|
By:
phillyharper
The real problem is as has been previously mentioned is that your figures are derived from known favourites after the event. When actually supporting with cash before the race has begun, the horse that when you lay it at that point in time is the fav. but might not end up being fav. after you have placed your bet. A prepost favourite that has drifted so it is no longer the favourite post race is usually more likely to lose so it might even make your profits larger although you have ended up not laying the official favourite |
|
By:
tobemory
In fairness to phillyharper he has stated BSP so I dont't think your post ~ "You can't lay at those odds" is applicable. |
|
By:
Pitssburgh Phil.
Looking the example of backing the 10th favourite returning a 143%, can I not look at that significantly larger return and offer a hypothesis? Elsewhere, we are seeing returns in laying of 5-10%, sometimes more, sometimes less, but the 10th favourite at 143% return over 2300 bets seems to be significant? So my hypothesis for why that is happening would be that there is less action and money going on the outsiders, therefore, with less people 'guessing' their value, their value is further out than it should be. I guess my hypothesis would be that Betfair is a layers market, and with so few laying the outsiders (for high liability) they are offering large odds in order to entice in backers, and thus their overall valuation is wrong? |
|
By:
As someone has already pointed out you can't back the 10th favourite before the race has run. Position in betting at Betfair SP is only known as soon as the race is off - when it is too late to get your bet on at those odds.
Look at the 7:40 tonight. Which horse will be the 10th favourite? You want to Maharnee right now? Look again in 5 minutes: you want to back Ballinarh Boy? Look 2 minutes to the off will Trojan Rocket be the 10th fav? Even if you did have the knack of identifying the 10th fav early on you the fact that you would be backing the wrong horse a percentage of the time would mean that your would not be able to achieve 143% return. Yes there is merit in backing outsiders on Betfair because layers tend to give way over the odds (you can get 40/1 for a 20/1 shot etc). But just as with favourites it is far better to be more selective. Don't back all of them; back the ones that do have a chance of winning. This requires hard work. You seem to be looking for an easy punting-by-numbers solution. It can't be done so simply. You still have to work at it. I would suggest sticking with favs, or one or two particular jockeys or trainers. Learn everything about the fav / trainer or jockey. Once you know them you will know when they are more likely to win, and when the market have got them totally wrong. That will distinguish you from the crowd. That is how you can get an edge. There are very few flaws in the numbers of racing. If there are flaws they are so tiny that any mistake can negate the flaw. However, there are many flaws with favs (67% of them lose), and there are many flaws with trainers and jockeys. |
|
By:
ok CM i just read his later post now
as said though the fav changes pre race often enough, the 4th , 6th , 10th fav must change in almost every race multiple times |
|
By:
What if you were to back the 10th favourite (or lay the third favourite) 10 seconds before the off in each race? Would this be as close to the SP as you could get?
|
|
By:
Philly, what are the figures for the 9th fave and the 11th fave winning? Do they differ much from those for the 10th?
|
|
By:
I really can't see why those figures would be sustained .
Different outcomes over a couple of years for 10th favs 4th favs etc No one backs or lays horses on the basis that they are 4th or 10th in the betting order, just likely a completely random statistic. |
|
By:
Don't mean to be rude guys but I think you've completely missed a beat here!
To the original poster, this is my constructive reply: If you'd backed every favourite you would have got back 92% of your stake (and lost 8% which you correctly said)... this does NOT mean you can simply lay every favourite and expect an 8% profit for the following reasons: The 92% figure is based on actual bookmaker SP... when you lay on here you are ALWAYS going to be laying at much higher odds than this. You would have to almost pre-empt in some occassions just before the race starts which horse will be SP favourite, then lay it on here, however this is flawed in that the actual SP favourite is always the shortest price, so even after you've layed it it could shorten futher and you lose even more percentage. The only realistic thing you could do is lay every horse at Betfair SP, however their claim that Betfair SP is 20% higher than bookmaker's SP means that your perceived winning edge of 8% would turn into a loss of 12%... There is actually a site somewhere (I've deleted it from my favourites) and it compared Adrian Masseys figures in every scenario to actual SP and Betfair SP and it showed that in 2010 if you'd backed every favourite you would have made a profit of about 3%, but this fluctuates and changes every year, so you can't possibly know before each year if the edge would be to back every favourite or lay them. End result of this long winded post is that there are no long term edges following trends. There are short term edges, which will over time be negated by short term losing edges, so my advice would be to try and find an edge elsewhere other than by following trends. Ask yourself "would anyone else have thought of this?" and if the answer is "yes" then I can almost guarantee you that there is no edge... so try and be creative. |
|
By:
Phillyharper : To calculate the 5% commission do we simply say that on average, for every £10 staked, we profit £1.30, and remove the 5% comission from that average return, which is 6.5 pence. So we would profit an average of £1.23 for every £10 staked. At a turnover of 4466 bets, that is £5493 profit?
