How scoring works

What gets recorded

When you submit a price target, we store:

• Ticker — the stock symbol
• Price target — where you think the price is headed
• Horizon — 1 day, 1 week, 1 month, 3 months, or 12 months
• Entry price — the market price at the moment you submit (extended-hours price for 1-day calls when applicable)
• Timestamp — exactly when the call was made (entered_at)
• Evaluate date — when the prediction matures and gets scored

This record is permanent and public on your predictor profile. You cannot edit a prediction after submission.

The five horizons

Each horizon is scored independently. Short-term skill and long-term skill are tracked separately.

When predictions are scored

After the evaluate date passes, we fetch the closing price for that trading day and compare it to your target. The result — actual price, accuracy percentage, and evaluation timestamp — is added to your public record.

Predictions still waiting to mature appear as pending on your profile. Only evaluated predictions count toward leaderboard rankings.

Accuracy score

Accuracy measures how close your predicted move was to the actual move, as a percentage from 0 to 100.

We compare the predicted percentage change (target minus entry, divided by entry) to the actual percentage change (actual minus entry, divided by entry). The closer the two moves align, the higher the score.

A perfect forecast — same direction and magnitude — scores 100%. A call in the wrong direction scores 0%. Partial credit is given when direction is correct but magnitude differs.

Direction score

Direction is simpler: did the stock move the way you predicted?

• Correct — you predicted up and it went up, or you predicted down and it went down
• Incorrect — you predicted up and it went down, or vice versa

Your direction score is the percentage of evaluated predictions where direction was correct. On the leaderboard, horizon columns show direction as correct / total evaluated (e.g. 8/10).

What wrong calls cost

Every evaluated prediction becomes part of your permanent public record. There is no way to delete a scored call — accountability is the point.

• Wrong direction — scores 0% accuracy on that call and lowers your average
• Right direction, wrong magnitude — earns partial credit (above 0%, below 100%)
• Right direction and magnitude — scores up to 100%
• Leaderboard impact — rankings use your average accuracy across all evaluated calls; bad guesses drag you down over time

Random guessing is cheap before evaluation, but wrong calls leave a visible trail. The leaderboard requires at least 10 evaluated predictions, so consistency matters more than one lucky shot.

Leaderboard ranking

To appear on the leaderboard, a predictor needs at least 10 evaluated predictions.

Default ranking order:

1. Average accuracy — mean accuracy across all evaluated predictions (higher is better)
2. Direction accuracy — percentage of calls with correct direction (tiebreaker)
3. Total predictions — more evaluated calls break remaining ties

Leaderboard modes

• All — ranks across every horizon combined
• Active thesis — ranks on 1 month, 3 month, and 12 month calls only
• Pulse — ranks on 1 day and 1 week quick calls only

Sorting by a specific horizon column requires at least 10 evaluated predictions in that horizon. Horizon sorts rank by direction first, with accuracy as tiebreaker.

View the leaderboard →

Worked example

Every evaluated call on GetTheTape follows this same process. Your full history — wins and misses — is visible on your public profile.