Trading What I see !?
- ShaunWhite
- Posts: 9731
- Joined: Sat Sep 03, 2016 3:42 am
So have you stopped backtesting? The idea of paper trading is just to iron out any execution issues, if you're using it to test a strategy it'll take months to get a useable sample.
It's a little bit more tricky backtesting with scraped form data, but not impossible..ShaunWhite wrote: ↑Tue Nov 23, 2021 9:34 pmSo have you stopped backtesting? The idea of paper trading is just to iron out any execution issues, if you're using it to test a strategy it'll take months to get a useable sample.
my flaw is with pure backtesting, I fall into the trap of data mining, i'd rather think positively forward.....
- ShaunWhite
- Posts: 9731
- Joined: Sat Sep 03, 2016 3:42 am
There won't be anyone who can backtest that doesn't. And if backtesting is your weakness then fix the weakness rather than deliberately handicapping yourself by not doing it? It's not like you don't know how to make sure you're not just backfitting. It just doesn't make sense.
"thinking positively forward" sounds like management speak for guessing and hoping again to me.
So i've definitely not left backtesting behind, but it's not worked for me the last 6months, i've found many patterns, so the way i've been using it wasn't right. The problem is it doesn't make you broaden your mind to other possibilities, I just kept digging and digging for patterns, and I found lots of historical patterns in the random data!ShaunWhite wrote: ↑Wed Nov 24, 2021 1:39 am
There won't be anyone who can backtest that doesn't. And if backtesting is your weakness then fix the weakness rather than deliberately handicapping yourself by not doing it? It's not like you don't know how to make sure you're not just backfitting. It just doesn't make sense.
"thinking positively forward" sounds like management speak for guessing and hoping again to me.
like the little Acorns recommended by weatherfox OR the dutching by mobius grey . with a robust selection process and the staking formulae (square root ).
have you taken them onboard?
- ShaunWhite
- Posts: 9731
- Joined: Sat Sep 03, 2016 3:42 am
I'm afraid this is where a lack of methodology is letting you down. Backtesting isn't where you look for edges/patterns. Backtesting is confirmation of your analysis. The analysis is designed around your theory or idea. And your idea comes from observation or other inspiration. Each discrete step asks specific questions and involves the minimum amount of work to answer that question. And you only proceed to the next stage if the answer was satisfactory. Any wonder I'm suprise you're only doing paper testing which is actually just the final shakedown. Everything before that has been consigned to the bin.
You might say that's slow and long winded but the tortoise will always beat the hare and I'm afraid you're proving that. Tbh it's nothing special, it's just a fairly standard problem solving methodology. The above won't suit people without the necessary tech or be necessary if they're implementing something they do manually but that's your situation.
We keep mentioning backtesting and re being hasty, you did some for about 6 months and didn't get much added value from it. That's because you just started running tests. If you'd spent 3 months(!) just thinking about what it could tell you, all the possible pitfalls and a solution for each, checks and recs, limitations, tools you'll need etc and learning how to test....and then tested, then it wouldn't have been such a waste of time and you'd have a reusable skill.
I know it always sounds like a bollocking goaty and I'm sorry about that because for you this is just a bit of fun, but I really wish you'd stop doing and start thinking.
I agree with part of this, but you're missing my objective. I agree with "your idea comes from observation or other inspiration", and this is what i'm trying to do, work out how to develop a model and how to use it. Peter's re-posted video https://www.youtube.com/watch?v=DZEiNaXVWSw is what I am trying to do, and as he states he's dubious on backtesting and rather prefers to dabble in the live market. Paper testing lets me do that with reasonable accuracy without losing money!ShaunWhite wrote: ↑Wed Nov 24, 2021 4:04 pmI'm afraid this is where a lack of methodology is letting you down. Backtesting isn't where you look for edges/patterns. Backtesting is confirmation of your analysis. The analysis is designed around your theory or idea. And your idea comes from observation or other inspiration. Each discrete step asks specific questions and involves the minimum amount of work to answer that question. And you only proceed to the next stage if the answer was satisfactory. Any wonder I'm suprise you're only doing paper testing which is actually just the final shakedown. Everything before that has been consigned to the bin.
You might say that's slow and long winded but the tortoise will always beat the hare and I'm afraid you're proving that. Tbh it's nothing special, it's just a fairly standard problem solving methodology. The above won't suit people without the necessary tech or be necessary if they're implementing something they do manually but that's your situation.
We keep mentioning backtesting and re being hasty, you did some for about 6 months and didn't get much added value from it. That's because you just started running tests. If you'd spent 3 months(!) just thinking about what it could tell you, all the possible pitfalls and a solution for each, checks and recs, limitations, tools you'll need etc and learning how to test....and then tested, then it wouldn't have been such a waste of time and you'd have a reusable skill.
I know it always sounds like a bollocking goaty and I'm sorry about that because for you this is just a bit of fun, but I really wish you'd stop doing and start thinking.
- wearthefoxhat
- Posts: 3221
- Joined: Sun Feb 18, 2018 9:55 am
The gift that keeps on giving..(2 out of 3 today)
You do not have the required permissions to view the files attached to this post.