Did I go wrong somewhere? Yes, you've grossly underestimated commission as far as I can see. At a guess I would say 6.5 pence is around 10 fold too low. Imagine having 10 winning £10 lays each paying 5% commission (£5) followed by a loss of £90. You've now made £10 gross profit and paid 50% commission leaving you with £5 net profit which would not be unusual. Never under estimate the basic commission amount, it's huge! |
|
By:
@phillyharper
I don't know what you mean by this? Lay all of the horses in a race? Forgive me for not understanding, but could you please clarify? Yes, lay all horses in every race @ Betfair SP to the same set liability. This is too tedious unless you use a bot. There are plenty of cheap well constructed bots around these days. I use Betting Assistant. I guess I shouldn't be amazed, but I have posted a guaranteed winning strategy and had hardly a reply. How often do you get handed something that works faultlessly for free. Either the sharper minds are not so sharp, or the sharp minds are using it ;) You'll get a better return laying all horses below 10.00 for the same liability @ betfair SP. If in doubt, download a year's worth of Betfair SP historical data and do the exercise yourself. It's all there in black and white! P.S. Don't forget to allow for commission in your calculations. Best of luck. |
|
By:
Hi Beat The OverRound,
A couple of questions. Using betting assistant (gruss) you can't lay to a fixed liability only an approximate. Gary Russel posted in his forum ~ "In Excel it's up to you to calculate the stake required, liability staking mode is for manual bets only. If you want to lay to liability then use a formula to calculate it. For example insert the following formula in cell S5: =5/(H5-1) to always have a liability of 5." How do you do it? You state ~ "Yes, lay all horses in every race @ Betfair SP to the same set liability" This may work if you laid every race, however in my case only betting on Saturdays metro meetings since BSP was available in Melbourne would have produced a loss at level liabilities but a profit at level stakes. If I halt guss updating at the 4 minute countdown, the fav at that point if excluded will produce a level liability profit, not 5% though. So two problems, laying to an exact liability is not possible with gruss and level stake laying in Betfair is only possible if the level stake will produce the minimum liability required (30.00) and only if you place the stake and click take SP Say I wanted to have a level stake bet of 5.00 on each selection to BSP, only selections over 6.99 would get matched. If I upped the stake so all would be matched the liability on the rank outsiders would be horrendous. Any comments? Cheers. |
|
By:
Hi CM,
Sounds like you are from my neck of the woods (Aus). Using betting assistant (gruss) you can't lay to a fixed liability only an approximate. Gary Russel posted in his forum ~ "In Excel it's up to you to calculate the stake required, liability staking mode is for manual bets only. If you want to lay to liability then use a formula to calculate it. For example insert the following formula in cell S5: =5/(H5-1) to always have a liability of 5." How do you do it? Any lay bet at SP is to liability. So I use the excel interface and just specify say $100 as the stake. The stake then becomes the liability. In the above instance Gary is talking about the betting mode liability, rather than the Betfair SP liability. You state ~ "Yes, lay all horses in every race @ Betfair SP to the same set liability" This may work if you laid every race, however in my case only betting on Saturdays metro meetings since BSP was available in Melbourne would have produced a loss at level liabilities but a profit at level stakes. Are you sure about this? I have layed every metro race and every non metro race for a positive return. It may be the case that Melbourne metro is showing a loss at the moment, but I don't think that's going to continue if that's the case. For example, there have been a handful of horses scooped before the jump so their