[realtraders] CL_Evaluating & Optimizing systems {02}

["Gary Fritz" <fritz@xxxxxxxx>]

Herr Zebra wrote:
> > % wins:  Obviously, the higher the better.  
> I think thats a very difficult and dangerous number. While you are
> right about  the PSYCHOLOGICAL effects of "having a winner or a
> looser", the bottom line is the $ results over all. 

I absolutely agree that the bottom line is what counts, **IF** it's a 
realistic and really-tradeable result.  But system trading is a 
psychological challenge for most people, and a higher-win system is 
going to be easier to follow than a low-win system.

Furthermore, I don't agree that bottom-line profits are the MOST 
important number.  As I said in my post, high profits from a system 
can be very misleading.  I've seen too many newbies get excited about 
an S&P system that shows huge profits but averages only $50 a trade, 
or other craziness.  You've got to have a system that you can trade 
successfully in REAL LIFE.

Let's say I had to choose between two systems that reported $100k 
profit over some period.  If one was a 20% winner that averaged $1000 
a trade and had a reasonable drawdown, and the other was a 90% winner 
that averaged $50 a trade or had a $100k drawdown, I'd definitely go 
for the safer low-% system.  High win % alone is NOT a good way to 
choose a system, any more than high net profits is.  But if I had to 
choose between two systems that were fairly similar except one had a 
10% higher win percentage, I'd definitely go for the higher-% system. 
 It's easier to trade and it reduces your risk.

Gary