Title
Logo

Documentation

Learn how to use Fidenly and all its features

Price Predictions

Learn about price predictions and scoring

The Price Prediction feature allows users to make predictions about stock prices. You can predict where a stock will be at a specific target date, indicating whether you're bullish (expecting price to rise) or bearish (expecting price to fall).

Table of Contents

Understanding Price Predictions

Key Components

ComponentDescription
TickerThe stock symbol you're predicting
Target DateWhen your prediction expires
Initial PriceStock price when prediction was made
Target PriceYour predicted price
Actual/Final PriceCurrent market price (or final price if expired)
Partial/Final ResultHow accurate your prediction is

Prediction Accuracy Color Coding

The result is color-coded based on the absolute percentage difference between the actual price and your target price. This helps you quickly identify how accurate your predictions are.

Color Scale

RangeRatingDescriptionIndicator
0-5%ExcellentVery accurate prediction🟢
5-7%GoodReasonably accurate prediction🟡
7-10%FairModerate accuracy🟠
>10%PoorSignificant deviation from target🔴

Note: The color coding is based on the absolute percentage difference. This means that -5% and +5% are treated the same and both receive a green color if within the excellent range.


Examples

Excellent Prediction

  • Target: $100
  • Actual: $103
  • Result: +$3 (3.00%)

Good Prediction

  • Target: $100
  • Actual: $106
  • Result: +$6 (6.00%)

Fair Prediction

  • Target: $100
  • Actual: $109
  • Result: +$9 (9.00%)

Poor Prediction

  • Target: $100
  • Actual: $85
  • Result: -$15 (15.00%)

UserScore - How Your Prediction Score Works

Your Score

Your score measures how accurately you predict stock prices over time. It rewards both precision and consistency, with higher scores for correct predictions over longer timeframes.


How Scoring Works

Direction Matters First

  1. Your prediction must get the direction right (up/down)
  2. Wrong direction = 0 points, even if the price is close

Points Formula

Score = Days × (Precision ^ Exponent)

Where:

  • Days: How far ahead you predicted
  • Precision: How close your target was to the actual price (0-100%)
  • Exponent: Gets smaller for longer predictions (makes them easier)

Why This Matters

TimeframeRequirementPoints Potential
Short-term (1-7 days)Need high precision to score wellLower maximum points
Long-term (30+ days)More forgivingHigher maximum points

Your Accuracy Rating

Accuracy shows your average error rate across all predictions, weighted by duration:

Accuracy = 1 - (Total Weighted Error / Total Days Predicted)

Example: A 10% error on a 30-day prediction affects your accuracy more than a 10% error on a 1-day prediction.


Tips for Better Scores

  1. Get the direction right - Bullish or bearish correctness is mandatory
  2. Longer predictions = more points - But only if you're accurate
  3. Consistency pays off - Your score builds over time
  4. Close counts - Even 95% precision scores well for long-term predictions