Restating relative-strength-index parameters after a smoothing change

What the construction states

The assigned relative-strength-index here is a forecast from ordered price observations over a stated sampling interval and lookback. The historical workflow begins with a rate-of-change-sum: a 14-period price rate of change is added to an 11-period price rate of change before any smoothing is applied.

The same 10-period window can then be applied either as a front-weighted moving average or as a weighted-moving-total. Those two choices produce different numerical scales. In the weighted-moving-total version, the latest combined rate-of-change value is multiplied by 10, the value one period earlier by 9, and so on until the value nine periods earlier is multiplied by 1.

Restated cutoffs and condition rules

After the construction was aligned with the original arithmetic, a parameter of 1.3 on the averaged scale was treated as parameter-equivalence for 7.15 on the totaled scale, whereas the original published cutoff was zero.

On the totaled scale, an upward turn while the index was below 7.15 was used as the long condition, and a decline of 60.5 points while the index was above 7.15 was used as the opposite condition, compared with an 11-point decline on the 1.3 scale. Each figure is a signal-threshold that is valid only on the scale produced by the chosen lookbacks and weighting.

Switching from the averaged form to the totaled form left the shape of the series and the calendar dates of those conditions unchanged, because the difference was a rescaling rather than a new timing rule.

How the lookbacks were chosen

The 11-period and 14-period lookbacks were motivated by an informal one-year recovery analogy, not by a reported statistical search over alternative lookbacks.