I agree with Peter's video - develop an idea/strategy and 'forward test'/analyse your idea as you develop it going forward.goat68 wrote: ↑Wed Nov 24, 2021 4:44 pmI agree with part of this, but you're missing my objective. I agree with "your idea comes from observation or other inspiration", and this is what i'm trying to do, work out how to develop a model and how to use it. Peter's re-posted video https://www.youtube.com/watch?v=DZEiNaXVWSw is what I am trying to do, and as he states he's dubious on backtesting and rather prefers to dabble in the live market. Paper testing lets me do that with reasonable accuracy without losing money!ShaunWhite wrote: ↑Wed Nov 24, 2021 4:04 pmI'm afraid this is where a lack of methodology is letting you down. Backtesting isn't where you look for edges/patterns. Backtesting is confirmation of your analysis. The analysis is designed around your theory or idea. And your idea comes from observation or other inspiration. Each discrete step asks specific questions and involves the minimum amount of work to answer that question. And you only proceed to the next stage if the answer was satisfactory. Any wonder I'm suprise you're only doing paper testing which is actually just the final shakedown. Everything before that has been consigned to the bin.
You might say that's slow and long winded but the tortoise will always beat the hare and I'm afraid you're proving that. Tbh it's nothing special, it's just a fairly standard problem solving methodology. The above won't suit people without the necessary tech or be necessary if they're implementing something they do manually but that's your situation.
We keep mentioning backtesting and re being hasty, you did some for about 6 months and didn't get much added value from it. That's because you just started running tests. If you'd spent 3 months(!) just thinking about what it could tell you, all the possible pitfalls and a solution for each, checks and recs, limitations, tools you'll need etc and learning how to test....and then tested, then it wouldn't have been such a waste of time and you'd have a reusable skill.
I know it always sounds like a bollocking goaty and I'm sorry about that because for you this is just a bit of fun, but I really wish you'd stop doing and start thinking.
Back Testing is a waste of time imo.
- Realrocknrolla
- Posts: 1903
- Joined: Fri Jun 05, 2020 7:15 pm
I don’t think back testing is a waste of time. But i do think it depends on what strategy/sport you are testing.
Is checking a horse form over a certain period of time back testing?
Is checking if a horse has DOBBED more than 50% of its runs back testing?
Is looking for an average goal time in the last ten meetings of a football match back testing?
I never back test dogs though
Is checking a horse form over a certain period of time back testing?
Is checking if a horse has DOBBED more than 50% of its runs back testing?
Is looking for an average goal time in the last ten meetings of a football match back testing?
I never back test dogs though
How many successful strategies have you developed using 'back testing' ?Realrocknrolla wrote: ↑Wed Nov 24, 2021 7:08 pmI don’t think back testing is a waste of time. But i do think it depends on what strategy/sport you are testing.
Is checking a horse form over a certain period of time back testing?
Is checking if a horse has DOBBED more than 50% of its runs back testing?
Is looking for an average goal time in the last ten meetings of a football match back testing?
I never back test dogs though
- Realrocknrolla
- Posts: 1903
- Joined: Fri Jun 05, 2020 7:15 pm
Plenty if the rules above apply.Cardano wrote: ↑Wed Nov 24, 2021 7:28 pmHow many successful strategies have you developed using 'back testing' ?Realrocknrolla wrote: ↑Wed Nov 24, 2021 7:08 pmI don’t think back testing is a waste of time. But i do think it depends on what strategy/sport you are testing.
Is checking a horse form over a certain period of time back testing?
Is checking if a horse has DOBBED more than 50% of its runs back testing?
Is looking for an average goal time in the last ten meetings of a football match back testing?
I never back test dogs though
Define a time frame on back testing.
Backtesting allows you to simulate a trading strategy using historical data to generate results and analyse risk and profitability before risking any of your balance. I would recommend paper trading if you find anything that has legs and bin something that doesn’t. Also you have to be honest with yourself going forward otherwise a strategy you are running won’t be true.
As Peter said in his video, he has tweaked and branched off many times. Probably like a lot of us.
I have a Devil Bot. I called it that because it is version 6.66.
I would recommend 'forward testing' combined with analysis as Peter advises.Realrocknrolla wrote: ↑Wed Nov 24, 2021 7:46 pmPlenty if the rules above apply.Cardano wrote: ↑Wed Nov 24, 2021 7:28 pmHow many successful strategies have you developed using 'back testing' ?Realrocknrolla wrote: ↑Wed Nov 24, 2021 7:08 pmI don’t think back testing is a waste of time. But i do think it depends on what strategy/sport you are testing.
Is checking a horse form over a certain period of time back testing?
Is checking if a horse has DOBBED more than 50% of its runs back testing?
Is looking for an average goal time in the last ten meetings of a football match back testing?
I never back test dogs though
Define a time frame on back testing.
Backtesting allows you to simulate a trading strategy using historical data to generate results and analyse risk and profitability before risking any of your balance. I would recommend paper trading if you find anything that has legs and bin something that doesn’t. Also you have to be honest with yourself going forward otherwise a strategy you are running won’t be true.
As Peter said in his video, he has tweaked and branched off many times. Probably like a lot of us.
I have a Devil Bot. I called it that because it is version 6.66.
Wear-the-fox-hat …apologies for slaughtering your name. I find the little acorns strategy one that has proven all kinds of weather just like a foxwearthefoxhat wrote: ↑Wed Nov 24, 2021 5:58 pm
The gift that keeps on giving..(2 out of 3 today)
Weth 1.15.png
Dundalk 2.00.png
Weth 4.20.png
the results vary depending on how you define a favourite and at what time you enter the market .
My selection process has now been extended into the in -play trading window. There is more volume and higher levels of participation in this markets and I guess traders have gained more confidence provided by TPD data .