Sample sizes

Hi guys,

I’m trying to figure out what would be good sample sizes (SS) to calculated the average ranges for all the time frames.

m5 SS=288 bars (in a day)

m15 SS=96 bars (in a day)

m60 SS=24 bars (in a day but I think maybe too small)

4hr SS=30 bars (for a week, same too small?)

daily SS=251? trading days in a year??

weekly SS=50? (say 250 trading days, 5 days a week)

Monthly SS=12?

More samples better or worse? Is there a logical way to determine or is it a traders choice?

I just want to know what is optimal for most traders so that I can make more informed choices, reasons would be very cool too :slight_smile:

Thanks!

the “norm” for calculating average daily range is 14 periods.

since ATR measures volatility, you want to have the best recent reading, so my advice is that you shouldn’t go beyond 28 periods.

Hello!

Thanks for your reply.

It makes sense to measure recent volatility.

Is there a personal reason for 28?

Not really… I just do not like to have data more than a month old for ADR… I use 14 periods… :slight_smile:

Thanks yunny! :slight_smile:

Mmm what is called recent volatility? Can I see this on the charts or elsewhere to find a good sample size? Anyone?

I just use sixths of the median range of the last three weekly candles projected from the low and high extreme of the current week.

Right smack dab in the middle of the two zones is the median range.

CrashZone.zip (2.07 KB)

Hello SOM,

Thanks for this. Hmm thinking about it more I think you are grouping weekly bars together to get a target??

Not sure how you are calc the targets from the last three candles, but from what I can gather so far, the targets are based off fib extensions from this average range, is that correct?

Currently in an AU short average 1.0560 and using the crashzone indi I see the first target 1.430 is coming up, can you say that I see something?

EDIT: Nvm, its heading for the green target 1

He said how he calculated it in his post - he takes the median range of the last 3 weekly bars - i.e. out of the last 3 weekly bars he takes the middle one w.r.t how big it is. Then he calculates 6ths of that bar and projects those 6ths from this?? weeks high and low (surely last weks high and low otherwise the zones are going to move as this week’s range expands???), if you want to correlate 6ths with fibs, that’s up to you.