OZmium Sports Betting and Horse Racing Forums

OZmium Sports Betting and Horse Racing Forums (http://forums.ozmium.com.au/index.php)
-   Horse Racing (http://forums.ozmium.com.au/forumdisplay.php?f=5)
-   -   Betfair - Chasing Prices (http://forums.ozmium.com.au/showthread.php?t=15465)

Chrome Prince 18th February 2007 06:07 PM

Betfair - Chasing Prices
 
Hi all,

I've done some research on how the actual chasing of prices to lay horses on Betfair impacts the bottom line.

While bettors are clambering over themselves to snip a few cents here and there, I've come up with some amazing research.

Notwithstanding my hypothesis does not extend to those laying to a book or arbing or hedging, just those that are straight out laying.

As I've mentioned before, generally weight of money indicates a horse will win. Therefore getting "unders" about your lay price actually means less profit and more risk. On the flip side getting more "overs" on your win bets means less profit and more risk.

I have identified that this is where part of my profit is evaporating.

Here is my hypothesis:

The worse the win price on Betfair, the more profit. (to a point)
The better the lay price on Betfair, the less profit. (to a point)


This flies in the face of every mathematical solution we've been familiar with in the past.

Why is it so?

a) Betfair is an exchange, it is not a pari mutual pool, therefore the price on other runners is not necessarily influenced by movements on one particular runner.
b) Seasoned exchange users know where to get accurate pre post prices and know if they deduct or add on X percentage, they will make a profit.
c) UK pools are up to Caulfield Cup financial level at some meetings, so the weight of money (speculation) is greater with professional market setters offloading huge amounts into the pools. (Sometimes up to 100K on a single horse at a single price).
d) As there is as much as 5% takeout (commission) and as low as 5% minus 60% commission, there is a greater margin for error and still profit versus 14.5% takeout from the start.

I would appreciate any other ideas on this.

The proof

I took my last 500 win bets and my last 500 lay bets,which were staggered from Opening Bookie prices to greater percentages of profit or better prices.
These bets were left sitting in the market with no manual intervention.
Here is the proof by list, with the rank of 1 being the opening price and 2 onwards being better:

RANK LAYPROFIT
1 14.07
2 2.30
3 4.30
4 4.72
5 1.98
6 0.64
7 1.80
8 2.48
9 4.07
10 4.18
11 3.76
12 1.83


The exact opposite of normal expectation!

The figures correspond almost exactly on the Bet side of the ledger as well.

So why is it so?

I will answer this one after some feedback - I think I have an angle on why ;)

Chrome Prince 18th February 2007 10:24 PM

I would have thought this topic would generate a lot of discussion, but perhaps it's not of great interest???

To provide my answer, it is because having locked in the bookies SP, anything less means the horse has a better chance of winning (apparently the figures say so), any drifting would enhance the lay profit (apparently).

Bottom line, better price does not always equal better profit.

A real bombshell to conventional punting maths.

Remember we are talking about market movements after the opening prices, not the opening prices.

crash 19th February 2007 05:40 AM

After 4 days use of Betfair, I'm a bit disappointed. The pool sizes are mostly dreadful and the odds on offer for prices above $5 to $6 are like bookies prices, mostly short compared to Stab ect. where the pools dictate the SP price, not exposure fear. For short priced horses, the prices are generally good.

Looked at arbitrage possibilities and personally I wouldn't touch it on Oz. Betfair. Being left uncovered on either the lay or bet side seems to be a real risk, even on a Saturday. The markets here are too small and OS would be the go there but I'm not about to start trying to learn all about British racing and betting.

I did find a few juicy early odds here on Saturday on some runners that I knew were going to later wind in a lot closer to the jump.

Anyway, early days.

Chrome Prince 19th February 2007 09:18 AM

crash,

I agree with what you've said in part, however, to some extent each individual drives that market. Do not be afraid to throw a price into the mix very early, bets are traded in milliseconds so you might get matched at some stage.

I have thrown prices into the market and been surprised to find them matched even in the 20 to 25/1 range.

