Overview:
We hear a lot about Gartleys, bats, crabs and the rest of the barnyard crew, but have you ever wondered what other creatures might be lurking out there yet to be discovered? Well wonder no longer, it's time to find out for yourself! The Custom XABCD Validation and Backtesting Tool allows you to define retracement ratios and targets for your very own patterns.
Tips:
(1) Adjust the patterns entry/stop/target configuration and see how it affects the pattern's backtesting results.
(2) Adjust the weights of pattern score components (% error, PRZ confluence, Point D/PRZ confluence), along with the entry minimum score requirements ('If score is above'), and see how it affects the patterns' results.
Pattern Scoring:
The pattern's score is an attempt to represent the quality of a pattern with a single metric. This is one of the most powerful aspects of the tool because it can quickly tell you whether a trade is worth entering. The score is based on 3 components:
(1) Retracement % Accuracy - this measures how closely a pattern's retracement ratios match your defined theoretical values. You can change the "Allowed ratio error %" in Settings to be more or less inclusive.
(2) PRZ Level Confluence - Potential Reversal Zone levels are retracements of the XA, BC, and/or XC legs. These levels indicate where a potential reversal might occur (i.e. pivot point D). The PRZ Level Confluence component measures the closeness of the two closest PRZ levels, relative to the height of the of the XA leg.
(3) Point D / PRZ Confluence - this measures the closeness of point D to either of the two closest PRZ levels (identified in the PRZ Level Confluence component above), relative to the height of the XA leg. In theory, the closer together these levels are, the higher the probability of a reversal.
While the score is percentage-based, it should not be confused with a probability. A score of 96% does not imply a 96% chance of success. It simply represents the average of the three components mentioned above, weighted according to the defined weight parameters. A score of 100% would mean that (1) all leg retracements match the defined theoretical retracement ratios exactly, (2) all PRZ retracement levels are exactly the same value, and (3) pivot point D occurred exactly at the confluent PRZ level.
Pattern scoring research has been ongoing since I introduced the concept with my Harmonic Pattern Detection, Prediction and Backtesting Tool (see below). So the way that the score is calculated is subject to change based on the results of that research.
We hear a lot about Gartleys, bats, crabs and the rest of the barnyard crew, but have you ever wondered what other creatures might be lurking out there yet to be discovered? Well wonder no longer, it's time to find out for yourself! The Custom XABCD Validation and Backtesting Tool allows you to define retracement ratios and targets for your very own patterns.
Tips:
(1) Adjust the patterns entry/stop/target configuration and see how it affects the pattern's backtesting results.
(2) Adjust the weights of pattern score components (% error, PRZ confluence, Point D/PRZ confluence), along with the entry minimum score requirements ('If score is above'), and see how it affects the patterns' results.
Pattern Scoring:
The pattern's score is an attempt to represent the quality of a pattern with a single metric. This is one of the most powerful aspects of the tool because it can quickly tell you whether a trade is worth entering. The score is based on 3 components:
(1) Retracement % Accuracy - this measures how closely a pattern's retracement ratios match your defined theoretical values. You can change the "Allowed ratio error %" in Settings to be more or less inclusive.
(2) PRZ Level Confluence - Potential Reversal Zone levels are retracements of the XA, BC, and/or XC legs. These levels indicate where a potential reversal might occur (i.e. pivot point D). The PRZ Level Confluence component measures the closeness of the two closest PRZ levels, relative to the height of the of the XA leg.
(3) Point D / PRZ Confluence - this measures the closeness of point D to either of the two closest PRZ levels (identified in the PRZ Level Confluence component above), relative to the height of the XA leg. In theory, the closer together these levels are, the higher the probability of a reversal.
While the score is percentage-based, it should not be confused with a probability. A score of 96% does not imply a 96% chance of success. It simply represents the average of the three components mentioned above, weighted according to the defined weight parameters. A score of 100% would mean that (1) all leg retracements match the defined theoretical retracement ratios exactly, (2) all PRZ retracement levels are exactly the same value, and (3) pivot point D occurred exactly at the confluent PRZ level.
Pattern scoring research has been ongoing since I introduced the concept with my Harmonic Pattern Detection, Prediction and Backtesting Tool (see below). So the way that the score is calculated is subject to change based on the results of that research.
Notes de version:
Added parameters for removing or moving location of the results table
Notes de version:
Fixed bug with incomplete patterns getting skipped
Notes de version:
Lots of bug fixes
Notes de version:
(1) Fixed matrix.rows runtime error
(2) Made entry checking more efficient
(2) Made entry checking more efficient
Notes de version:
Fixed "objects cannot be positions more than 500 bars in the future" runtime error
Notes de version:
Fixed bug with invalid incomplete patterns being added
Notes de version:
Bug fixes for drawing/updating potential patterns
Notes de version:
Better handling PRZ levels < 0
Notes de version:
Now drawing potential/incomplete pattern labels on projected point D/entry/PRZ level to reduce label overlapping
Notes de version:
Updated to use latest version of Fib library (improved efficiency for precise fibs)
Notes de version:
Updated to use latest version of Fib library (improved efficiency for precise fibs)
Notes de version:
Firing alert only on bar close until issue with real-time bar alert spamming is resolved