SP price has been well below even SP odds. I can guarantee, if you only bet metro saturdays but included all states you will be in front - longterm that is. If I halt guss updating at the 4 minute countdown, the fav at that point if excluded will produce a level liability profit, not 5% though. Yes because in Australia with low liquidity, the price on the longer runners is way off what the BFSP will actually be. This is designed laying all runners including the favourite. So two problems, laying to an exact liability is not possible with gruss and level stake laying in Betfair is only possible if the level stake will produce the minimum liability required (30.00) and only if you place the stake and click take SP Say I wanted to have a level stake bet of 5.00 on each selection to BSP, only selections over 6.99 would get matched. If I upped the stake so all would be matched the liability on the rank outsiders would be horrendous. Yes, it's possible. Set your stake to $30.00 for all runners and click lay SP, you don't want to lay them in the open market or you will be below Betfair's minimum stake. So don't lay them and choose the option Take SP, click on the pink SP box and specify $30.00 liability - and you're done. I think there's some confusion over stake, liability, live market convert to SP and actual SP bets. As long s you're laying SP to a minimum of $30.00 liabilty (stake), you'll have no problems at all. Whether the horse is 5/1 or 500/1, the $30.00 liability is the same. Only the amount matched will change to take you to exactly $30.00 liability. |
|
By:
Although I was confused, I am a bit too old now and injured these days, I believe that greater minds than mine have confirmed my own belief and experience that ITCANTBEDONE!
Thank you all for your confirming postings, and thanks to Mr OverRound for revealing his winning method- a very unsual and philanthropic gesture on here and certainly something I would not be prepared to do! If my own latest horseracing method fails- 2 unsuccessful lays in 192 selections- I will try it. Good luck to you all! |
|
By:
Beat The OverRound,
Just simulated your case ... with a fixed Liability of 100 euros, starting from the beginning of the year you would have lost ~ 25.000 euros ... |
|
By:
That's certainly not my results.
1. How did you calculate liability? 2. How did you calculate commission? Remember commission is paid on the profit on the race, not winning bets. 3. How did you handle deadheats? |
|
By:
@Acornel
How did you simulate that then? |
|
By:
1. Liability = 100 euros per horse having odds
|
|
By:
1. Liability = 100 euros per horse having odds less than 10
2. 5% commission ... OK, not correct To be quick I put 1% per winning bets (really don't think it will be less than that on average) 3. Not handled Result: down 1900 euros |
|
By:
Precision: Down 1900 euros from 1st of May until now
|
|
By:
I mean more specifically, how did you simulate it? With software? Manually?
I'm still not sure I entirely understand Beat The Overounds strategy of laying every horse to a set liability. I can write a trigger to do exactly this - to lay every horse at SP that is below 10. Will it not work if you do it to level stakes rather than to level liability? I'm interested in this system for its simplicity - are you suggesting this works on all horse races over time? UK, Irish, AUS? I'm trying to get my head around why this works - could you please offer your thoughts? |
|
By:
Manually ... it takes ages :-)
With a software: Excel |
|
By:
Where did you get all the data from? I'm no good at Excel. Wish I was!
|
|
By:
How did you calculate liability, something is skewiff.
Say odds are 10, it's 100/9 not 10 for liability. More precisely 100 / (odds -1) |
|
By:
Also did you take out all the non race field sp results, like beaten lengths and other products?
|
|
By:
Still too many betting by numbers strategies here. Believe me they are not going to work unless you have a specific reason as to why they work. If you can't justify why 10th fav is profitable then you should not use it.