Case in point (not bragging because it was a fluke really), I wanted to have some fun in Mentality's race in Sydney, I had lay bets set on it, but the price kept drifting out, so I saw old Desert War was first up and remembered he had led all the way and won at juicy odds in the past and I couldn't remember but thought it was first up.

I got set at $6.60.

TAB price was $5.80 in the end!

I also had a bet on Fashions Afield.

Also if you are looking early at the prices, wait, like the TAB most money is traded in the last few minutes and prior to this people are trying to suck others in with very poor prices.

There is an excellent series of articles on Betfair trading at articlemotron dot com by Mike Davies. Type Betfair into the site search and they are there.
Makes for interesting reading.

crash 19th February 2007 09:43 AM

I noticed the early prices were a suck-in. However some were quite good. Lad of the Manor I picked up early at 27/1 which OK, the horse lost but at the odds it was still a good bet and more than double the Tote.

Mark 19th February 2007 10:49 AM

My 2c worth gents.

Our turnover (Aust races) will never ever reach the heights of UK racing. Simple population numbers tell us this.

Chrome, I can't agree that weight of money will indicate the winner. I know what you're getting at re value, but if this was true most winners would shorten on the tote, bookies, and Betfair. This just does not happen. Having said that, I have no proof.

You also stated that the price on other runners is not necessarily influenced by movements on one particular runner. If only this were true. Unfortunately the bots see to it that the market is immediately brought back into the 99% - 101% range, which makes it very hard to make a book.

Crash, if you can't get huge overs on longer priced runners, you mustn't be trying hard enough yet. I backed a horse at 120's that was only 20's on IAS, and even shorter on the totes. Minutes later I was able to lay it off at 32. The 'way overs' on bolters has just about done away with the favourite/longshot bias.

I can understand your disappointment as BF has not grown in the last 3 years, IMO. I think the BF pools have stagnated, and will not increase any further until BF is able to advertise nationally and some of the bad press has been rectified.

As for arbing, there are still opportunties around. There was a Cummings first starter on Saturday that I though had some chance. (it ran last!!) The official prices were 17, 31, 51. I backed it at 65 and laid off at 38, so had it going for nothing, just the way I like it.

AngryPixie 19th February 2007 11:27 AM

Quote:
Originally Posted by Chrome Prince

There is an excellent series of articles on Betfair trading at articlemotron dot com by Mike Davies. Type Betfair into the site search and they are there.
Makes for interesting reading.

Chrome

Thanks. Thought provoking ideas in this series of articles.

Pixie

AngryPixie 19th February 2007 12:36 PM

I was waiting for him to try and sell me something at the end though :)

Mark 19th February 2007 12:51 PM

From the above mentioned article.

"It is always more expensive to lay a horse than to back it, this has to be otherwise the book would not add up. So the lay price of a favourite at traditional 3/1 will be around 4.7 decimal..

If you lay this horse to lose for £1 at 4.7 and it does lose, you stand to make £1 minus Betfair commission of 5% so you get your £1 stake back and a profit of 95p. All very good.

However, if perchance the contenders do not perform, your horse suddenly wakes up and goes on to win, then you have to pay for every other horse in the race that lost. So your loss would be your £1 stake* 4.7 = £4.70"

How can a so called professional get this basic concept so wrong???
Your loss is only 3.70.

Mark 19th February 2007 01:03 PM

"In fact you will never get a lay price at SP full stop."

Plain silly.



Maybe I'm reading the wrong article, but this one is extremely basic and doesn't really tell you anything.

crash 19th February 2007 01:17 PM

[QUOTE=Mark]
Crash
I can understand your disappointment as BF has not grown in the last 3 years, IMO. I think the BF pools have stagnated, and will not increase any further until BF is able to advertise nationally and some of the bad press has been rectified.

QUOTE]

Thanks for the insight Mark. Like I said though, it's early days and I was expressing 'initial' thoughts from [naturally] a shallow experience of BF so far. I'm still are a long way yet from seeing into the depths of possibility that are, as you point out, there to be exploited within BF.

AngryPixie 19th February 2007 01:20 PM

Quote:
Originally Posted by Mark
From the above mentioned article.

