Well that was pretty shallow. Only a single pattern was included and even that performed worse than the index. If you were to include the transaction costs and

Well that was pretty shallow. Only a single pattern was included and even that performed worse than the index. If you were to include the transaction costs and the cost of time spent analysing that, the perf would be even worse. So a shallow article with a wrong conclusion. Since past returns are not a reliable indicator for future returns any back testing will only show if a thing would have worked in the past not in the future.