{{ block title }}How the Trials Work{{ endblock }}

{{ block content }}
<style>
    .task-instructions {
        max-width: 680px;
        margin: 0 auto;
        font-family: Georgia, 'Times New Roman', serif;
        line-height: 1.7;
        color: #2c3e50;
    }
    .task-instructions h2 {
        font-family: 'Helvetica Neue', Arial, sans-serif;
        color: #1a1a2e;
        font-weight: 600;
        margin-top: 1.5em;
    }
    .task-instructions h2:first-child { margin-top: 0; }
    .highlight-box {
        background: #f0f4f8;
        border-left: 4px solid #3498db;
        padding: 1em 1.2em;
        margin: 1.5em 0;
        border-radius: 0 4px 4px 0;
    }
    .key-point {
        background: #fef9e7;
        border-left: 4px solid #f39c12;
        padding: 1em 1.2em;
        margin: 1.5em 0;
        border-radius: 0 4px 4px 0;
    }
</style>

<div class="task-instructions">

    <p>
        Now that you have seen how the Plinko board works, here are the details
        of the experiment.
    </p>

    <h2>The Board</h2>
    <p>
        In the demo, you used a {{ R_DEMO }}-row board. The actual trials use a longer
        <strong>{{ R }}-row</strong> board, so the balls will scatter more widely. The
        board has columns numbered from <strong>0 to 1000</strong>. On each trial,
        the drop position is chosen uniformly at random from these columns, so every
        position from 0 to 1000 is equally likely. The drop position range is much wider than
        what you saw in the demo.
    </p>

    <h2>Signals</h2>
    <p>
        As you saw in the demo, when multiple balls are dropped from the same position,
        their average landing position is a better estimate of the drop position.
        More balls means more precision.
    </p>
    <p>
        In each trial, you will receive two signals about the same unknown drop position.
        Each signal is the average landing position of some number of balls, rounded to
        the nearest integer. You will <strong>not</strong> see the drop position, the path
        of the balls, or where individual balls landed.
    </p>

    <div class="highlight-box">
        <strong>Step 1:</strong> You will see the average landing position of some number
        of balls dropped from the unknown position. This is the first signal.<br><br>
        <strong>Step 2:</strong> You will see the average landing position of another set
        of balls dropped from the same position. This is the second signal.<br><br>
        Your job is to combine both signals to estimate the drop position.
    </div>

    <h2>Tasks and Trials</h2>
    <p>
        You will complete <strong>{{ NUM_ROUNDS }} trials</strong> across two tasks.
    </p>
    <p>
        <strong>Task 1 (4 trials):</strong> You will be told exactly how many balls
        were used for each signal.
    </p>
    <p>
        <strong>Task 2 (4 trials):</strong> You will be told how many balls were used
        for the first signal, but not the second.
    </p>

    <div class="key-point">
        <strong>Independence:</strong> Each trial is completely independent. The drop position
        is drawn fresh each trial. The result of one trial tells you nothing about the next.
    </div>

    <h2>Scoring</h2>
    <p>
        Each trial is worth up to <strong>{{ PER_TRIAL }} points</strong>, with a penalty
        for estimation error.
    </p>

    <div class="highlight-box">
        <strong>Points per trial</strong> = {{ PER_TRIAL }} &minus; (error)<sup>2</sup> / 4<br><br>
        Your error is the distance in columns between your estimate and the true drop position.<br><br>
        Examples:<br>
        Off by 10 columns: {{ PER_TRIAL }} &minus; 25 = <strong>75 points</strong><br>
        Off by 20 columns or more: <strong>0 points</strong>
    </div>

    <p>
        Your total score is the sum of points earned across all trials.
    </p>

    <p>When you are ready, click Next to begin Task 1.</p>

</div>

{{ next_button }}
{{ endblock }}