"It is always more expensive to lay a horse than to back it, this has to be otherwise the book would not add up. So the lay price of a favourite at traditional 3/1 will be around 4.7 decimal..

If you lay this horse to lose for £1 at 4.7 and it does lose, you stand to make £1 minus Betfair commission of 5% so you get your £1 stake back and a profit of 95p. All very good.

However, if perchance the contenders do not perform, your horse suddenly wakes up and goes on to win, then you have to pay for every other horse in the race that lost. So your loss would be your £1 stake* 4.7 = £4.70"

How can a so called professional get this basic concept so wrong???
Your loss is only 3.70.

Mmm yes I skimmed over that bit on first reading. That is a silly mistake.

I wouldn't want any contributors to be scarred off though. Please share the resources you find. I for one am here for the diversity of views.

mad 19th February 2007 02:15 PM

$18 about Sky Biscuit on BF. $5.50 on Unitab, bit better elsewhere. Crazy stuff!!!

The opportunites are certainly there.

Chrome Prince 19th February 2007 05:56 PM

Quote:
Originally Posted by Mark

Chrome, I can't agree that weight of money will indicate the winner. I know what you're getting at re value, but if this was true most winners would shorten on the tote, bookies, and Betfair. This just does not happen. Having said that, I have no proof.

You also stated that the price on other runners is not necessarily influenced by movements on one particular runner. If only this were true. Unfortunately the bots see to it that the market is immediately brought back into the 99% - 101% range, which makes it very hard to make a book.

As for arbing, there are still opportunties around. There was a Cummings first starter on Saturday that I though had some chance. (it ran last!!) The official prices were 17, 31, 51. I backed it at 65 and laid off at 38, so had it going for nothing, just the way I like it.


Hi Mark,

I did some extensive research regarding prices versus winners.
This research was only on favourites.
My conclusion is based on value and not number of winners.
Let me explain entirely why I hold this point of view.

I did an exercise in the thousands of manually comparing the movements of TAB IAS and Betfair on favourites. I also tried two approaches in real time over a thousand bets (at the start of this thread).
What got me started was the Betfair Advantage Tool, and realising that the price comparison was to a degree flawed, I wanted to compare the best and worst prices and the actual fluctuations rather than just the final price available on Betfair.

As this data is not available in Betfair's historical data, I had to do it all manually (ugh).

But I did only record various prices, not every price (that would be near impossible).

I recorded favourites or joint favourites on IAS.

Backing every one on IAS at opening quote returned only -1.43% POT.
That surprised me greatly.

Backing every one at final tote price returned 0.87% POT.
That astounded me.

It would seem that IAS are a better judge than tote punters - I had held this belief for many years anyway.

Now what about Betfair?

Well the final available price meant nothing in the scheme of things - even though it showed a profit, so I set about throwing (small) real money at various percentages over and under the IAS opening price to back and lay.

Surely if you got 10% 20% 30% 40% 50% better than IAS - you'd kill them!
Surely if you layed in the same increments shorter than IAS - you'd brain them as well.

Here's what happened....

I lost $1300 - but it was well worth it for the learning experience.

This happened because over time a pattern emerged, when my "overlays" were matched I made a huge loss, and when my lay bets were matched I also made a huge loss.

The pattern was clear enough to me, that although weight of money didn't necessarily indicate a winner for this particular race, as a long term proposition as far as value or profit goes, it certainly was a very strong indicator.

Note these bets were left floating in the market for at least 30 minutes.

Determined to make a bit of cash to pay for my expensive exercise, I decided to fly with taking advantage of this knowledge.

So I reversed my situation and put my a lay and back bet into the market at a few cents over and a few cents under the IAS price. This was more of a "watching" alert as I was trading on almost every horse race in Australia at that point.

When the lay price was matched, if the horse kept firming to a pattern I backed the horse for a greater amount, because it was firming. What I was doing was getting one of the best prices available at the time on firmers.

When the back price was matched, if it kept drifting I would lay it for a greater amount.

I didn't do anything silly like lay it at huge prices over, just close to the set price.

I made $3,450.52 clear profit.

