diff --git a/examples/README.md b/examples/README.md
index 0297343..0312a72 100644
--- a/examples/README.md
+++ b/examples/README.md
@@ -646,52 +646,3 @@ and the matcher reproduces that order by assigning each shape to a slot, then su
out of N. Shows that "sequential" and "unconstrained" are one plugin with `targets` turned from
length-1 to length-N. Run it the same two-tab way.
-## `reference-game-hawkins.html` and `reference-game-cwg.html`
-
-Two **faithful replication** builds of the reference game, distinct from the two demos above: where
-`reference-game.html` / `reference-game-match.html` are lightweight teaching demos (synthetic shapes,
-short schedules), these two are **study artifacts** parameterised to reproduce a specific published
-paper, sharing the canonical 12-tangram stimulus set via [`tangrams.js`](tangrams.js). The tangrams in
-[`assets/tangrams/`](assets/tangrams/) are the exact set from
-[hawkrobe/tangrams](https://github.com/hawkrobe/tangrams) (Hawkins, Frank & Goodman 2020), themselves
-reproduced from Clark & Wilkes-Gibbs (1986) — so both builds are directly comparable to both papers.
-
-| | `reference-game-hawkins.html` | `reference-game-cwg.html` |
-| --- | --- | --- |
-| Replicates | Hawkins, Frank & Goodman (2020), *Cognitive Science* 44 | Clark & Wilkes-Gibbs (1986), *Cognition* 22:1–39 |
-| Condition | Sequential — **single** target, one click | Full-board — all **12** tangrams as an ordered target |
-| Schedule | 6 blocks × 12 = **72 trials** (`sequentialSchedule`) | **6 trials**, fresh full-board order each (`fullBoardSchedule`) |
-| Fidelity note | Near-exact (the paper was already web + text chat), including Exp. 2's matcher-click gate | Text chat **substitutes** for the original **spoken** dialogue — the one deliberate deviation, noted in-file |
-
-Both fix director/matcher for the whole game, reveal the target to the director only ("cued"), and use
-**unrestricted two-way chat** (`chat_role: "both"`, no count/length limits) — faithful to both papers,
-and load-bearing for the reference-shortening effect. `tangrams.js` builds each schedule
-**deterministically** from a seed, and both files seed from the shared session id (`?mp_session=`):
-each dyad therefore gets its own trial order while its two partners still agree on it without
-exchanging anything — a single fixed order across dyads would confound item order with repetition
-block, which is the axis both papers measure. Both files also **preload** the 12 tangram PNGs before
-pairing, since `rt` is a reported DV and `stimuli` is an object parameter that jsPsych's automatic
-media preloading can't see into. The faithful schedules are long for manual piloting — lower `BLOCKS`
-(Hawkins) / `TRIALS` (C&WG) at the top of each file for a quick smoke test, then restore.
-
-Both configs were verified parameter-by-parameter against the papers **and** the original
-[hawkrobe/tangrams](https://github.com/hawkrobe/tangrams) experiment code (which pins the details the
-prose leaves ambiguous — e.g. the code confirms director and matcher grids are *independently*
-scrambled, `notMatchingLocs`, and that the schedule is 72 rounds / 6 per tangram for the sequential
-build and `numRounds = 6` full boards for the unconstrained build). Both have been play-tested two-tab
-at full faithful schedule length, collecting both participants' data.
-
-The Hawkins build sets `require_message_before_response: true`, reproducing Exp. 2's gate on the
-matcher's click (`game.client.js`: `if (globalGame.messageSent)`) so a referring expression exists on
-every trial. The C&WG build leaves it off, matching its free-form protocol.
-
-> **`round_timeout` is not from either paper.** Both builds set one (60s Hawkins, 180s C&WG) purely so
-> a partner who disconnects or walks away cannot hang the trial forever — neither original was timed.
-> Rounds it ends are logged as `ended_by: "timeout"` with a null assignment, so they are easy to
-> exclude; raise the values if piloting shows genuine trials running long, especially C&WG's first
-> full-board trials, which are the longest in the study.
