defmodule ChesstrainerWeb.RatingComponents do @moduledoc """ Visual guide to a user's rating. Places the rating on a fixed scale and shades the range the true rating is likely to sit in, derived from the rating deviation (rating ± 2×RD). The band is intentionally wide for a new or returning player: it narrows as rated attempts accumulate, which is exactly what the deviation measures. """ use Phoenix.Component # Glicko ratings are conventionally shown on a 0–3000 scale. The default # rating is 1500 and the deviation is capped at 350, so 500–3000 keeps the # useful range readable without letting a stray value push the marker off. @scale_min 500 @scale_max 3000 @doc """ Renders the rating gauge with its confidence band and explainer. """ attr :rating, :float, required: true attr :deviation, :float, default: nil attr :id, :string, default: "rating-guide" def rating_guide(assigns) do low = assigns.deviation && assigns.rating - 2 * assigns.deviation high = assigns.deviation && assigns.rating + 2 * assigns.deviation assigns = assign(assigns, low: low, high: high, scale_min: @scale_min, scale_max: @scale_max, marker_pct: pct(assigns.rating), band_left: low && pct(low), band_width: low && pct(high) - pct(low), label: aria_label(assigns.rating, low, high) ) ~H"""

Your rating: {round(@rating)}

likely {round(@low)}–{round(@high)}

{@scale_min} {@scale_max}

Your rating estimates how well you solve endgames. It moves after each rated attempt — solving a harder puzzle gains more, and replaying the same puzzle within a week does not count. The shaded band is the range your true rating is likely to sit in (±2 rating deviations); it narrows as you play.

""" end # Position on the gauge as a percentage, clamped to the drawn scale. defp pct(value) do clamped = value |> max(@scale_min) |> min(@scale_max) Float.round((clamped - @scale_min) / (@scale_max - @scale_min) * 100, 1) end defp aria_label(rating, nil, nil) do "Rating #{round(rating)} on a #{@scale_min} to #{@scale_max} scale" end defp aria_label(rating, low, high) do "Rating #{round(rating)}, likely between #{round(low)} and #{round(high)}, " <> "on a #{@scale_min} to #{@scale_max} scale" end end