stevehopwoodforex.com
https://www.stevehopwoodforex.com/phpBB3/
Print view

Crossfader thread
https://www.stevehopwoodforex.com/phpBB3/viewtopic.php?t=4110
Page 1 of 1
Author:  AnotherBrian [ Sat Feb 14, 2015 4:33 pm ]
Post subject:  Crossfader thread

I use a link on my desktop to get ot SHF, it takes me to my CrossFader Gold thread, the link is dead. I tried a search and nothing came up. Anybody know where it is?
Author:  Reaper [ Sat Feb 14, 2015 4:47 pm ]
Post subject:  Crossfader thread

Crossfader subforum as been removed from this forum completely. I don't know anymore than that. I guess for further information contact the powers that be.
Author:  dudest [ Sat Feb 14, 2015 6:59 pm ]
Post subject:  Crossfader thread

Steve locked the TOR thread (I think on Thur) saying that VM was either a crook or an (idiot?) for relying on optimized backtesting to make his EAs work. By that time VM had been banned and his posts deleted. Don opened a new thread at the same time to re-assure CF followers that it's all good. He got banned as well, and the entire Crossfader subforum was deleted.

IMHO, it was uncalled for. But I'm not the forum owner so that doesn't count for much...

Cheers
Author:  SteveHopwood [ Sat Feb 14, 2015 10:29 pm ]
Post subject:  Crossfader thread

There were umpteen crossfader threads - most of them dead.

The single live one left as I recall - and I really don't care if this was a mis-recall - was based on optimised backtesting.

Now, anybody stubbornly refusing to read this thread http://www.stevehopwoodforex.com/phpBB3 ... 83#p108983 explaining why Crapergy Tester backtesting does not work, or worse reading it and refusing to believe it, needs therapy.

vm's thread appeared to base the success of his ea on optimised backtesting. No, I didn't read it all, but post 1 was quite specific.

Optimising is great fun if you really have nothing to do on a wet Sunday afternoon.

It is the method used by criminals to make their ludicrous custom indicators look as though they actually work before parting you from $500 to pay for it. Innocent punters do not realise how good an indi that crosses a 10 moving average by a 5 moving average can look when passed through optimising. I could name it "Hyper-Sensitive Forex Index", code an EA to trade it, create a fantastic equity curve with optimised backtesting and market it for $999.99. The whole process would take 10 minutes. There are idiots out there who would buy it.

Play with optimisation folks. Run crapergy tester when you are next bored. Backtest your fav EA - or any EA. Enter different values into the Start, Step, Stop fields. Make a note of your best results.

Then ask yourself this: ok, I was trading this EA during the period I was optimising the EA; would I have known that these were the best settings?

Then ask yourself this: ok, I am using settings optimised by me on a bored Sunday afternoon recently; they showed how successful I would have been had I used them over the optimisation period - although I didn't they were for obvious reasons; I am not a total tit and so understand that market conditions change; how do I know that my highly optimised settings will be profitable over a future period equivalent to the past? Given that the only data held in the candle database is ohlc and open time? How much of Steve's hard won expertise over the last 9 years can I afford to ignore? In other words, how thick am I?

Now, I grant you that my understanding of how craptesting in crapergy tester works has only emerged over time. It is probably true that vm joined this forum before I gained this understanding.

Fact is, I have it now. Only an idiot tries to run a thread about an ea that depends on its apparent success for optimised craptesting, in the teeth of the howling gale of information here that it doesn't work.

Idiot?

Or crook?

I don't care.

Fact is, discussion about craptesting is unwelcome here. The only reason I do not subject noobs to instant ban fore even mentioning is that it would be unfair on the inexperienced.

So ask yourself this: how pleased do you think I was to find a member whom I know has commercial interests in promoting his own work elsewhere, was basing the so-called success of his work on optimised backtesting?

This thread is now locked and will be deleted as soon as I remember to do so.

In future, any mention of cf without prior consultation with me will result in the thread deletion. On a good day when I am in a good mood. Ask around if you are unsure what will happen if I am in a bad one.
All times are UTC Page 1 of 1