What stats are important? What's good enough to trade?

So I finally got around to downloading some good data and trying some things out.

The curves are nice, drawdown to rate of return is good. Profit factor is low though. It's easy to say all we need is a 51% win rate consistently, it's another to put money behind it.

So my question is, what stats do you think are important and what are the numbers they should be?

Many thanks for any contributions.

Hotch

Who told you that you need a 51% win rate?

100 trades:

49 trades lose for a loss of $4,900.00
51 trades win for a gain of $4,000.00

or 100 trades:

90 lose for a loss of $4,900.00
10 win for a gain of $5,000.00

which is better?? Yeah I know it is an extreme example, but it is just to illustrate a point. My current EA has had days where the accuracy was only 25% and yet the account went up.

For me the only statistic that is important is what direction the account balance is going. I don't consider a method valid until after it has completed 250 live forward trades. Back testing is just to give an initial view of the potential. Trading live is the only way to determine if a system can be traded. Why 250 trades? I dunno... I just picked that number you may like 50 or 5000, whatever it takes to get you to the point of belief in the method.

You can crunch all the numbers you want but in reality the account balance is the only really useful number.

Actually the guy who told me was head trader at a large bank

I guess what we both mean was profit factor just needs to be anywhere > 1, but I thought the meaning was obvious (was late).

Pretty sure I must disagree with balance just going up. If you risk 100% of your account every day just to make 0.001%, even if it works for your 250/5000 trades whatever, it's not a good RoI, too much risk etc, I'd not be happy trading it (again extreme example), but you get my meaning.