-
-Like the demos they run on `adapter-multiplayer-local` for two-tab piloting; the header comment marks
-the **one-line swap to `adapter-multiplayer-firebase`** (plus a real waiting room) for the paid
-Prolific run. Run them the same two-tab way (build packages, serve the repo, copy the `?mp_session=…`
-URL into a second tab).
diff --git a/examples/assets/tangrams/tangram_A.png b/examples/assets/tangrams/tangram_A.png
deleted file mode 100644
index 951c12a..0000000
Binary files a/examples/assets/tangrams/tangram_A.png and /dev/null differ
diff --git a/examples/assets/tangrams/tangram_B.png b/examples/assets/tangrams/tangram_B.png
deleted file mode 100644
index 49e6578..0000000
Binary files a/examples/assets/tangrams/tangram_B.png and /dev/null differ
diff --git a/examples/assets/tangrams/tangram_C.png b/examples/assets/tangrams/tangram_C.png
deleted file mode 100644
index 114e18c..0000000
Binary files a/examples/assets/tangrams/tangram_C.png and /dev/null differ
diff --git a/examples/assets/tangrams/tangram_D.png b/examples/assets/tangrams/tangram_D.png
deleted file mode 100644
index 34d89b3..0000000
Binary files a/examples/assets/tangrams/tangram_D.png and /dev/null differ
diff --git a/examples/assets/tangrams/tangram_E.png b/examples/assets/tangrams/tangram_E.png
deleted file mode 100644
index c11e77c..0000000
Binary files a/examples/assets/tangrams/tangram_E.png and /dev/null differ
diff --git a/examples/assets/tangrams/tangram_F.png b/examples/assets/tangrams/tangram_F.png
deleted file mode 100644
index 7138a3c..0000000
Binary files a/examples/assets/tangrams/tangram_F.png and /dev/null differ
diff --git a/examples/assets/tangrams/tangram_G.png b/examples/assets/tangrams/tangram_G.png
deleted file mode 100644
index fb9fba8..0000000
Binary files a/examples/assets/tangrams/tangram_G.png and /dev/null differ
diff --git a/examples/assets/tangrams/tangram_H.png b/examples/assets/tangrams/tangram_H.png
deleted file mode 100644
index 2c55c72..0000000
Binary files a/examples/assets/tangrams/tangram_H.png and /dev/null differ
diff --git a/examples/assets/tangrams/tangram_I.png b/examples/assets/tangrams/tangram_I.png
deleted file mode 100644
index 45acac0..0000000
Binary files a/examples/assets/tangrams/tangram_I.png and /dev/null differ
diff --git a/examples/assets/tangrams/tangram_J.png b/examples/assets/tangrams/tangram_J.png
deleted file mode 100644
index 93ab788..0000000
Binary files a/examples/assets/tangrams/tangram_J.png and /dev/null differ
diff --git a/examples/assets/tangrams/tangram_K.png b/examples/assets/tangrams/tangram_K.png
deleted file mode 100644
index 844801d..0000000
Binary files a/examples/assets/tangrams/tangram_K.png and /dev/null differ
diff --git a/examples/assets/tangrams/tangram_L.png b/examples/assets/tangrams/tangram_L.png
deleted file mode 100644
index 2a9496b..0000000
Binary files a/examples/assets/tangrams/tangram_L.png and /dev/null differ
diff --git a/examples/reference-game-cwg.html b/examples/reference-game-cwg.html
deleted file mode 100644
index 6653eae..0000000
--- a/examples/reference-game-cwg.html
+++ /dev/null
@@ -1,211 +0,0 @@
-
-
-
- Reference Game — Clark & Wilkes-Gibbs 1986 replication (full-board) — jsPsych multiplayer
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
diff --git a/examples/reference-game-hawkins.html b/examples/reference-game-hawkins.html
deleted file mode 100644
index c076124..0000000
--- a/examples/reference-game-hawkins.html
+++ /dev/null
@@ -1,214 +0,0 @@
-
-
-
- Reference Game — Hawkins et al. 2020 replication (sequential) — jsPsych multiplayer
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
diff --git a/examples/tangrams.js b/examples/tangrams.js
deleted file mode 100644
index 54b3402..0000000
--- a/examples/tangrams.js
+++ /dev/null
@@ -1,72 +0,0 @@
-// =================================================================================================
-// Canonical tangram stimuli + faithful round schedules, shared by the two reference-game replications
-// (reference-game-hawkins.html and reference-game-cwg.html).
