Acceptable Variance?

[allphar]

Hi guys,

When applying/testing a particular bet selection approach (not mechanical system), what do you consider to be an acceptable variance/drawdown over a range of shoes.

For example say an approach hit minus 40 units over 100 shoes. Is that too much?

What should the max be? I know the idea is to win more bets than lose but not sure of where I should set the drawdown line.

Thanks.

 

 

[kachatz1]

Not sure there is any magical formula, but over 100 shoes?

     -you should be AH.E.A.D. with an " accurate bet selection approach" with almost  7500 chances/100 shoes

Since we don't espouse betting every single hand, even you bet 1x/shoe, you should be ahead of the game...

( if not, better get another methodology to follow...)

or, in the alternative, maybe even better, is to use your experience as you go along to identify what it is about that " system" that a " workaround/fix it" could turn it into a winner...

this is why this forum so valuable...plenty of tools in the toolbox to keep you far away from the woodshed...

[wolfat]

allphar,

at first we have to define the word "variance".

For me that's the spread between losing and winning decisions, disregarding the amount bet.

The avg for a statistical spread like this is the square root of the total decisions played.

So, on avg. in 100 decisions played you should be +/- 10, to be on avg

IMPORTANT: I am speaking about decisions PLAYED, not only recorded.

If you are at +20, you are performing better than avg spread.

In 100 shoes, that are ar. 7000 decisions, if you bet ar. 50% of them (so 3500 decisions) your avg should be +/-59 

...and so on.

I always keep record of my spread for many reasons but the most two important are:

a) I track how I am performing compared to the past

b ) I know, as I track my highs and lows, "more or less" if some winnings or losses are expected

A

[allphar]

Thanks Kevin and wolfat. Great information.

Learning, learning, learning.