2 SMAs (-1SD and +1SD) around the smoothed mean (0SD; ie. in total 3 SMAs) forming a "channel" is good, IMO. The height can be made a range going from 100 (top) to 0 (bottom), and the middle band being 50, so then one immediately can read where the current stock price lies in this range between 0 to 100. Above 70 or so would mean overbought, and below 30 oversold, similar to the standard interpretation of the RSI indicator. See also Bollinger Bands etc.
For an indicator to rightfully be considered an overbought/oversold variety shouldn't it at least factor in ummm volume? TD Pressure (below) which uses ranges and volume. But nothing is perfect and judgement has to be applied. Confirmations and divergences, reverse, hidden, etc.
Couldn’t you look at Bolinger bands with standard deviation for RSI or all of them and then look at volume separately for confirmation?
I don't think there is one indicator that is good by itself. To MacBookPro's point, there is an art / craft to this game.
I don't really care for RSI...sure it looks great in hind sight but on the hard right edge it just didn't do much for me. Admittedly I could have put more effort into it though to develop probabilities...ex: RSI>80 for "n" units of time are this likely to revert this much...experimenting with different time frames until you have a set of probabilities that merit some level of confidence...then have a plan b of what likely happens if you are wrong and the appropriate reaction etc... I do like Keltner channels...more in the realm of volatility but mean reversion after momentum slows sounds pretty similar to overbought/oversold. I have also come to accept that Price action is a legit skill that is really worth developing. Accumulation/distribution is worth looking into as well IMO...Is the price trending up or down? accumulating or distributing? Pretty fundamental stuff...meaning it is extremely important.