Different Results on different machines

[daven]

I recently started testing my strategies on a second machine running a demo-only version of NT. I'm doing this because I have found that backtesting strategies and collecting live market data or running strategies at the same time can cause lockup problems. So I develop and test on one machine and run live on the other. Here's my question. I back tested the same exact strategy with the same exact parameters and got wildly different results from the two different machines. How do I know which one is the correct answer, and why is there a difference. The only significant difference is one machine is downloading live data from my live account for the test, while the other is downloading data from a sim account. Both are going through the same broker. Is live data different going into a sim account than it would be going into a live account. For reference here are snapshots of the two backtests, as well as the data that was used to run them:
I just learned that I can only attach one file because of size restrictions. I will add the second image to a reply to this message. To clarify the issue though on one machine I get a net profit of 24,101.96 at a profit factor of 1.79, while on the other machine, same settings I get a net profit of 13,765.50 at a profit factor of 1.45. Everything else is the same as you will see in the two images.

[daven]

Here's the second image of the results from the second machine.

Image 2 from the previous message.

[daven]

Serial Tries, different results

I ran the strategy a second time on both of the machines. On my test machine I got exactly the same result as the first run, while on my live trading machine I got a very significant difference. This time when I ran the strategy I got a net profit of 3,882.66 versus a value of 24,101.96 the last time I ran it. I also got a different profit factor, 1.12 for the second run versus a profit factor of 1.79 for the first run. It appears I cannot trust the results on this machine and I'm wondering what I should be doing. Do I need to completely delete and reinstall NT on this machine?
DaveN

[NinjaTrader_RJ]

Hello,

Thank you for your post.

I'm not sure why those would be different. The data should all be the same since it is from the same data feed.

Have you checked the code to both strategies and verified they are the same?

Also try repairing both databases and attempt the backtest again: Tools --> Options --> Misc --> Repair DB.

As a further test, can you try testing one of the simple default strategies such as the MA crossover and see if the results differ?

Thank you!

[daven]

Different Results

I already did what you have asked. I repaired both data bases. Took the strategy that I used for the tests, saved it with a new stategy name, exported it over to the other machine, imported it to that machine and reran the tests. Now, I am getting negative results on the live machine with both the new renamed strategy and the old strategy. I opened them both up and verified they were the same line-by-line. Is there a way I can look at the data used by the strategy analyzer to ensure it is the same on both machines?

[NinjaTrader_RJ]

Thank you for your reply.

There can be small differences in recorded data in each computer that can change results.

Try exporting the database from one computer and restoring it to the other computer: http://www.ninjatrader-support.com/H...verview50.html