News Focus
News Focus
icon url

FreshlyBlended

06/14/09 6:33 PM

#664 RE: QUIKTDR #663

Sounds very weird. Ideally the back testing engine should be the exact same as the live engine. The only different would be the data passed into the engine. If this is not the case then back testing is completely useless.
icon url

mattie633

06/14/09 8:27 PM

#666 RE: QUIKTDR #663

The issue with all this is that what we see on the charts and what the math actually picks up are two different things when viewed tick by tick. So, for example when I look at a day chart, I might see that when the sto crosses 20 from below and the MACDh crosses 0 from below it can be a good entry point. But what that candle represents is that they crossed sometime that day... most doubtful that both of those conditions happened on the same tick.

Until SD allows us to run our live strategies on stale data, the challenge remains how to state criteria in a more general way so that all conditions are picked up at once... like maybe sto going up and below 50 and macdh going up. It seems to me that more conditions of a general nature is the better combo than just one or two very specific conditions. Does that make sense?