Explanations

Strike Rate

• Your strike rate is determined by how many wins vs losses you have.
• Open predictions are not counted towards this number.
• Your strike rate does not get affected by your swan score.

Swan Score

• Your swan score is used to determine your performance and rank among your closed predictions.
• It is calculated using a combination of price delta, elapsed time, and asset volatility.

Viper Charts

• Viper Charts show the price data from the date a prediction has been made until the specified expiration date of the asset of interest.
• Please see Viper Charts Twitter for fully open source charting.

Confidence

• Confidence is used as a way to mitigate the risk of your prediction's effect on your swan score.
• By adding a level of confidence in your prediction, you are reducing the amount of swan score you are allowing yourself to risk on the prediction.
• Confidence level effects your swan score by multiplying it by the fraction of your confidence
  • 10% confidence = 1/10 Swan Score prediction total
  • 70% confidence = 7/10 Swan Score prediction total

Prediction Feed

• The prediction feed is the right hand side of the homepage.
• This gives a summary of all predictions made within the Capswan app, both publicly and privately.
• Private predictions will show who made the prediction and what asset the prediction was made on.
• Public predictions are fully visible to all Capswan users.

Progress Bar

• The progress bar is visible on all open predictions. It is used to track how far the prediction is from completion.

Prediction Metadata

• To expand the full metadata available on a prediction see Metadata Instructions.
• From the metadata, we can see the full range on information gathered when a prediction is made. This information is visible below.

Natural Language Processing (NLP)

• Capswan uses NLP in order to extract metadata from a prediction.
• The NLP is continuously being updated to keep up to date with current prediction styles.
• To see the currently supported NLP, please see Prediction Text Styles.