Strange and disturbing

[mainstream]

Any thoughts on why testing the same strategy on two different machines would produce drastically different results?

Same contract
Same date range
Same time rang
Same tick bar size
Same data provider

No optimization

Look forward to any feedback or troubleshooting techniques. Thanks.

[NinjaTrader_Jason]

I suggest to delete all historical data you have in the database for the instrument in question via Tools-->Historical Data-->Delete for both machines.

Backtest the strategy once more on both machines, while connected to the same data feed and using the same parameters as you did before.

Check if the results resemble each other.

[mrlogik]

mainstream,

What broker are you using?
If NT is doing a time stamp based on your system clock, and the two computers are off a few seconds from each other, you can have different results.

[mainstream]

Thanks for the replys. That's is a great point about the system clocks. Unfortunately it's a little more mundane an error.

The tick data was accumulated by the NINJA, not imported from IB... So one machine had different data than the other.

Follow up question who has cheap historical tick data? Minute models are great, however, I'm interested in tick/scalp models right now.

Thanks again!

[NinjaTrader_Jason]

You can find all supported connectivity providers at the link below.

http://www.ninjatrader-support.com/HelpGuideV6/HistoricalData.html

Unfortunately, we do not have information regarding the pricing models of the supported connectivity providers. You will need to contact the providers themselves regarding this inquiry.