M Alpha Core logoM·ALPHA·COREBuy a license

Home / Videos / A real SUI scan tested on data the search never saw

A real SUI scan tested on data the search never saw

Published on November 3, 2026 · Length 1:46 · Last updated: November 13, 2026

Download M Alpha Core Watch on YouTube TikTok Instagram X

About this video

Everyone shows you the backtest. Almost nobody shows you what happens next. A setting can look perfect on old charts and still fall apart in the first month of real prices, because it never found an edge. It only memorised the past. This problem is called overfitting, and this video shows how M Alpha Core deals with it on a real market.

We type in one coin, SUI, and load a full year of real candles: about nineteen thousand half hour candles plus the hourly ones. Then the scan starts. It tries twenty thousand different settings of the built-in strategy. The important part: 35 % of the data stays locked away. While the scan searches, it never sees those months. Only at the end is the best setting run on them.

The result for SUI on the one hour chart: 97 trades. 66 of them fell in the data the scan searched, 31 in the months it had never seen. The profit factor was 4.13 on the searched data and 5.47 on the locked months. In plain words, this setting did not just fit the past. It also held up on data it had never seen.

Is that a guarantee? No, and the video says so clearly. 31 trades are a start, not proof. That is why the app offers more checks before real money goes in. At the time of the video, the next step was a second check of the same setting in TradingView. Today the scans run on our own scan servers and no TradingView subscription is needed. You can also let a bot trade with play money first, see our page about paper trading.

The same method stands behind every result in M Alpha Core: search on the first 65 % of a year, test on the last 35 %. Our page about out of sample testing explains why this split matters and how to read the two numbers side by side.

In the next video of this series, a setting like this one is turned into a bot that trades long and short on your own account.

The numbers in this video

More on this topic: Out of sample testing explained.

Related videos

Learn more

Download for Windows, Mac and AndroidHome

All figures come from backtests on past market data. No result is a promise of future profit. Trading, especially with leverage, involves high risk. M Alpha Core is a software tool, not investment advice.