From e0b446bbfb3642222f5f720323e7a351700d12c0 Mon Sep 17 00:00:00 2001 From: Owen Rees Date: Tue, 17 Nov 2026 14:28:24 +0100 Subject: [PATCH] Weight puzzle selection and stop repeats inside a session (#183) MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Selection took a hard rating band (medium = user rating ± 100), removed everything attempted in the last 30 days, and drew uniformly from what was left. Because the bulk importer gives every endgame in a batch the same rating, the band held either the whole library or — once a user's rating drifted past it — only the endgames that user had already played. With `fresh` empty, the fallback re-drew from the same band ignoring recency, so recently played puzzles came straight back. An empty band used a deterministic nearest-match, alternating between the same two puzzles. Difficulty is now a bias rather than a filter: candidates are weighted by `exp(-|rating - target| / spread)` and drawn with an exponential race, so the whole library stays reachable and a thin region around the user's rating can never empty the pool. `easy` and `hard` keep one-sided guards so the labels stay honest. When every candidate is excluded, selection serves the least recently attempted endgame instead of a random recent one. Sessions now carry the ids they have served, which fixes the observed in-session repeat (only the current puzzle used to be excluded), and guest training gets the same treatment. The 30-day window is configurable as the repeat window and is now named in GLOSSARY.md, distinct from the rating window and from the tablebase cooldown. Difficulty copy is qualitative, since the band numbers no longer describe how selection works. --- GLOSSARY.md | 20 ++- config/config.exs | 15 +- lib/chesstrainer/attempts.ex | 4 +- lib/chesstrainer/puzzles.ex | 143 +++++++++++------- .../live/endgame_live/guest_train.ex | 22 ++- .../live/endgame_live/train.ex | 24 +-- test/chesstrainer/puzzles_test.exs | 87 +++++++---- .../live/endgame_live/train_test.exs | 12 ++ 8 files changed, 212 insertions(+), 115 deletions(-) diff --git a/GLOSSARY.md b/GLOSSARY.md index 90b7884..7f22a54 100644 --- a/GLOSSARY.md +++ b/GLOSSARY.md @@ -105,6 +105,14 @@ carries a start time and a rating baseline for the timer and rating-gain monitor. Ephemeral — it lives only in the running LiveView, never persisted. _Avoid_: Run, game, attempt +**Repeat window**: +The minimum interval before the same endgame may be served to the same user +again. Selection skips endgames attempted inside the window; only when nothing +else is available does the least recently attempted endgame come back. It is +independent of the rating window, which decides whether a replay may move a +rating. +_Avoid_: Cooldown (reserved for tablebase rate-limiting), due date + ## Ratings **Rating**: @@ -128,11 +136,13 @@ systems; the trainer deliberately does not track it. _Avoid_: Variance, inconsistency **Difficulty**: -A choice — `easy`, `medium` or `hard` — that selects which puzzles to serve -relative to the current user's rating: `medium` is near the user's rating, -`easy` below it, `hard` above it. The band width is a configured constant. It is -a property of a session, not of an endgame — the same endgame is easy for one -user and hard for another. +A choice — `easy`, `medium` or `hard` — that biases which puzzles are served +relative to the current user's rating: `medium` centres on the user's rating, +`easy` below it, `hard` above it. Candidates are weighted by how close their +rating is to that centre, so a session draws mostly from near the target but the +whole library stays reachable. The spread is a configured constant. It is a +property of a session, not of an endgame — the same endgame is easy for one user +and hard for another. _Avoid_: Level, rank, tier ## Infrastructure diff --git a/config/config.exs b/config/config.exs index 9991f63..6db7d31 100644 --- a/config/config.exs +++ b/config/config.exs @@ -61,11 +61,16 @@ config :chesstrainer, :ratings, window_days: 7, rd_inflation_c: 10.0 -# Difficulty bands for puzzle sessions. An endgame is "medium" when its rating -# falls within `band_width` of the user's rating, "easy" below that band and -# "hard" above it. Tweak to widen or narrow the served difficulty (e.g. 150 or -# 200). -config :chesstrainer, :puzzles, band_width: 100 +# Puzzle selection. Difficulty is a weighted bias towards a target rating — +# `medium` targets the user's rating, `easy` 200 below it and `hard` 200 above +# it — and `spread` is how many rating points it takes for a puzzle's chance to +# fall off by a factor of e. `repeat_window_days` is how long an attempted +# puzzle stays out of the pool. Tweak `spread` to widen or narrow the served +# difficulty (e.g. 100 or 250). +config :chesstrainer, :puzzles, + spread: 150, + repeat_window_days: 30, + targets: [easy: -200, medium: 0, hard: 200] # Public codes of the endgames pinned to the landing-page puzzle grid, in # display order. When empty (or short) the grid is topped up with a date-seeded diff --git a/lib/chesstrainer/attempts.ex b/lib/chesstrainer/attempts.ex index 74f3766..4b4fc31 100644 --- a/lib/chesstrainer/attempts.ex +++ b/lib/chesstrainer/attempts.ex @@ -106,8 +106,8 @@ defmodule Chesstrainer.Attempts do @doc """ The distinct endgame ids `user_id` has attempted in the last `days` days. - Used by puzzle selection to prefer fresh puzzles; a retry still counts as an - attempt, even though the rating window leaves it unrated. + Used by puzzle selection to skip puzzles inside the repeat window; a retry + still counts as an attempt, even though the rating window leaves it unrated. """ def recent_endgame_ids(user_id, days \\ 30) do cutoff = DateTime.add(DateTime.utc_now(), -days, :day) diff --git a/lib/chesstrainer/puzzles.ex b/lib/chesstrainer/puzzles.ex index 9b13ac0..0b241d1 100644 --- a/lib/chesstrainer/puzzles.ex +++ b/lib/chesstrainer/puzzles.ex @@ -2,14 +2,21 @@ defmodule Chesstrainer.Puzzles do @moduledoc """ Selects endgames to serve as puzzles. - A puzzle is played at one of three difficulties — `:easy`, `:medium` or - `:hard` — defined relative to the current user's rating: `:medium` is within - a configurable band of the user's rating, `:easy` below it and `:hard` above - it. The same endgame is therefore easy for one player and hard for another. + Difficulty is a bias, not a filter. Every endgame is weighted by how close its + rating is to a target derived from the user's rating — the user's own rating + for `:medium`, below it for `:easy` and above it for `:hard` — and one + candidate is drawn from that weighted distribution. Nearby endgames are much + more likely, but the whole library stays reachable, so a thin region around + the user's rating can never leave the pool empty. - Selection prefers endgames the user has not attempted in the last 30 days. - When that pool is empty it falls back to any endgame in the band; the rating - window keeps a recent replay unrated until it expires. + Endgames the user has attempted inside the repeat window are skipped, as is + anything passed in `:exclude_ids` (the puzzles already served in a session). + Only when every candidate is excluded does selection fall back to the least + recently attempted endgame, so the repeat is the most overdue one rather than + an arbitrary one. + + The same endgame is therefore easy for one player and hard for another: a + rating is what makes the two comparable. """ import Ecto.Query, warn: false @@ -17,95 +24,115 @@ defmodule Chesstrainer.Puzzles do alias Chesstrainer.Repo alias Chesstrainer.Accounts.User alias Chesstrainer.Attempts + alias Chesstrainer.Attempts.Attempt alias Chesstrainer.Endgames.Endgame - @default_band_width 100 + @default_spread 150 + @default_repeat_window_days 30 + @default_targets [easy: -200, medium: 0, hard: 200] @difficulties [:easy, :medium, :hard] @doc "The