I could list a dozen similar systems and show the figures. Most make a loss. Some have shown a profit but can not easily be operated in the future. I can assure you I have looked at these over the years and covered every angle. Most of these systems were devised years ago - when Betfair SP did not exist and the results of backing on the exchanges could not be accurately reported. Here is one: 4th+ in the Betting 10+ Runner Handicaps This one is best in the AW season. It is based on backing all horses 4th or more in the betting in 10+ runner handicaps on the sand. Why? Because in these races the winner is from the first 3 in the betting 52% of the time ... thus this bet will have a winner 48% of the time. And because the higher the price, the higher the Betfair advantage (the greater price difference over SP, e.g. 10/1 pays 14/1, 20/1 pays 30/1, 40/1 pays 100/1) then this system is going to payout over the odds on winners. The results at Betfair SP as of today is a 1.1% profit at 5% commission. So this one STILL WORKS! But remember the problem of identifying the contenders? How do you know which horses will be 4th+ in the betting at the off? Lucky Jim 2.1 Old Dog 3.75 Green Banana 5.6 House Boat 5.8 Water Wheel 5.9 Foot Rest 10.5 Cheesemaker 28.0 Horsey 32.0 No Fancy 48.0 Outlier 100.0 You don't know which of Green Banana, House Boat, Water Wheel will be 3rd favourite. You don't know which 2 of the 3 to include in your bet. This is what screws up the system. You can get a 1.1% profit (after 5% commission) if you made the selections perfectly but you can't do that as by the time the race is suspended the 4th+ in the betting can never be accurately known. |
|
By:
phillyharper --> http://form.horseracing.betfair.com/horse-racing/090911
Beat The OverRound, If odds = 5 and Liability = 100 then you are going to win 25 euros if your horse does not Win. Right? |
|
By:
Going through my results, betting to $100 liability there has been a profit of $6,098 this year laying the entire field at SP.
Laying those under $10.00, the profit has been $9,969.89 May to Sept under $10.00 profit is $4,195.59 for every $100 SP liability. |
|
By:
Yes, win EUR 25.00 if odds are 5.
But how can are results be so different? |
|
By:
Is that just with AUS races or are you doing UK and Irish too?
|
|
By:
The figures I just posted are for UK and IRE races only to match acornel's dataset.
The figures are in fact better on Aus races [;)] |
|
By:
What commission figure are you using there BTOR? I can see a 0.1% loss this year at 5% commission when laying the field at BSP, and a slight profit at 1% commission.
|
|
By:
I'm simply at a loss to explain -1900 EURO to +4,195 AUD
It's either the calculation or the data. My data reflects my betting records as well. I have a databse of all my bets and also another database of Betfair historical prices. These coincide give or take one or two minor prices. |
|
By:
Hello Phil,
I've browsed your posts often over at your other forum! I'm using 5% commission (for these figures), but 5% of race profit, not individual runner profit. Commission is calculated on race profit when laying more than one runner, so when you have a winning lay, but lose on the race, you pay zero commission. When you win on the race, you only pay comission on the race profit, not the entire portion of the winning bet. |
|
By:
I know this is potentially valuable information, but does anyone have the data and excel sheet they could send to me so I can run these tests? I'm using Adrian Massey at the moment and it's not perfect for this kind of backtesting.
|
|
By:
Beat The OverRound,
I am going to improve my calculation by using commission / race and not per winning bet. I'll keep you posted. |
|
By:
BTOR,
Let's take a simple example. These 2 races: http://form.horseracing.betfair.com/horse-racing/010511/Hamilton_Park-GB-Ham/1430 and http://form.horseracing.betfair.com/horse-racing/010511/Hamilton_Park-GB-Ham/1505 ==> what is the P/L per race (5% commission) if you LAY every horse having odds < 10 with a maximum liability of 100 euros? |
|
By:
Melbourne Australia metro Saturdays only, this year (All lays BSP)
283 races 3112 runners. Laying to a liability of 100.00 produced a loss of 485.24 Commission at 5% was 130.06 so a total loss of 615.30 Deleting the official favourite (283) from calculations which won 105 races, Resulted in a gross profit of 1525.56 Commission at 5% = 364.9 giving a total net profit of 1160.66 |
|
By:
http://form.horseracing.betfair.com/horse-racing/010511/Hamilton_Park-GB-Ham/1430
Zero commission as you would have lost $73.01 on the race. |
|
By:
Philly, you should be able to download the data here:
http://data.betfair.com/ You'd be well advised to try and learn some Excel skills, as they'll help enormously when it comes to analysing bets. |