What is the Pair Prediction Function and how does it work?

Most of us have heard of indicators like RSI and Bollinger Bands and have likely used them. These are among the most well-known tools in technical analysis. A significant use of RSI is in identifying “oversold” and “overbought” conditions. Oversold is generally seen as a potential buying point, while overbought is considered a potential selling point. However, this does not mean that an asset bought at oversold must be sold at overbought; rather, overbought is used independently as an exit signal.

When we look at RSI on a chart, we often observe that when RSI reaches the oversold zone—typically below levels like 30 or 35 in traditional markets—the price tends to reverse and return to the same oversold level. While this concept is simple to state, estimating the exact oversold price in real-time is extremely difficult. It is challenging to determine how far below the oversold threshold the price may go and at what point one should enter a trade. Similarly, identifying where the price will rebound to, allowing for a profitable exit, is also complex. The same challenge exists for overbought levels—knowing where to sell and when to re-enter.

Bollinger Bands operate on a similar principle. The bands consist of three lines, and when the price falls below the lower band, it often returns to that band, just as it tends to return from above the upper band. While this doesn’t occur every time, it happens frequently enough to be relevant. The question remains: how far below the lower band should one consider buying, and how can this point be identified? Similarly, how far above the upper band should one sell, and how can that level be determined?

The purpose of the current function is to address this exact complexity. Given the volatility of the crypto market, this function sets RSI thresholds at 20 for oversold and 80 for overbought. It first estimates the price levels at which these RSI values will occur. Then, using historical data, it analyzes how much the price has historically dropped below (in the case of oversold) or risen above (in the case of overbought) those thresholds. Based on this data, it further evaluates where the price typically rebounds to after such movements. The function then provides predicted levels based on current and projected prices.

RSI is commonly calculated using either 7 or 14 candles; this function uses 7 candles. Another important aspect of the function is that it identifies strong support or resistance zones around each expected level. The strength of these zones is determined by prioritizing higher timeframes first, followed by medium and then lower timeframes. For example, if a daily support level appears within an hourly chart, it is given preference. Furthermore, each support or resistance level is only considered valid if the price has touched that zone at least three times historically.

The function applies the same process to Bollinger Bands, identifying levels, support, and resistance zones accordingly. However, its functionality does not end there. As mentioned earlier, both indicators reflect market conditions, but prices do not always return to their levels—especially in lower timeframes where accuracy drops. Therefore, the function combines both indicators to generate consolidated levels, supports, and resistances. When both indicators are combined, accuracy improves significantly, often reaching 95% to 98%. Each coin includes a success report where one can view historical performance of the function. For most strong coins, the historical performance has been excellent, especially in predicting price rebounds from oversold conditions. The function performs particularly well on four-hour charts.

The function’s process can be summarized as follows:

For oversold conditions, it first identifies the price at which a coin is expected to become oversold. Then it calculates three further price levels below that point:

1. Minimum Level: The function examines historical data to determine the lowest point price has dropped below oversold. It then identifies this level and calculates the minimum, average, and maximum rebound levels observed from that point.
2. Average Level: Using a weighted calculation method, it determines the average drop below the oversold level based on historical patterns. From this level, it also calculates three rebound levels.
3. Maximum Level: It identifies the maximum historical drop from oversold and calculates corresponding rebound levels.

For overbought conditions, it first finds the price at which the coin is expected to be overbought, and then calculates three levels above it:

1. Minimum Level: It determines the smallest historical increase after reaching overbought and calculates three potential pullback levels.
2. Average Level: It finds the average historical rise above overbought and derives the corresponding three pullback levels.
3. Maximum Level: It identifies the highest historical rise beyond overbought and calculates three pullback levels accordingly.

To enhance usability, the function includes several visual aids:

1. Table rows and live prices are color-coded. The closer a coin is to oversold, the lighter the green shade; the closer to overbought, the lighter the red. This makes it easier to interpret market conditions at a glance.
2. When a coin reaches oversold or overbought, a blinking dot appears. A blue dot indicates an RSI-based signal, a red dot for Bollinger Band signals, and a green dot if both conditions are met. Combining row color and dot indicator gives an immediate sense of a coin’s condition. For example, a green row with a green dot means both indicators signal oversold, while a red row with a green dot indicates both are overbought.
3. Live prices update every minute. Below each coin, a historical performance report is available to assess how well this strategy has worked in the past. Coins are gradually being added to ensure stability of the function, and currently, only Binance-listed coins are being considered.
4. For Pro users, alerts are available when the price crosses the average of both oversold or overbought levels. For example, if a price drops below the average oversold level, an alert is triggered. A user may then consider entering at that point and exiting near the original oversold level.
5. An additional option for high-risk alerts is also provided. Instead of triggering strictly at the average level, these alerts are generated slightly above and below the average. This allows users to split their entry into two parts, resulting in an average entry near the lower boundary. This is particularly useful for coins that may not always reach the average level.
6. The webhook management feature is designed to integrate Botslash with 3Commas, an automated trading platform. Once connected, alerts generated from Botslash can send buy or sell signals to 3Commas, allowing for fully automated trading from entry to profit-taking. By joining 3Commas using our referral link, users can get a one-time discount—40% on annual membership and 20% on monthly membership:
   https://app.3commas.io/auth/registration?utm_source=referral&utm_medium=cabinet&c=tc858563

According to our experience, this strategy is nearly fail-proof for good coins on the daily and four-hour timeframes. However, since all data is collected and analyzed automatically, errors are still possible. Therefore, users are advised to make trading decisions based on their own research. Botslash does not provide financial advice nor accepts responsibility for any trading risks.

Owais Paracha
Botslash Team