Forums

General Betting

Welcome to Live View – Take the tour to learn more
Start Tour
There is currently 1 person viewing this thread.
hummingsox
13 Nov 09 16:45
Joined:
Date Joined: 24 Jul 06
| Topic/replies: 60 | Blogger: hummingsox's blog
Hypotheticaly, If I backed every dog at one particular track that was 3/1 and in the same race laid every dog that was 6/1, I could expect roughly that my backs would win 25% of their races and my lays just over 14% of theirs. right ?

Now suppose that trap 1 won 19% of the races at this track, and trap 2 won 14%.

If I backed every trap 1 that was 3/1 and laid every trap 2 that was 6/1, what would the odds be against these runners now.

This is hypothetical of course, but I am interested to know what you think, and which takes precedence, the trap advantage, or the price.

thank you.

hummingsox
Pause Switch to Standard View question for stats boys
Show More
Loading...
Report pots November 13, 2009 4:51 PM GMT
3/1 shpts do not win 25% of races. if that were the case they would be 7/4
Report kenilworth November 13, 2009 5:34 PM GMT
Exchange price 3/1 chances would show 25%.
Report hummingsox November 14, 2009 1:41 PM GMT
Anyone any opinion on this please ?

cheers,

hummingsox
Report Muppetier November 14, 2009 1:52 PM GMT
The trap bias is included in the price so a 3/1 in trap1 will have about the same chance as a 3/1 in trap2.
Report aye robot November 14, 2009 1:54 PM GMT
You would have to work on the assumption that all the information is already built into the price, IE any stall bias is accounted for, and the odds of the selection winning at odds of 3/1 are still 0.25 (25%).

Personally when I'm working on a new theory/strategy I start by assuming that it's zero sum (offers no value) then try to prove otherwise by gathering and anylising data. If I can prove that it's anything other than zero sum (+ or -) then theorhetically I can make money from it if I can get the bets matched (which is the difficult part).
Report hummingsox November 14, 2009 3:07 PM GMT
aye robot, thanks for that.

would you just confirm then in that case, that based on the figures in my first post if I have achieved 87% more winners than losers, there must be some element at work other than just the prices.

And would you make the assumption that the trap advantages were probably that element.

Thank you for your time, much appreciated

hummingsox
Report zipper November 14, 2009 3:21 PM GMT
Heres my take on odds .. give you four ...... evens 50%...... 2/1 33% 7/2 22% 12/1 8 % thats in theory .. over a long time they all return much less than theory ... EG evens dont win 50% of the time ......
Report hummingsox November 14, 2009 3:32 PM GMT
Hi Zip,

I cant argue with you on that, they are just ballpark figures, something to use for comparison purposes etc, though I think they are probably within 3 or 4% over a LONG period.
Report theonlywayforward November 14, 2009 3:33 PM GMT
It would depend on which sp you use to calculate your % .
Report theonlywayforward November 14, 2009 3:44 PM GMT
As there are more factors creating the sp than the draw adv.
Report zipper November 14, 2009 3:57 PM GMT
Hummingsox ....... lets take your 3 or 4 %... so evens win 46% .. 2/1 win 29% 7/2 win 18% and heres the real killer 12/1 win 4% ......... By the way 4% of a million is £40.000..... must punters good punters and i meam very good punters .... can only make 10% of turnover ..
Report kenilworth November 14, 2009 4:22 PM GMT
must punters good punters and i meam very good punters .... can only make 10% of turnover ..

Very, very few people make 10%, more like 2% to 5% I assure you.
Report aye robot November 14, 2009 5:14 PM GMT
"if I have achieved 87% more winners than losers, there must be some element at work other than just the prices"

Sorry to be equivocal, but it depends....

To start with your terminology is a bit odd- when you say "I have 87% more winners than losers" I assume that you mean that about 65% of the bets you place are winners- which is a better way of expressing it.

Now, if all of those bets were at odds of 4 and there were a lot of them that would be a fairly draw dropping record, but I suspect at least one of those things isn't true.

What you need to do is compare the rate at which you are actually winning with the rate at whcih you should expect to win if the prices are "true" and you need to have a large sample to do this with. If you're betting at around 4 then you'd probably want about 500-1000 bets (actual bets- not "back testing") under your belt before you start to relax.

Once you have a large sample you can anylise it for statistical significance using a variety of techniques, but personally I don't bother. Once I've established that a strategy has a decent chance I just give it a small bank and stake a small percentage of that bank on each bet (usualy 2%) - the winning strategies grow, and the bad-uns go extinct.

If your strategy is good you will find that the bank will grow incredibly quickly even if it only shows a small edge, although as stakes get larger things get harder.

Consider this:
If you have a 4% edge and you stake 2% of your bank on each bet you will double your bank about every 100 bets (ignoring comission).


Above all- stay away from "staking plans" - by which I mean anything that says "If you win do X and if you loose do Y". Plans like this are always a bad idea- and yes I do mean always, there are no exceptions.
Report FatalStatistics November 14, 2009 9:49 PM GMT
good thread, wise words from the 'robot' again !...
Report hummingsox November 15, 2009 12:10 AM GMT
aye robot,

Many thanks for your most helpful thread.

I can confirm that the win/loss ratio is 65/35 and that it is indeed based on actual figures not backtesting and also that is all based on betfair prices, over a 2000 race.sample.

Would'nt use a staking plan under any circumstances.

Cheers, I owe you one.

hummingsox
Post Your Reply
<CTRL+Enter> to submit
Please login to post a reply.

Wonder

Instance ID: 13539
www.betfair.com