The best indicator is the one the is invisible. When price retraces, it retraces in many fashions/manner/forms. So use naked eyes to look out for the end of retracement and get ready to enter the trade. There is no indicator to substitute what an experienced naked eye can see. Trading involves tons of hard work. No indicator can substitute for that.
You can never truly know when a stock will pullback, or when it will pullback significantly. I use the following, because they are what most other traders are watching and theres not too much clutter on my charts. Price action, volume, s/r, MA 20/50/200 + RSI The more that line up the better. For example, a stock trending up on the weekly, pulls back hard to touch the 50 EMA, is approximately at the last pivot low making it a support area, gives you a daily green doji candle, with >1 relative volume, and RSI <80. For this example, you have many different traders all looking for different signs of a pullback, all identifying what they like to see before going long.
Pullback signals - This might be worth a minute. Pop an 8ema on your charts to see. Trading the T-line - Stephen Bigalow cued....
In trading there is nothing that can tell you truly what will happen. If you really need that, better stop searching and waste your time on other things. You can only improve the probablilitty that you are right.
You may want to look into the more complex type of averages and filters. Kaufman's Adaptive Moving Average (KAMA) uses a nice weighting scheme based on the efficiency ratio. You can use it to denoise trending time series. Linear Regression Moving Averages (LRMA) are standard for detrending, and can even be used iteratively. You can also use the differential of price and an LRMA. Summing the differential will remove trends from the time series and smooth it. Recursive Median Filters (RMF) are a nice little tool from Ehlers that can denoise a complicated time series. Signal processing is hard, but this one's so good it's worth learning. Most of the more advanced tools use recursion in the calculation of the average, meaning they take the previous periods calculation, and weigh that against new information and price to aid in filtering out volatility or noise from the data.
I've been toying with these 3 in different ways to try and figure out how to utilize them during ranging to assist staying out of bad trades as well as during trending (since it'll be on all the time). Are there any tips/hints/guides that describe how to utilize them to minimize entering trades during the chop? That's probably a loaded question, but I was looking at their relationship to EMA, RSI, etc... ie, when the LRMA 8 passes below the LRMA 29 and is above the EMA 20 then go short, etc... I was trying out tons of combinations and trying to evaluate it, but any cheat sheets to read would be great.
Not sure what software you're using. I do all that stuff with thinkScript. I like to use a 150min LRMA to give me a reliable trend estimate, then do a bunch of smoothing/filtering/differencing or momentum stuff. I don't have a simple algo using these, but I do something like that. It's more like calculus, though. If you sum the differential of price against an LRMA, you remove trend from a data series, but preserve the changes in momentum. It's kind of like a filter. The RMF is awesome because it's a non-linear filter that ignores outlier price spikes, and yet has no lag on trend estimation (very similar to KAMA). While the others, like exponential, or wilders smoothed exponential, will follow spikes. If you chart the divergence, you can isolate extremal events in a data series. I use them to isolate extremal momentum events by applying them to a spread of the summed differential and a momentum differential. It's complicated, but I would suggest you use them for their intended purpose, whether its noise filtering, rapid tracking due to aggressive front weighting, or measuring trend strength with varying sensitivity to certain behavior. Not sure how to make this simple... This is a chart that uses an LRMA on the upper, then below is price vs the LRMA with two more LRMA on that. It's detrended and the LRMA pair is used to track momentum within the trend. (just a way to use some of them) (this is a treasury fly with an ultra bond overlay) I almost always use VWAP for any kind of trend analysis, though.
"LRMA on the upper, then below is price vs the LRMA with two more LRMA" Measuring trend strength would be a necessary 1st step in producing what OP is looking for. For the 2nd step "predict what a reasonable pullback zone can be based on the strength", is your LRMA study designed to predict the depth or duration of a Reasonable Pullback Zone Valley for the Correction Wave based on the Trend Strength?