Ah man, this is frustrating. I watched your video and debated on how to approach. I will attempt not bash and to explain what my intentions were/are:
First and foremost, the quantity and the quality of your backtesting data is extremely important. I understand that it is time consuming to backtest and it sounds like you invested a lot of time for those five weeks of data but five weeks is not enough to come to a "close to accurate conclusion" if your system could/would work. Imagine if you had cancer and your doctor offered you two drugs; one was tested on only 300 patient and cured 270 of them with very little short term side effects and not enough data on long term side effects. The second drug was tested on 30,000 patients and cured 75% it's short term side effects were slightly worst with almost no long term side effects. Let's assume you are 27 and 67. Which drug would you pick. Then by quality, I mean the timing of your tests. The MLB in Aprilis nothing compared to the MLB after the break. Rosters are extended, players have been traded, prospects are brought up, teams begin to dump, teams begin to grind, lineups are out of whack... I recommend oyu try this out in July, not now.
Then, I didn't realize you had success in the first week of the season and that is why I didn't stop by. Along with the fact that it was mentioned (see below) earlier that five weeks of data might not be enough, I most likely would've not mentioned anything anyway just purely out of courtesy. The reason why I am trying to help you is because it looks like you will make more mistakes. On to my third point....
When you lose confidence in a system, you do one of two things. Shut it down and paper trade it (bet air) or reduce your risk. Let assume you are driving down the highway and all of the sudden you hear a loud crack under your hood. What do you do, keep on driving, stop and check or slow down?
So to sum it all up. Do yourself a favor and either reduce your unit size or bet air. Then, learn either SDQL or find another way to test your data. I'll even volunteer and give you a short assessment of what it could take to run a thorough test. I don't need your exact variables and parameters, hell you can even provide false data if you fear I might steal your system. I just need to know roughly what you are looking at.
One more thing, I was warned to not do this and "waste my time on you". I decided not to, just because I would like to help.
I do not believe in luck so I am not going to wish you luck, I wish you success.
First and foremost, the quantity and the quality of your backtesting data is extremely important. I understand that it is time consuming to backtest and it sounds like you invested a lot of time for those five weeks of data but five weeks is not enough to come to a "close to accurate conclusion" if your system could/would work. Imagine if you had cancer and your doctor offered you two drugs; one was tested on only 300 patient and cured 270 of them with very little short term side effects and not enough data on long term side effects. The second drug was tested on 30,000 patients and cured 75% it's short term side effects were slightly worst with almost no long term side effects. Let's assume you are 27 and 67. Which drug would you pick. Then by quality, I mean the timing of your tests. The MLB in Aprilis nothing compared to the MLB after the break. Rosters are extended, players have been traded, prospects are brought up, teams begin to dump, teams begin to grind, lineups are out of whack... I recommend oyu try this out in July, not now.
Then, I didn't realize you had success in the first week of the season and that is why I didn't stop by. Along with the fact that it was mentioned (see below) earlier that five weeks of data might not be enough, I most likely would've not mentioned anything anyway just purely out of courtesy. The reason why I am trying to help you is because it looks like you will make more mistakes. On to my third point....
When you lose confidence in a system, you do one of two things. Shut it down and paper trade it (bet air) or reduce your risk. Let assume you are driving down the highway and all of the sudden you hear a loud crack under your hood. What do you do, keep on driving, stop and check or slow down?
So to sum it all up. Do yourself a favor and either reduce your unit size or bet air. Then, learn either SDQL or find another way to test your data. I'll even volunteer and give you a short assessment of what it could take to run a thorough test. I don't need your exact variables and parameters, hell you can even provide false data if you fear I might steal your system. I just need to know roughly what you are looking at.
One more thing, I was warned to not do this and "waste my time on you". I decided not to, just because I would like to help.
I do not believe in luck so I am not going to wish you luck, I wish you success.
