Backtesting irregularities

[Korattt]

Unfortunately, Steve isn't the messiah, he's..

a sports trader

Was Euler going with..... a liar?

I've done Steve's course and been part of his mentor group.

been?, as in no longer?, if so why?

[deansaccount]

a sports trader

Was Euler going with..... a liar?

I've done Steve's course and been part of his mentor group.

been?, as in no longer?, if so why?

Correct, correct again. I'm not full time, change in job and life situation. I believe the initial membership was for around 6 months, I let it expire,

[foxwood]

Life of Brian movie explained in 9 seconds https://www.youtube.com/watch?v=plZRe1kPWZw

[ShaunWhite]

Nope never seen it, quick google search has revealed I was not even born when it was released in 8th Nov 1979

I always feel a bit saddened by comments like this. 100yrs of great movies and most people choose to watch movies because they're new rather than because they're good.

Hollywood needs to close for a year to let people catch up.

[Korattt]

Korattt, I've spoken with people who have put every indicator from betfair and many others into machine learning software and I can tell you that a magic bullet indicator really doesn't exist.

true, but by piecing together a few BA videos & by studying the charts of the day thread, there is practically the fundamentals there to put oneself in positive trading territory more often than not, works for me now, (apart from Saturday’s for some reason)

[deansaccount]

Korattt, I've spoken with people who have put every indicator from betfair and many others into machine learning software and I can tell you that a magic bullet indicator really doesn't exist.

true, but by piecing together a few BA videos & by studying the charts of the day thread, there is practically the fundamentals there to put oneself in positive trading territory more often than not, works for me now, (apart from Saturday’s for some reason)

How many races per day (on average this time of year) do you trade?

[Korattt]

I can tell you that a magic bullet indicator really doesn't exist.

true, but by piecing together a few BA videos & by studying the charts of the day thread..

How many races per day (on average this time of year) do you trade?

depends if I’m able to find an angle in or not on the market

[ruthlessimon]

true, but by piecing together a few BA videos & by studying the charts of the day thread, there is practically the fundamentals there to put oneself in positive trading territory more often than not, works for me now, (apart from Saturday’s for some reason)

I think that's a very very dangerous assumption. In fact, I had your exact mentality - but instead of trading it, I took the time to formulate a basic strategy (in excel) that would test if "getting in" on every "graph of the day" (favs only) was profitable (with caveats i.e. market selection). I found it wasn't & was much much worse; compared to a strategy that just backed at "x" time

This is why you must challenge your assumptions Korattt, you need to really test what you "think" is an edge

[ruthlessimon]

Let's take the last two shorties. Not graph of day material, ugly markets




But IMO, Peter woulda have really enjoyed those two races.

[ShaunWhite]

works for me now

you need to really test what you "think" is an edge

He is testing it, and his BF balance is the evidence?

It might be gone next week and it might not have been there 6 months ago, but that's were continual reassessment comes in. The docrine of continual reassessment doesn't leave much space for the 'everything must be backtestable' way of thinking.

We mustn't end up here though .....

[ruthlessimon]

Perhaps; but I'm about furthering understanding & challenging trading dogma He's been trading 4 yrs, I'm expecting Webb style trading from Korattt with that level of experience (i.e. regular 3fig days & quitting the day job permanently)

True. I don't have the current understanding for dealing with "changes", but like Tony mentioned - I think this pattern has been there for yr's, but "concept drift" is certainly a bridge I wanna cross soon

[ShaunWhite]

Old data might tell you the climate, but the weather is what's tight up against the right hand side of the charts.

Take the WC late goals phenomenon: markets seemed efficient re late goals in the past (climate), something changed and opportunity arose (weather), the market adjusted to take account of the weather (opportunity gone). Your question now is has there been climate change, or will futher opportunity arise when the climate returns and everyone is still dressed for the weather?

I'm guessing still, but I think I'm not far off with that. It fits the continual reassessment idea more closely than anything else I've thought of so far.

[ShaunWhite]

This 'quick look at the forum' has blown away what''s supposed to be an afternoon off (shame the sun's gone in).
I'll catch up with you later.

[ruthlessimon]

But what's your defence for this?? (http://www.wikinvest.com/wiki/Recency_bias)

-k catch ya soon

[ruthlessimon]

by studying the charts of the day thread, there is practically the fundamentals there to put oneself in positive trading territory

Although the way I tested it was very crude (i.e. only price). I made sure the strategy was guaranteed to trade all the markets that would make graph of the day (short fav's only). Here's what the equity curve looks like for April, May, start of June



The hallmarks of a random strategy