I can't seem to get my PCF for a Pull back to work:

C > (C + C1 / 2) > (C + C1 + C2 / 3)

Is there a better way to get a short pull back of a trending stock?

You need to make each comparison individually and connect them with AND instead of putting the comparisons in a series.

C > C + C1 / 2 AND C + C1 / 2 > C + C1 + C2 / 3

I'm guessing you actually mean the following.

C > (C + C1) / 2 AND (C + C1) / 2) > (C + C1 + C2) / 3

Which could be done with moving averages.

C > AVGC2 AND AVGC2 > AVGC3

Although I'm pretty sure this would only be true if the following is also true.

C > C1 AND C1 > C2

As far as pullbacks go, most requests are usually about price pulling back to some level (such as a moving average) or within a certain range of time. I'm going to go with within a certain range of time definition. Could you describe the features of the type of pullback you are trying to identify?

I'm trying to get 3 candle pull back in price with either a 1 minute or 5 minute chart.  I was hoping an average of C + C1 / 2 AND C + C1 + C2 / 3 would give a short-term pull back. from the previous close.

I would think that in a pullback price would be going down. Your formula would seem to require price going up. If you want price going down for the most recent 2 bars you could just use the following (which is the opposite of what your formula boils down to).

C < C1 AND C1 < C2

I would suspect you want more than just that though. If it is a pullback, you probably want it to have been trending up prior to the pullback. The following would check for price to have two bars ago have the highest close over a 10-bar period with a positive linear regression slope over the same period as well.

C < C1 AND C1 < C2 AND C2 > MAXC9.3 AND 9 * (C2 - C11) + 7 * (C3 - C10) + 5 * (C4 - C9) + 3 * (C5 - C8) + C6 > C7

Using Linear Regression vs Classical Peaks/Valleys for Divergence Analysis

Yes, I think this works... Let me test today... Thank you.

You're welcome.