How long do you need to paper trade a strategy?

How log=ng so far ?

* long

I think it depends on whether you can set up a model based on past results. With some systems you can and it gives you the degree of confidence required when added to your paper trade results.
I don't know what your proposed strategy is, it may not lend itself to the system builders on horseracebase and adrianmassey.

If you cannot do this, I don't think you can put a time on the paper trade because it depends on your own personality. If you are naturally cautious, you won't mind spending 8 months say paper trading but most people wouldn't have the patience for that. So a reasonable alternative is to do it for real with small stakes at first.

I think paper trading is completely useless, unless you plan to trade by only backing and laying prices offered by others, rather than putting prices up yourself.

The reason being that you will not be able to factor in the likelihood of your bets getting matched. Trading with real money albeit at limited stakes, is a much better option.

about 5 years

the odds range needs to be taken into account to determine the statistical significance of the available hindcast dataset.

i paper traded my system for 6 months with good results before going live.... almost 3 months in and its still going strong....

however, i still worry that it may have all been luck so far and could go tits up but so far so good....

I agree with investor entirely. I do very little simulation these days- I just start up based on any old theory with tiny stakes and see if I can make it work. It's more realistic, it's easier to stay interested and there's no temptation to make retrospective changes to your strategy to make it look like it would have worked when the reality was that it didn't. I also find that I generate new ideas this way.

Thanks for all the replies,I appreciate them.Its only been 2 weeks but i might try the small stakes route,cheers.

KEEP TO PAPER GAMBLING AT LEAST IF YOU LOOSE YOUR PAPER ,YOU CAN CHALK ON THE WALL

I agree just do it with small stakes.

As to how long for i really don't know.

My own £2 horse laying system went from £100 to £307 in 341 races , then from £307 to £279 in the next 300 odd so i was none the wiser as to its merit.

tobermory, nothing wrong with £100 up to £279 in 600 odd bets.

I'm in the camp with invester and aye robot in that paper trading is very likely to give you a false result.

Would you have got matched?  Would your money have affected the market (£2 jumps in front of you and soon the price is a new level)? Would you have had the bottle to place the bet?

The best illustration of this is those who look at 1.01 horses getting beat and work out they could make x profit blindly laying each race.  Reality is of course they would get matched on very few losing 1.01s as they would be too far down the queue.

Rather than paper trade you can bet with stakes down to 10p with Gruss. You can learn things by putting down small stakes that you dont with paper trading.

Does that mean some are making money blindly laying 1.01 on horse racing?

You can go down to 1p with Gruss.

I follow in-running prices on the horses and often put a 'lay-the-field' bet of 0.01p at 3.40 to save my head bobbing up and down trying to follow the price movements.

Agree with all of the above comments. Use money otherwise you find yourself making excuses why your strategy failed.

Silca Dan - yes I am sure the first man in the 1.01 queue is clearing up in the long run.  Similarly the person who is first to lay the horse at 1000 that can no longer win again will be well up in the long run despite the odd blip.

When you analyze this, it becomes obvious why this is the case.

I saw a horse (fav) get matched at 1000 to over £1300
the other day, after unseating rider. I find that
asyonishing.

*astonishing.

K - that is probably people freeing up funds having already layed the horse.

Totally agree in not paper trading - better to use low stakes instead

I don't think it's a case of how long but how many.
I can't see the difference between paper trading and actually placing a bet as
long as you use the same price reference, say Betfair SP.

As a rule of thumb I use 100 times the average odds you will be betting on, so if you betting just favourites at an average odds of evens the sample needed to get a fair idea would be 100, if your average odds of your selections was 5/1 you would need a sample size of 500 etc....

My advice:

Is there any logic to your strategy? That is, do you actually know why it might/should work? If not you need to think long and hard about this - as it's the key to success.

Thanks for the advice,so far its 16 winning days out of 18.I  fear  Sod's Law so I  shall stick with small stakes for a while yet.

To state the obvious, paper trading is not worth the paper it's written on.
So many unforeseen factors come into play when you start live betting. It's much better to start with small real stakes, than small or large paper ones.
You will learn much more and much faster about the actual or potential weaknesses of your trading system, when you are actually staking real money.

If you have found a factor that seems to be ignored, or
overlooked by the market, then a paper trial is unnecessary.
Go for it to small stakes and assuming you are correct,
gradually increase your stakes. It's that easy.

kenilworth  28 Aug 10 06:23   

If you have found a factor that seems to be ignored, or
overlooked by the market,.......

kenilworth

How would you know that you've found a factor that seems to be ignored, or overlooked by the market without knowing all of the factors that the market makers have used?

I think "seems " to be the operative word, don't you ?

about half an hour

top2rated, there are many factors to be considered and their degree
of importance are arguable. For instance some people think previous
results between protagonists have a bearing, others don't etc,
I concur with FINE AS FROG HAIR'S reading of my post.

about 10 seconds

Sorry to intrude though I have a question, let's say instead of using a strategy it's picking selections based purely on your form reading abilities. How many selections - days, weeks or months would have to past before an individual can be confident in there results not being purely luck? Thanks in advance.

i suppose if you consistently make profits year in year out
then youre on the right track.

tobermory, nothing wrong with £100 up to £279 in 600 odd bets.

yes  Ken , the overall results were good , just it was divided into 2 halves in such an odd way .

First 7 weeks excellent , next 7 very tough going . Would have been happier with it had it been just slow , steady, progress all the way . Maybe the result is what i might be able to expect from it over any spell of 600+ races but hard to be sure after 300+ of going nowhere in the 2nd half , and fear if i had another go might not win at all.

This was how it went btw....

https://spreadsheets.google.com/ccc?key=0AtEfsI5oIBTGdEhuWUlkdjcwVkFRMUhiQnpHVkFUa2c&hl=en&pli=1#gid=0

More to the point, if the last 300 of your bets had occurred first, then you would probably have given up the ghost before the 2nd 300 turned good.

Definitely.

Particulary as it was taking up 4-8 hrs a day.

i am currently cleaning up on the play money poker, cant believe how loose they all are!

The problem  with paper trading  is  you dont  get to see the real problems... !


You place you Lay bet  mid way through the match  60 secs  later  you place the Back side of your spread trade  on another   betting site...   Then  just as your abour  to undind  the 2 trades  at a profit..

The following happens..



1,   Cat jumps  on keyboard   hitting keys

2,  Cup of  tea  pours all over lap top

3,  Internet connection goes...

4   your mates  set the smoke alarm off.


and it all goes wrong! 




Now  were did I leave  betfair's  phone number   Grrrrr

Wednesday was a 'Daniel Powter' for my strategy,some fine tuning required.