A Testing environment that closely mimics Real Time results, for strategies.

[NinjaTrader_ChrisL]

Hi Isaac, thanks for your reply. I'm suggesting debugging techniques you can use for your strategy. Checking the BarsRequiredToTrade property is one suggestion to look for in explaining why you are getting different results between a backtest and real time strategy runs. Tick Replay will not have an effect on strategies that run OnBarClose. I have already given you all of the information we have on correcting the differences between real time and historical backtest. Did you try using Print and comparing the trades like I have already suggested? Please also test the SampleMACrossover that comes with the platform to see if you can reproduce on that strategy. If you can, I will be happy to take a look given the steps to reproduce are provided.

Kind regards,
-ChrisL

[IsaacBillion]

Thank Chris.
I'm working a long time with Print() and even found a bug in the Mechanism of ClearOutpuWindow().. ask NinjaTrader_ChelseaB .
I'm talking always on the SampleMACrossover as i wrote in the previous messages.
I'll post here the data if I won't succeed to get same results between backtest and PlayBack (again: not Realtime , but Playback with same historical data of backtest Played back)
Thank you very much.
Isaac.

[IsaacBillion]

Hi Isaac, thanks for your reply. I'm suggesting debugging techniques you can use for your strategy. Checking the BarsRequiredToTrade property is one suggestion to look for in explaining why you are getting different results between a backtest and real time strategy runs. Tick Replay will not have an effect on strategies that run OnBarClose. I have already given you all of the information we have on correcting the differences between real time and historical backtest. Did you try using Print and comparing the trades like I have already suggested? Please also test the SampleMACrossover that comes with the platform to see if you can reproduce on that strategy. If you can, I will be happy to take a look given the steps to reproduce are provided.

Kind regards,
-ChrisL

Hi Chris.
As you requested, I'm attaching My Sample MA crossover like in the NT8 samples (with little tweaks inside).
As you can see for the backtest this is a Rocket Profit Strategy. there is a day with %50 profit for only one day (6.5 hours trading)
For 3 month there is insane %420,000 (420 thousands percent gain). I wish it will be reality ...
But in PlayBack as you can see from the xlsx and from the pics, it's horrible!!
The Symbol is TQQQ for the day 28.06.2022 ( 28-29.06)
I wanted to attach here but it won't let me because it's a Tick Data over 2.5MB zip file (original text is 32M),
and there is a problem i think that the data is normalized to my Time Zone so I'm afraid it's better that you download
yourself the data from your NT8/Kinetic servers.
I think you can try it on any day and you'll see the diff between Playback and Backtest (even i enabled TickReplay On and same diff
between backtest and Playback.

So i want to know what to do to have SAME results for Playback and Backtest - this way i can optimize and get some reality settings that
will be close to Real Realtime.
Tell me if you need something more to do or send you.
Thank you very much for your help.
Best.

Isaac.

[IsaacBillion]

More Pics and Uploaded the XML template file to load all settings to Backtest.
It's a 15 Ticks TQQQ Data Series.

[NinjaTrader_ChrisL]

Hi Isaac, Thanks for the follow up. We have already given you all of the information on how you can get accurate results in a backtest. We will not be able to review or debug your strategy or results of the strategy for you. If you require further assistance, we recommend hiring a developer from the NinjaTrader Ecosystem website that will be able to work one on one with you.

https://ninjatraderecosystem.com/search-results/?fwp_category=programming-services

Please let us know if there is anything else we can assist with.

The NinjaTrader Ecosystem website is for educational and informational purposes only and should not be considered a solicitation to buy or sell a futures contract or make any other type of investment decision. The add-ons listed on this website are not to be considered a recommendation and it is the reader's responsibility to evaluate any product, service, or company. NinjaTrader Ecosystem LLC is not responsible for the accuracy or content of any product, service or company linked to on this website.

[IsaacBillion]

Hi Isaac, Thanks for the follow up. We have already given you all of the information on how you can get accurate results in a backtest. We will not be able to review or debug your strategy or results of the strategy for you. If you require further assistance, we recommend hiring a developer from the NinjaTrader Ecosystem website that will be able to work one on one with you.

https://ninjatraderecosystem.com/sea...mming-services

Please let us know if there is anything else we can assist with.