difficulties a session can be played at." def difficulties, do: @difficulties @doc """ - The half-width of the medium band, in rating points. + How far, in rating points, a puzzle's likelihood falls off from the target. - Configured under `:chesstrainer, :puzzles, :band_width`. + A larger spread draws more widely around the target. Configured under + `:chesstrainer, :puzzles, :spread`. """ - def band_width do - :chesstrainer - |> Application.get_env(:puzzles, []) - |> Keyword.get(:band_width, @default_band_width) - end + def spread, do: setting(:spread, @default_spread) @doc """ - The `{min, max}` rating range for `difficulty` relative to `rating`. + The minimum gap, in days, before the same endgame may be served again. - A `nil` bound is open-ended. + Configured under `:chesstrainer, :puzzles, :repeat_window_days`. """ - def difficulty_range(rating, difficulty) when difficulty in @difficulties do - width = band_width() + def repeat_window_days, do: setting(:repeat_window_days, @default_repeat_window_days) - case difficulty do - :easy -> {nil, rating - width} - :medium -> {rating - width, rating + width} - :hard -> {rating + width, nil} - end + @doc """ + The rating `difficulty` aims at, given the player's `rating`. + + The offsets are configured under `:chesstrainer, :puzzles, :targets`. + """ + def difficulty_target(rating, difficulty) when difficulty in @difficulties do + rating + Keyword.fetch!(setting(:targets, @default_targets), difficulty) end @doc """ Picks the next puzzle for `user` at `difficulty`, using `user.rating` as the baseline. - `:exclude_ids` (default `[]`) skips endgames by id — used to avoid repeating - the puzzle just played. Returns `nil` when no endgame exists at all. + `:exclude_ids` (default `[]`) skips endgames by id — used for the puzzles + already served in this session. Returns `nil` when no endgame exists at all. """ def next_endgame(%User{} = user, difficulty, opts \\ []) when difficulty in @difficulties do exclude_ids = Keyword.get(opts, :exclude_ids, []) - band = difficulty_range(user.rating, difficulty) - recent_ids = Attempts.recent_endgame_ids(user.id) + recent_ids = Attempts.recent_endgame_ids(user.id, repeat_window_days()) + target = difficulty_target(user.rating, difficulty) - fresh = - Endgame - |> in_band(band) - |> skip_ids(exclude_ids) - |> skip_ids(recent_ids) - |> random_one() - - fresh || fallback(user.rating, band, exclude_ids) - end - - defp fallback(rating, band, exclude_ids) do Endgame - |> in_band(band) - |> skip_ids(exclude_ids) - |> random_one() + |> on_side_of(user.rating, difficulty) + |> skip_ids(exclude_ids ++ recent_ids) + |> weighted_one(target) |> case do - nil -> nearest(rating, exclude_ids) + nil -> least_recently_attempted(user.id, exclude_ids) endgame -> endgame end end - defp nearest(rating, exclude_ids) do + # Keep the labels honest: "easy" never serves above the user's rating plus the + # spread, and "hard" never below it minus the spread. "medium" is open, which + # is what lets it reach the rest of the library. + defp on_side_of(query, _rating, :medium), do: query + + defp on_side_of(query, rating, :easy), + do: where(query, [e], e.rating <= ^(rating + spread())) + + defp on_side_of(query, rating, :hard), + do: where(query, [e], e.rating >= ^(rating - spread())) + + # Draw one row with probability proportional to + # `exp(-|rating - target| / spread)`: an exponential race, where the row with + # the smallest `-ln(U) / weight` wins. + defp weighted_one(query, target) do + spread = spread() * 1.0 + target = target * 1.0 + + query + |> order_by( + [e], + fragment("-ln(1 - random()) * exp(abs(? - ?) / ?)", e.rating, ^target, ^spread) + ) + |> limit(1) + |> Repo.one() + end + + # Every candidate is excluded, so the repeat window is overruled: serve + # whatever the user has waited longest to see again. + defp least_recently_attempted(user_id, exclude_ids) do + attempted = + Attempt + |> where([a], a.user_id == ^user_id) + |> group_by([a], a.endgame_id) + |> select([a], %{endgame_id: a.endgame_id, last_at: max(a.inserted_at)}) + Endgame |> skip_ids(exclude_ids) - |> order_by([e], asc: fragment("ABS(? - ?)", e.rating, ^rating)) + |> join(:left, [e], a in subquery(attempted), on: a.endgame_id == e.id) + |> order_by([_e, a], asc_nulls_first: a.last_at) + |> order_by([e], fragment("random()")) |> limit(1) |> Repo.one() end - defp random_one(query) do - query - |> order_by([e], fragment("RANDOM()")) - |> limit(1) - |> Repo.one() - end - - defp in_band(query, {nil, max}), do: where(query, [e], e.rating <= ^max) - defp in_band(query, {min, nil}), do: where(query, [e], e.rating >= ^min) - - defp in_band(query, {min, max}), - do: where(query, [e], e.rating >= ^min and e.rating <= ^max) - defp skip_ids(query, []), do: query defp skip_ids(query, ids), do: where(query, [e], e.id not in ^ids) + + defp setting(key, default) do + :chesstrainer + |> Application.get_env(:puzzles, []) + |> Keyword.get(key, default) + end end diff --git a/lib/chesstrainer_web/live/endgame_live/guest_train.ex b/lib/chesstrainer_web/live/endgame_live/guest_train.ex index 885fdbb..84408e3 100644 --- a/lib/chesstrainer_web/live/endgame_live/guest_train.ex +++ b/lib/chesstrainer_web/live/endgame_live/guest_train.ex @@ -135,7 +135,10 @@ defmodule ChesstrainerWeb.EndgameLive.GuestTrain do @impl true def mount(_params, _session, socket) do - socket = assign(socket, :limit, GuestLimits.limit()) + socket = + socket + |> assign(:limit, GuestLimits.limit()) + |> assign(:served_ids, []) cond do signed_in?(socket) -> @@ -217,11 +220,7 @@ defmodule ChesstrainerWeb.EndgameLive.GuestTrain do |> assign(:limit_reached, retry_ms) {:ok, _remaining} -> - endgame = - Endgames.random_endgame(exclude_ids: [socket.assigns.endgame.id]) || - Endgames.random_endgame() - - case endgame do + case random_unserved(socket) do nil -> assign(socket, :empty, :exhausted) endgame -> start_guest_puzzle(socket, socket.assigns.guest_address, endgame) end @@ -234,13 +233,21 @@ defmodule ChesstrainerWeb.EndgameLive.GuestTrain do {:ok, limit_wall(socket, address, retry_ms)} {:ok, _remaining} -> - case Endgames.random_endgame() do + case random_unserved(socket) do nil -> {:ok, empty_pool(socket, address)} endgame -> {:ok, start_guest_puzzle(socket, address, endgame)} end end end + # Prefer an endgame this visitor has not seen in this session. Once the whole + # pool is served — only possible on a very long guest session — start over + # rather than dead-end. + defp random_unserved(socket) do + Endgames.random_endgame(exclude_ids: socket.assigns.served_ids) || + Endgames.random_endgame() + end + defp start_guest_puzzle(socket, address, endgame) do case GuestLimits.take(address) do {:ok, remaining} -> @@ -250,6 +257,7 @@ defmodule ChesstrainerWeb.EndgameLive.GuestTrain do |> assign(:guest_remaining, remaining) |> assign(:limit_reached, nil) |> assign(:empty, nil) + |> assign(:served_ids, [endgame.id | socket.assigns.served_ids]) |> PlayCore.init(endgame, page_title: "Train") {:limit_reached, retry_ms} -> diff --git a/lib/chesstrainer_web/live/endgame_live/train.ex b/lib/chesstrainer_web/live/endgame_live/train.ex index f2b2738..e93d204 100644 --- a/lib/chesstrainer_web/live/endgame_live/train.ex +++ b/lib/chesstrainer_web/live/endgame_live/train.ex @@ -106,7 +106,7 @@ defmodule ChesstrainerWeb.EndgameLive.Train do id="pool-exhausted" class="rounded-xl border border-hairline bg-canvas-raised p-6" > -

You've finished every puzzle at this difficulty

+

You've played every puzzle available right now

There are no more {difficulty_label(@session.difficulty)} puzzles right now. Try again, or pick another difficulty. @@ -161,7 +161,7 @@ defmodule ChesstrainerWeb.EndgameLive.Train do

Choose a difficulty

- Difficulty bands are ±{Puzzles.band_width()} around your rating. + Easy draws easier endgames, hard draws harder ones — your rating sets the centre.

