biasShown) {
t = `<p><b>The meadow Blue appeared in ${rateTxt(cb.overall)} of logged meadow visits, but that figure is inflated.</b> Players chasing a Blue report it in ${pct(chaseB.p)} of their visits. Bulk logs, which record every run, show ${rateTxt(bulkB)}. The game cannot see what you are chasing, so that gap comes from which runs get reported.</p>`;
t = `<p><b>The meadow Blue appeared in ${rateTxt(cb.overall)} of logged meadow visits, but that figure is inflated.</b> Players chasing a rare dragon report it in ${pct(chaseB.p)} of their visits. Bulk logs, which record every run, show ${rateTxt(bulkB)}. The game cannot see what you are chasing, so that gap comes from which runs get reported.</p>`;
t += `<p>A realistic expectation is nearer the bulk-log rate, about one meadow visit in ${Math.round(1 / Math.max(bulkB.p, 0.001))}, than the headline figure.${otherB ? ` Form players who were not chasing a Blue report ${pct(otherB.p)}, which still carries some selective reporting.` : ""}</p>`;
} else {
t = `<p><b>The meadow Blue appeared in ${rateTxt(cb.overall)} of logged meadow visits in the current selection.</b> Compare the data-source filter settings to see how much this depends on who logged the runs.</p>`;
@@ -161,9 +161,9 @@
const rk = E.rankRoutes(state.runs, "see_Blue", state.rtMin);
const adjSig = rk.rows.filter((x) => x.mh_q !== undefined && x.mh_q < 0.05);
let t = sig.length
? `<p><b>After adjusting for reporting bias, ${sig.length === 1 ? "one recorded choice was" : sig.length + " recorded choices were"} linked to the Blue appearing:</b> ${sig.map((s) => `${esc(s.term)} (OR ${fmtOR(s.or)}, 95% CI ${fmtOR(s.lo)}–${fmtOR(s.hi)}, p = ${fmtP(s.p)})`).join("; ")}.</p>`
? `<p><b>After adjusting for reporting bias, ${sig.length === 1 ? "one recorded choice was" : sig.length + " recorded choices were"} linked to the Blue appearing:</b> ${sig.map((s) => `${esc(labelFactor(s.factor))}: ${esc(s.level)} (OR ${fmtOR(s.or)}, 95% CI ${fmtOR(s.lo)}–${fmtOR(s.hi)}, p = ${fmtP(s.p)})`).join("; ")}.</p>`
: `<p><b>No recorded route choice changed the chance of the Blue appearing once reporting bias is held fixed.</b> Direction, first sighting, how you handled the cadet and who you entered with all have adjusted odds ratios whose 95% intervals include 1 (n = ${mb.n} meadow visits).</p>`;
t += `<p>Of ${rk.n_routes_tested} full routes with at least ${state.rtMin} meadow visits, ${adjSig.length} beat the others after adjustment and correction for multiple testing. The "precise pathways" shared on Reddit may still matter through choices the form does not record, such as how you react to the first dragon.</p>`;
t += `<p>Of ${rk.n_routes_tested} full routes with at least ${state.rtMin} meadow visits, ${adjSig.length} beat the others after adjustment and correction for multiple testing. The "precise pathways" shared on Reddit may still matter through choices the form does not record, such as the answers given inside each encounter.</p>`;
out.push(["blue", "Blue", t]);
}
if (encBlue && encBlue.overall.n) {
@@ -186,7 +186,7 @@
const ms = r.models.shadow;
let t = `<p><b>Shadow leads to test, not established:</b> ${sh.map((x) => `<i>${esc(labelFactor(x.factor))}: ${esc(x.level)}</i> had the shadow in ${rateTxt(x)} vs the rest (OR ${fmtOR(x.or)}, q = ${fmtP(x.q)})`).join("; ")}.</p>`;
const att = ms && ms.terms.find((tm) => tm.factor === "attempt" && tm.level === "First attempt");
if (att) t += `<p>The first-attempt pattern survives adjustment for source, goal, direction and date (OR ${fmtOR(att.or)}, 95% CI ${fmtOR(att.lo)}–${fmtOR(att.hi)}). That matches a theory on Reddit that a fresh account rolls the shadow more often, but players who strike lucky on a first run may also be more likely to submit the form.</p>`;
if (att) t += `<p>The first-attempt pattern survives adjustment for data source, chase goal and direction (OR ${fmtOR(att.or)}, 95% CI ${fmtOR(att.lo)}–${fmtOR(att.hi)}). That matches a theory on Reddit that a fresh account rolls the shadow more often, but players who strike lucky on a first run may also be more likely to submit the form.</p>`;
t += `<p>The cheapest test is a bulk log that records every tree climb, including the dull ones, with these choices noted.</p>`;
out.push(["lead", "Lead", t]);
}
@@ -208,14 +208,15 @@
const labelFactor = (f) => FACTOR_LABEL[f] || f;
// ------------------------------------------------------------ route map
const RARE_T = ["bond_Blue", "bond_Black", "see_Blue", "see_shadow"];
const OTHER_T = ["bond_Red", "bond_Green", "bond_Orange", "bond_Brown", "bond_any"];
// first row: bonding each dragon (all six treated alike); second row: the two rare sightings
const RARE_T = ["bond_Red", "bond_Orange", "bond_Brown", "bond_Green", "bond_Blue", "bond_Black", "bond_any"];
const OTHER_T = ["see_Blue", "see_shadow"];
function renderTargets() {
const chip = (t, featured) => {
const c = E.TARGETS[t].color ? DRAGON_VAR[E.TARGETS[t].color] : "--dr-any";
return `<button type="button" class="tchip${featured ? " featured" : ""}" data-t="${t}" aria-pressed="${state.target === t}" style="--c: var(${c})"><span class="sw"></span>${esc(E.TARGETS[t].label)}</button>`;
};
$("#targets-rare").innerHTML = RARE_T.map((t) => chip(t, t.startsWith("bond"))).join("");
$("#targets-rare").innerHTML = RARE_T.map((t) => chip(t, false)).join("");
$("#targets-other").innerHTML = OTHER_T.map((t) => chip(t, false)).join("");
$$(".tchip[data-t]").forEach((b) => b.addEventListener("click", () => { state.target = b.dataset.t; $("#best-routes").innerHTML = ""; renderTargets(); renderMap(); }));
}