-//
-// The 12 tangrams are the EXACT set from Hawkins, Frank & Goodman (2020), "Characterizing the Dynamics
-// of Learning in Repeated Reference Games" (github.com/hawkrobe/tangrams, tangram_A…L.png), which are
-// themselves reproduced from Clark & Wilkes-Gibbs (1986). Using this one set keeps both replications
-// directly comparable to both papers. The shapes are intentionally UNNAMED (ids A–L only): coining a
-// name for each hard-to-describe shape is the whole point of the task.
-// =================================================================================================
-(function (global) {
- const LETTERS = ["A", "B", "C", "D", "E", "F", "G", "H", "I", "J", "K", "L"];
-
- // Plugin stimulus objects: { id, src }. `src` is relative to the examples/ dir (works when served).
- const STIMULI = LETTERS.map((L) => ({ id: L, src: `assets/tangrams/tangram_${L}.png` }));
-
- // Small seeded PRNG (mulberry32). A schedule must be RANDOM-LOOKING yet DETERMINISTIC so that both
- // tabs/partners, loading the same seed, agree on the target sequence without exchanging it — and so
- // any given run is reproducible from its seed alone. Both examples pass the shared session id
- // (`?mp_session=`), so each dyad gets its own order while its two partners still agree; the default
- // seeds below are a same-order-every-time fallback for opening a file with no session in the URL.
- function mulberry32(seedStr) {
- let h = 1779033703 ^ seedStr.length;
- for (let i = 0; i < seedStr.length; i++) {
- h = Math.imul(h ^ seedStr.charCodeAt(i), 3432918353);
- h = (h << 13) | (h >>> 19);
- }
- let a = h >>> 0;
- return function () {
- a |= 0;
- a = (a + 0x6d2b79f5) | 0;
- let t = Math.imul(a ^ (a >>> 15), 1 | a);
- t = (t + Math.imul(t ^ (t >>> 7), 61 | t)) ^ t;
- return ((t ^ (t >>> 14)) >>> 0) / 4294967296;
- };
- }
- function shuffled(arr, rng) {
- const a = arr.slice();
- for (let i = a.length - 1; i > 0; i--) {
- const j = Math.floor(rng() * (i + 1));
- [a[i], a[j]] = [a[j], a[i]];
- }
- return a;
- }
-
- // Hawkins "sequential" schedule: `blocks` repetition blocks × 12 trials (72 total at blocks=6). Each
- // tangram is the SINGLE target exactly once per block; the within-block order is reshuffled each
- // block. Returns blocks*12 rounds of { round, targets: [id] } with unique, contiguous round indices.
- function sequentialSchedule(blocks, seed) {
- const rng = mulberry32(seed || "hawkins-sequential");
- const ids = STIMULI.map((s) => s.id);
- const rounds = [];
- let r = 0;
- for (let b = 0; b < blocks; b++) {
- for (const id of shuffled(ids, rng)) rounds.push({ round: r++, targets: [id] });
- }
- return rounds;
- }
-
- // Clark & Wilkes-Gibbs "full-board" schedule: `trials` trials (6 in the original), each a fresh
- // random ORDER of all 12 tangrams that the matcher must reproduce. Returns `trials` rounds of
- // { round, targets: [all 12 ids, ordered] }.
- function fullBoardSchedule(trials, seed) {
- const rng = mulberry32(seed || "cwg-fullboard");
- const ids = STIMULI.map((s) => s.id);
- const rounds = [];
- for (let r = 0; r < trials; r++) rounds.push({ round: r, targets: shuffled(ids, rng) });
- return rounds;
- }
-
- global.Tangrams = { STIMULI, sequentialSchedule, fullBoardSchedule };
-})(typeof globalThis !== "undefined" ? globalThis : this);