The NinjaTrader Ecosystem website is for educational and informational purposes only and should not be considered a solicitation to buy or sell a futures contract or make any other type of investment decision. The add-ons listed on this website are not to be considered a recommendation and it is the reader's responsibility to evaluate any product, service, or company. NinjaTrader Ecosystem LLC is not responsible for the accuracy or content of any product, service or company linked to on this website.

Sorry Chris.
I don't need a developer!
I'm a Senior .Net Developer myself with 20+ years of experience. I developed myself this quick sample.
Very annoying to hear you are falling back from what you have said and i quote (look above) "If you can, I will be happy to take a look given the steps to reproduce are provided."
I'm not doing this for myself. I'm doing it FOR YOU because you have a bug in the System that you get from the Backtest one result and from the same data you get DIFFERENT results
from PLAYBACK.
THIS IS A NINJATRADER BUG, not myne.
And I kindly ask you to check it out and to move it to your developers to :
1. Explain the public why from the same historical data you have 2 different results.
2. Find a way (some way) that both results will be the SAME.
3. Add a feature to NT8 to incorporate the PLAYBACK engine with the backtest that can produce REAL (as close to reality as can be)
Results that WON'T MISLEAD the customers of the good strategies they developed to smash it in their face when they're activating on PLAYBACK.

your answer now is very annoying and disappointing that shows lack of willingness to find bugs in NT8 and fix them.
Maybe Chelsea NinjaTrader_ChelseaB can you help?
In fact, I post the files only to quick the process because you can do it quickly yourself. you can take your own sampleMAcross , add the settings i provided
and see the bug yourself.
I wasted 2 hours doing so only to ease on you.
Regards.

Isaac.

[IsaacBillion]

rasdaka What do you think?

[NinjaTrader_ChelseaB]

Hello Isaac,

For a strategy to perform in backtest as closely as real-time, the data must be the same (meaning same exactly the same bars start to finish), the logic must be the same, the properties (like Calculate or inputs) must be the same, Tick Replay may need enabled if Calculate is OnPriceChange or OnEachTick, and the orders must be submitted to a 1 tick series.

Below is a link to a forum post that details.

https://ninjatrader.com/support/forum/forum/ninjatrader-8/strategy-development/94098-isfirsttickofbar-vs-onbarclose-for-backtest-live?p=773377#post773377

As well as a link to a forum post that demonstrates how to write the information to file to see what is causing differences.

https://ninjatrader.com/support/forum/forum/ninjatrader-8/strategy-development/100192-comparing-real-time-historical-and-replay-performance#post100192

Please let us know if you would like assistance analyzing output.

BarsRequiredToTrade ignores order submissions until there are enough bars, to give indicators a chance to start building up and producing values before orders start getting submitted. You can choose to set this to 1 if you want. However, it will need the same to get the same results.

[IsaacBillion]

Hello Chelsea.
I don't know what is happening with you guys , who writing answers WITHOUT ANY RELATION TO MY QUESTION !
I asked about PLAYBACK vs. BACKTEST ! Read the questions again and again.
They have the SAME CONDITIONS and SAME HISTORICAL DATA.
So why the hell there are horrible different results between playback and backtest (NOT REALTIME)???
Is backtest is a JOKE? TO MISLEAD the users??
I'm not the only one asking this!

https://ninjatrader.com/support/forum/forum/ninjatrader-8/platform-technical-support-aa/1039207-differences-between-historical-and-realtime-tick-processing?p=1039352#post1039352

and i'm sure there are others too.
So please show the users how they should make backtest results same like playback or... the better choice INCORPORATE THE PLAYBACK ENGINE IN THE STRATEGY ANALYZER BACKTEST no matter how much cpu cores will be required. Today I have 16 cores and there is no problem to buy Threadripper x64 cores.
Why this is happening at all? Didn't you ever opened your SimpleMACross and checked it YOURSELF in backtest and PLAYBACK and see TOTALLY diff results?
Very Annoying!

Regards.
Isaac.