I've been backtesting the oddball efs posted by Alexsis. The results look great. Does anyone have any realtime trading experience with this system? Are the results similar to the backtesting results?
Vince
First make sure you have the Time Template set up correctly.
The instructions are contained in the EFS itself. Use the Editor to open the EFS and read them.
Then right click on the chart and from the context menu select Tools->Back Testing...
When the Back Testing dialog box opens select the EFS to use in Back Testing, set the parameters and run the test.
Alex
Alexis, thank you for responding. I tried your suggestions and everything works perfectly! Thanks. One last question; is it possible to get back adjusted continuous data (hourly) out of esignal that goes back more than 3 months?
Vince
As of now eSignal provides only 60 days of intraday data, but that should
change in the near future when they will increase it to 5 months initially
and then build up from there to a maximum of 2 years.
Alex
How do you handle the missing bars problems in your code.
For example in May 01 there is a 4 bars shift between ES and $ADV.
This out of sync data remains for all bars before this date.
A simple request: odd ball has a good audience here in the forum.
Triyng to keep up with the latest releases or adjustments involving or not Dyna Order.
Would it be possible to open a folder in File sharing like for Pyratools or MM lines and ghater all the EFS , ( maybe with some .png copies from the detailed instructions of Alexis in the multipages forum?).
Missing data should not really pose a problem with this system. The "getValue" function used to pull the necessary data correlated the data between the two symbols (to the best of my knowledge).
The "outage" of the emini data a few weeks ago may have caused some problem for historical backtesting, but it should not pose any problem now with current data.
Mauro - From a personal point of view this is not an issue as I have already backtested the system and know its performance. As the data is synced going forwards the system should provide the correct signals (which are crosschecked with those from the Oddball site although I have found this no longer to be necessary).
Fabrizio - The three formulas that I posted originally when starting this thread contain all the necessary information and instructions to run Oddball.
They are also available here in the ACM Test group at FileShare.
Thanks Alexis & Brad
Alexis , I am trying to check by myself the merits of this strategy and of course I want to run a correct backtest.
Do you care sharing with us the results of your backtest ?
Mine is only valid from May 2 and the results are not very good but I understand that this is a very short period of time.
I have a little routine that solves the Missing bars problem but it is very time consuming. If a missing bar is detected the program asigns the last valid data to that bar, keeping the sync.
Mauro
I have dloaded the latest oddball_do.efs from doji3333 and tried to backtest but i got only zero numbers and no trade. Why ? Where is problem ?
I set various time templates and time bars but still got only no trade.
When i tested another systems everything works OK.
Comment