Paper trading - How long?

[skinnyman9000]

Just wondering how long people generally paper trade for when testing a system? Or is it just a case of doing it until you're satisfied?

I ask because i am currently testing my own system on the dogs. I have a full time job and other commitments, so the time i can spend in front of the computer is minimal so my system focuses on dog racing in the early evening, around 4-7pm, although his will be getting later, to around 7-10pm in the near future.

So far ive done 12 2hr test runs following this system. Overall its making good profit, but the outcome from each day varies massively, highest winning days have shown 15point profit over 2hrs, and highest losing days have shown 12point loss over 2 hrs.

As stated overall the system has made 36points over 12 2hours runs, not massive amounts of money but at £10 a point its working out at £15 profit per hour, and there is no form reading etc, so there is no preperation before the 2hours, and its easy to do other things in between the races.

Despite the decent results i still cannot seem to take the plunge and begin betting, so im finding myself continuing to paper trade.

So i was just wondering, when do people take the plunge?

[Wittmann 44]

There are statistical tests that can show whether your results are the result of luck or indicate whether or not you are beating luck / chance. The chi test is the most well know. The bigger the sample / longer the testing the better. It is common sense that turning a profit after 50 bets can be just luck but doing it over 500 is a lot more likely to be due to your forecasting skill. The fact that you haven’t gone to real money makes me think that you yourself aren’t fully convinced yet.

[skinnyman9000]

Originally Posted by Wittmann 44

There are statistical tests that can show whether your results are the result of luck or indicate whether or not you are beating luck / chance. The chi test is the most well know. The bigger the sample / longer the testing the better. It is common sense that turning a profit after 50 bets can be just luck but doing it over 500 is a lot more likely to be due to your forecasting skill. The fact that you haven’t gone to real money makes me think that you yourself aren’t fully convinced yet.

Im not fully convinced in myself as yet. The reason being is because in the past a system has proven fruitful during paper trading, and has then completely fallen over once i begin to use it. For this reason i am hoping to get a large paper trial before any betting takes place.

Each 2hr run last for 12 races, therefore 12 bets. Which means so far i have made 144 trial bets with the system.

Stastically this system will make a healthy profit. However, the dogs also has a large element of luck to it, with only 6 runners over a short distance a 1/1 favourite can be knocked straight out of the running within 100yards of the race, or 3 dogs can pile into each other in the corner, leaving the 20/1 dog to win. I think the dogs has a massive portion of luck attached to it.

[AnothrSpacecat]

Ive made a very good amount of money over the years predicting the bad luck youre talking about skinny, who crosses the line first is harder to predict than who will turn the first bend in front, yet its what happens at the first bend that decides who has a chance and who doesnt.

The number of dogs who can go slowly into the first bend and nearly always take a hit, but then have the potential to pick up again and win are very very very few and far between. Yes we have stayers and those who run from the back in almost every race, but all of them need a clear run which they can only get up the inner or by going very wide, again these are running traits which can be predicted.

I'd be interested to hear your system as it doesnt involve any form reading! If its returning money and youre not form reading then there will certainly be room for improvement, which would be even more profit.

[Snodser]

For what its worth.........

Take a sum of money, money that you can afford to lose,thats your budget. Make your picks then divide what you intend betting per race in half.

Every time your a/c is bigger than your starting budget withdraw the excess.
If you do that over a long enough period you will have a good idea about how "good" you are.

Its painless & once thru the first cycle/season you can always increase the initial budget and go to bigger stakes.

However, never add to your a/c over the "period/ season".
Reduce your bets if you have to otherwise you will only be conning yourself. If you can't stay afloat then the "system" is not working.........
Bets of luck.

[skinnyman9000]

I think that might be the best way, place some money on one side for a trial run.

I conducted another 4 paper tests this evening, each with 12 bets in place. I returned +14pts, -9.8pts, +9.2pts, and +10pts, so a very successful day, with the running total now at +57.8pts after 16 trial runs.

So at this stage it would appear to be profitable, although 20-25 tests will be conducted, and then the recorded data will be played around with to get the best betting strategy from it, then i will begin an actual run

[Snodser]

If I may add, when it gets to "real money" common sense can go out the window. .

The vast majority of "punters" lose in the long run .

I've been flying high all season (rugby) but took my first big hits over the last 2 weekends. Partly down to factors I could not possibly forsee but partly because I got cocky..

Thankfully they happened early on so I was able to spread things about so my powder is still dry. Very good lesson though .

[skinnyman9000]

I've suffered from being greedy/cocky in the past, just placing that one more bet, just pushing it a little more, obviously bit me in the ass.

Now im much more controlled, even thinking of having profit/loss stops per day, so that once the target is reached then thats it, i walk away for the day

[AnothrSpacecat]

profit/loss stops arent the way to control things. your system might find 20 good things one day and 2 the next, its the selections that should be controlling your betting not the earth spinning around the sun

[skinnyman9000]

I understand that, but ive now completed 16 trial runs, all containing 12 races at a single track, so 12 races from belle vue, 12 from sunderland etc etc.

Now so far ive found that the profits tend to come in sooner in the meeting, and then the losses slightly later on. In 10 of the 16 trials it has taken 6 races before a loss has been made, so with the trial data available it stands to reason that having a 'stop profit' will make more sense.

I'm planning on taking 20-25 trials runs of 12 races each, and then playing around with staking etc in order to find the most likely best profitable system

[AnothrSpacecat]

if thats what the stats say then go for it mate


ill buy the ebook of you later when its a winning system

[skinnyman9000]

Haha, well IF its a winning system (which to be honest i dont even think exists, but its worth a shot) then i wont be telling anyone anything, lol

[skinnyman9000]

Well another two trials runs today, each with 12 races.

1st run shows 11.8pts profit, happy days, but then the 2nd run , well that shows -18.4pts!! Just goes to show how this system can really throw things around.

Ya see, with a profit stop of 10pts and a loss stop of 5pts i would have been 5pts up today, whereas without them i finished it 6.6pts down. The massive losses were due to 2 of the races, so with some tweeking we might see big gambles like that disappear, which would have left the 2nd run in decent profit too, ah well, 18 trial runs now completed, over 50pts profit too.

[Snodser]

Skinny, shit, or get off the pot

[skinnyman9000]

Another 2 trials runs producing +8.4pts and +2.6pts.

Thats 20 trials so far, i reckon its about time to give this one a go.