Nearly Random Entry vs a High Probability Entry

[TriPack]

Quote from dbphoenix:



Yes, a true random entry will have an equal probability of going up or down. But so what? Unless you're going to scalp ticks, this information is of no use whatsoever.

If there were such a thing as a true random entry, at that moment price would need to have an even 50/50 probability of going up or down, and then you could flip the coin and go with either long or short and get the random entry. I presume this is what dbphoenix was referring to.

To the point made by damir, we may enter using a random number generator, but that doesn't mean the probability of going up or down is equal, just because we are using the random # generator. We may not know the actual probability of that particular moment, but the probabiliy doesn't revert to random, just because our entry method is random.

I sure would like to see somebody initiate and maintain a journal of this "entry doesn't matter as long as your exits are good" style of trading over several months.

Same here, regardless of whether dbphoenix participates, I'd like to see damir do profitable realtime random entry trading.

 

[TriPack]

Quote from damir00:



there is no edge in picking entries, none whatsoever. the only edge out there is money management, and the solution to that comes from you as a trader: what are you comfortable risking, holding etc.

That's a pretty heavy statement damir. I don't agree with it but I'm willing to hear you out on this subject. My personal belief is that there are edges to be had in both entries and exits with money management being a subset of the exit.

 

[nitro]

Quote from damir00:


...i'll wager you any amount you care to wager your "edge" is in fact a negative one and that the random entries will produce results that absolutely wipe the floor with your results.

you up for the challenge?

I will wager you that I will wipe the floor with you if you agree to 1 condition - that you trade REAL money on your so called random entries, not just claim that your are. So I will "flip a coin" each day, and tell you when to get in and which direction. It is up to you to get out. It is up to you how many times I "flip the coin" each day.

That is, your entries and exits have to verifiable (auditable) and that you have a blotter showing your entries and exits and P/L on each.

Up for the challenge?

nitro

 

[inandlong]

Will we be able to agree on what a random entry is?

My definition as it relates to this discussion would be saying 'you are long (or short) on this chart... trade.' Decide the entry before the name of the chart is given, give the name and go for it. I have never attempted it, but I am sure it can be done, because entry is nowhere near as important as trade management.

The one problem I see is that trade management for me includes reversing, which technically is an entry. But I think that if you are told to be long and you see quickly you should be short, then a reversal does not count as an entry, but as trade management.

 

[Lightningsmurf]

Ok, if you feel that you can successfully trade any entry simply by following good money management then it stands to reason that you'd be happy to let dbphoenix do your REAL MONEY picks for the day(pre-open). If you are successful under these sort of conditions then I would be more than convinced that your original post was accurate.

 

[nitro]

Picking entries pre-market is not random. Random means a random number generator picks the TIME of day to enter and he must use a market order at that precise moment.

We want the random time number generator under our control.

nitro

Quote from Lightningsmurf:

Ok, if you feel that you can successfully trade any entry simply by following good money management then it stands to reason that you'd be happy to let dbphoenix do your REAL MONEY picks for the day(pre-open). If you are successful under these sort of conditions then I would be more than convinced that your original post was accurate.

 

[ramora]

It would not only be helpful to define 'random' but also the timeframe that will be used to define success and failure of the test. What drawdown would define the test a failure? What profit would define a success? Over how many days?

Otherwise the test results on any day or number of days would also only be random.

Great thread!

ramora

 

[nitro]

What difference does it make what definition of random is? If the markets are random, he can have his cat walk on the keyboard for an entry and then use his money management to turn succesive trials of random cat trades of this kind into a winning week/month/whatever.

My idea is to write a simple random number generator that will take as input the day/month/year as a seed, e.g., 9022003 then output say 5 time of day/direction pairs entries, e.g., 10:31:43 Long.

I would be happy to write the software and upload it to ET. This way, we could see what the time of day entries for the day were and ALL of us could duplicate the results.

nitro

Quote from ramora:

It would not only be helpful to define 'random' but also the timeframe that will be used to define success and failure of the test. What drawdown would define the test a failure? What profit would define a success? Over how many days?

Otherwise the test results on any day or number of days would also only be random.

Great thread!

ramora

 

[Banjo]

" What difference does it make what definition of random is? If the markets are random, he can have his cat walk on the keyboard for an entry and then use his money management to turn succesive trials of random cat trades of this kind into a winning week/month/whatever."




That reminds me, once when I was using realtick with hot keys setup I had a list of SOX stocks up that I would trade all day. My cat jumped off a couch chasing a fly and landed on the keyboard which was in my lap , biggest freakin mess I was ever in. Don't let your cat trade for you.

 

[swoop[TR]]

I see where Damir is getting at, and I guess in absolute terms, every entry is random because we really never know if the next tick will go in our favor or not. The problem with random entry is the one of psychology, it's tough to buy in a downtrend and it's tough to short an uptrend.

It is true that exit is probably the most important factor in trading because only the exit will tell you if you have a profit or a loss.

It's a psychological issue, not a trading issue.
It's always better to see your trade go in the right direction right from the entry, but sometimes it will not, and sometimes it will for too short a period for you to take advantage of it.