I see 2 sorts of opportunities for "ideal" pullbacks:
1) "strong" trend - how one defines a "strong" trend isn't the point ... the point is, an "ideal" pullback is one which leads to a new leg for that trend with a very high probability. As already stated, in such conditions pullbacks are usually shallow, just enough to get counter-trend traders in - and stopped - again & again.
2) initial breakout out of a consolidation range ... ideally, that range has developed over a few days (for intraday trading) or a few weeks (for swing trading). There is a "classic" pullback after the break-out, which usually only retests the breakout level, the longer-term players get on board & price never looks back.
Outside of these ideal conditions, my favorite pullbacks are ABC-type pullbacks, where C comes in confluence with some other level (a prior S/R level, a retracement level, anything will do, really). I use 2 projection methods from AB to defined a maximum excursion for C, this gives fairly low $$$ exposure entries - but not too frequent either, because often-times C actually makes a DB(/DT) at B.
To the OP, even though retracement levels are sometimes useful, there are so many possible in a trend that of course 1 gets hit & is the pullback extreme, but which one is the million dollar question.
BUT, there is another tool that works pretty well in established trends, the popular & better known version of it is the channel.