<.link @@ -190,7 +190,7 @@ defmodule ChesstrainerWeb.EndgameLive.Train do
- <%= for {difficulty, label, description} <- difficulty_options(@user_rating) do %> + <%= for {difficulty, label, description} <- difficulty_options() do %> <.button id={"difficulty-#{difficulty}"} phx-click="start_session" @@ -315,7 +315,8 @@ defmodule ChesstrainerWeb.EndgameLive.Train do started_at_ms: System.system_time(:millisecond), baseline: socket.assigns.user_rating, count: 0, - sum_ms: 0 + sum_ms: 0, + served_ids: [endgame.id] } socket @@ -333,13 +334,16 @@ defmodule ChesstrainerWeb.EndgameLive.Train do if session do case Puzzles.next_endgame(selection_user(socket), session.difficulty, - exclude_ids: [socket.assigns.endgame.id] + exclude_ids: session.served_ids ) do nil -> assign(socket, :pool, %{difficulty: session.difficulty}) endgame -> + session = %{session | served_ids: [endgame.id | session.served_ids]} + socket + |> assign(:session, session) |> assign(:pool, nil) |> assign(:puzzle_started_mono, System.monotonic_time(:millisecond)) |> assign(:puzzle_counted, false) @@ -395,13 +399,11 @@ defmodule ChesstrainerWeb.EndgameLive.Train do defp parse_difficulty(_value), do: :medium - defp difficulty_options(rating) do - width = Puzzles.band_width() - + defp difficulty_options do [ - {:easy, "Easy", "below #{round(rating - width)}"}, - {:medium, "Medium", "#{round(rating - width)}–#{round(rating + width)}"}, - {:hard, "Hard", "above #{round(rating + width)}"} + {:easy, "Easy", "easier than your rating"}, + {:medium, "Medium", "around your rating"}, + {:hard, "Hard", "harder than your rating"} ] end diff --git a/test/chesstrainer/puzzles_test.exs b/test/chesstrainer/puzzles_test.exs index b60f174..29d5aef 100644 --- a/test/chesstrainer/puzzles_test.exs +++ b/test/chesstrainer/puzzles_test.exs @@ -7,6 +7,7 @@ defmodule Chesstrainer.PuzzlesTest do alias Chesstrainer.Accounts.Scope alias Chesstrainer.Attempts alias Chesstrainer.Puzzles + alias Chesstrainer.Repo # The shared `endgame_fixture/1` picks one of six FENs at random-ish, which # can collide when one test creates several endgames under async load. Pin @@ -22,71 +23,103 @@ defmodule Chesstrainer.PuzzlesTest do endgame_fixture(fen: Enum.fetch!(@fens, index), rating: rating) end - describe "difficulty_range/2" do - test "is centred on the rating and open-ended at the extremes" do - assert Puzzles.difficulty_range(1500, :medium) == {1400, 1600} - assert Puzzles.difficulty_range(1500, :easy) == {nil, 1400} - assert Puzzles.difficulty_range(1500, :hard) == {1600, nil} + defp attempt_days_ago(user, endgame, days) do + {:ok, attempt} = Attempts.record_attempt(Scope.for_user(user), endgame, :solved) + + attempt + |> Ecto.Changeset.change( + inserted_at: DateTime.utc_now() |> DateTime.add(-days, :day) |> DateTime.truncate(:second) + ) + |> Repo.update!() + end + + describe "difficulty_target/2" do + test "centres on the rating and offsets easy and hard" do + assert Puzzles.difficulty_target(1500, :medium) == 1500 + assert Puzzles.difficulty_target(1500, :easy) == 1300 + assert Puzzles.difficulty_target(1500, :hard) == 1700 + end + end + + describe "selection settings" do + test "uses the configured spread and repeat window" do + assert Puzzles.spread() == 150 + assert Puzzles.repeat_window_days() == 30 end end describe "next_endgame/3" do - test "medium picks an endgame within the band" do + test "draws far more often from near the user's rating" do user = user_fixture() - in_band = endgame_at(1550, 0) - _out_of_band = endgame_at(1900, 1) + near = endgame_at(1500, 0) + _far = endgame_at(4000, 1) - assert Puzzles.next_endgame(user, :medium).id == in_band.id + for _ <- 1..50 do + assert Puzzles.next_endgame(user, :medium).id == near.id + end end - test "easy picks an endgame below the user's rating" do + test "easy never serves above the user's rating plus the spread" do user = user_fixture() easy = endgame_at(1300, 0) - _medium = endgame_at(1500, 1) + _too_hard = endgame_at(3000, 1) assert Puzzles.next_endgame(user, :easy).id == easy.id end - test "hard picks an endgame above the user's rating" do + test "hard never serves below the user's rating minus the spread" do user = user_fixture() hard = endgame_at(1700, 0) - _medium = endgame_at(1500, 1) + _too_easy = endgame_at(100, 1) assert Puzzles.next_endgame(user, :hard).id == hard.id end - test "falls back to the nearest endgame when the band is empty" do - user = user_fixture() - only = endgame_at(2200, 0) - - assert Puzzles.next_endgame(user, :medium).id == only.id - end - - test "prefers endgames not attempted in the last 30 days" do + test "prefers endgames not attempted inside the repeat window" do user = user_fixture() attempted = endgame_at(1500, 0) fresh = endgame_at(1500, 1) - {:ok, _} = Attempts.record_attempt(Scope.for_user(user), attempted, :solved) + attempt_days_ago(user, attempted, 1) assert Puzzles.next_endgame(user, :medium).id == fresh.id end - test "falls back to a recent endgame once the fresh pool is empty" do + test "serves an endgame again once its attempt is outside the repeat window" do user = user_fixture() only = endgame_at(1500, 0) - {:ok, _} = Attempts.record_attempt(Scope.for_user(user), only, :solved) + attempt_days_ago(user, only, 31) assert Puzzles.next_endgame(user, :medium).id == only.id end - test "skips excluded ids" do + test "falls back to the least recently attempted endgame when all are excluded" do user = user_fixture() - excluded = endgame_at(1500, 0) + older = endgame_at(1500, 0) + newer = endgame_at(1500, 1) + + attempt_days_ago(user, older, 20) + attempt_days_ago(user, newer, 10) + + assert Puzzles.next_endgame(user, :medium).id == older.id + end + + test "skips endgames already served in the session" do + user = user_fixture() + served = endgame_at(1500, 0) other = endgame_at(1500, 1) - assert Puzzles.next_endgame(user, :medium, exclude_ids: [excluded.id]).id == other.id + assert Puzzles.next_endgame(user, :medium, exclude_ids: [served.id]).id == other.id + end + + test "returns nil when every endgame has been served and attempted" do + user = user_fixture() + only = endgame_at(1500, 0) + + attempt_days_ago(user, only, 1) + + assert Puzzles.next_endgame(user, :medium, exclude_ids: [only.id]) == nil end test "returns nil when no endgames exist" do diff --git a/test/chesstrainer_web/live/endgame_live/train_test.exs b/test/chesstrainer_web/live/endgame_live/train_test.exs index 6e5d887..889f23e 100644 --- a/test/chesstrainer_web/live/endgame_live/train_test.exs +++ b/test/chesstrainer_web/live/endgame_live/train_test.exs @@ -170,6 +170,18 @@ defmodule ChesstrainerWeb.EndgameLive.TrainTest do assert Enum.any?(endgames, &(&1.id == second.id)) end + test "does not serve the same puzzle twice in a session", %{conn: conn} do + {view, [only]} = start_medium_session(conn, 1) + + solve_current(view) + view |> element("#next-puzzle-button") |> render_click() + + assigns = :sys.get_state(view.pid).socket.assigns + assert has_element?(view, "#pool-exhausted") + assert assigns.endgame.id == only.id + assert assigns.session.served_ids == [only.id] + end + test "keeps next puzzle available while reviewing", %{conn: conn} do {view, endgames} = start_medium_session(conn, 2) first = :sys.get_state(view.pid).socket.assigns.endgame