Claims he worked for Ned Davis Research. A firm that uses back and forward testing. Or at least did so back in the day for their Four Percent Swing System. Prolly many others as well since.
With time-based charts, market sampling follows a traditional approach. However, with range-based charts, historical data is identical to future data—there is no variance in range, only in bar duration (the time it takes for a bar to complete). Every bar represents the same range, ensuring consistency in analysis. Once you catalog patterns, you'll notice a limited number of possible positions for the open and close of each bar. The high and low are always fixed at the extremes of the range. This structure creates a highly predictable environment where pattern analysis remains consistent across historical and future data. Unlike time-based charts, forward-walking the system is unnecessary—it either works or it doesn’t. The historical patterns you test will be the same patterns that appear in the future. Because there is a finite number of possible combinations and sequences, if you've mapped out how to handle them, you already have a plan for how the market will behave when specific patterns emerge. Out-of-sample testing is unnecessary with range bars because they always behave consistently. Unlike time-based charts, where future data can differ significantly from historical trends, range bars maintain a fixed structure. This means that once you've cataloged the possible patterns and combinations, the same sequences will occur again, with no unexpected variations. Since every bar is formed based on a predefined price range rather than time intervals, the market orders naturally align ahead of the patterns rather than reacting unpredictably. This structured predictability allows traders to preemptively plan their strategies with high confidence, knowing that the same setups will repeat under similar conditions. This makes range bar trading an entirely different world from time-based chart analysis—where constant adaptation and forward-walking tests are required. In contrast, range bars provide a finite, systematic approach where successful pattern recognition directly translates into repeatable trading performance.