I may have made a boo boo somewhere, I may have encountered luck or been unlucky, but after recording a thousand races manually and watching the movements and seeing the impact on profit/loss, I am convinced that my hypothesis is true.

Now it certainly may be well off the mark as far as others in the market go, but it's certainly true with favourites. The movement that is, not the actual price.

As further confirmation, I did another independant test in real time as outlined in Post #1 in this thread which further reinforces my findings.

I have also done the same test on U.K. races using the openers of 5 fixed priced U.K. bookies.

Exactly the same pattern emerged.

Chrome Prince 19th February 2007 06:09 PM

Quote:
Originally Posted by AngryPixie
Mmm yes I skimmed over that bit on first reading. That is a silly mistake.

I wouldn't want any contributors to be scarred off though. Please share the resources you find. I for one am here for the diversity of views.


I don't necessarily endorse everything he says, and I have actually looked at these minute traders who snip a few cents here and there - it's not for me at all. Although you can stop/loss your position, you can quite easily be left with your tail in the wind. One bad decision will wipe out all profit for the day.
That's just my take on it.
I would rather approach it in another similar manner.

Having said that I really liked a lot of his rules and suggestions, not necessarily his maths.

Chrome Prince 19th February 2007 06:49 PM

Actually, last Saturday was a very good demonstration.

Most of the favourites that eased on Betfair lost, most that firmed won.
I know because I was laying favourites all day.
I now use the BetAngel market graph (free version) and make decisions from there.

Mark 19th February 2007 09:27 PM

Quote:
Originally Posted by Chrome Prince
Hi Mark,

When the lay price was matched, if the horse kept firming to a pattern I backed the horse for a greater amount, because it was firming. What I was doing was getting one of the best prices available at the time on firmers.

When the back price was matched, if it kept drifting I would lay it for a greater amount.

I didn't do anything silly like lay it at huge prices over, just close to the set price.

I made $3,450.52 clear profit.

.


Chrome, do I understand you correctly? I'll try some examples of what I think you're doing.

(a)You back say $20 @ 3.0, if it drifts to 4.0 you lay $25. or
(b)You lay $20 @ 3.0, if it firms to 2.5 you back it for $30.

The odds and amounts are fictitious, but have I got the general idea correct?
(a) Is more likely to lose because of the drift, and (b) is more likely to win because it has firmed.
So in effect you've laid (a) for $5 @ 7.00. And you've backed (b) for $10 @ 1.50.

ps It's late so I may be confused.
Look forward to your reply.

Chrome Prince 19th February 2007 10:26 PM

Quote:
Originally Posted by Mark
Chrome, do I understand you correctly? I'll try some examples of what I think you're doing.

(a)You back say $20 @ 3.0, if it drifts to 4.0 you lay $25. or
(b)You lay $20 @ 3.0, if it firms to 2.5 you back it for $30.

The odds and amounts are fictitious, but have I got the general idea correct?
(a) Is more likely to lose because of the drift, and (b) is more likely to win because it has firmed.
So in effect you've laid (a) for $5 @ 7.00. And you've backed (b) for $10 @ 1.50.

ps It's late so I may be confused.
Look forward to your reply.


Not exactly, basically I cancel (effectively) the unwanted bet/lay. This unwanted or "flag" bet is merely a $6.00 alert bet to give me a reminder heads up, it isn't to try and even up or take advantage of using that initial bet.

For example
(a)You back say $6 @ 3.0, if it drifts to 4.0 you lay $60. or
(b)You lay $6 @ 3.0, if it firms to 2.5 you back it for $60 is more like the ratios involved.

(a) Is more likely to lose because of the drift, and (b) is more likely to win because it has firmed.

Yes that's the premise.
I realise what you're getting at, that I'm actually betting against myself and reducing my own odds to a severe degree, however, using much better ratios of investment this is reduced.
Also on another issue, it is quite often that I go to lay a horse at say evens and it's already at $1.80, so in that case it becomes a bet straight away with no "flag" bet.

It's not perfect, but it's certainly a very good indicator to start with.

Mark 19th February 2007 11:04 PM

Ok, I'm with you now.

At what point in the drift/firm do you place the "real"bet. Do you have a set % move or price in mind to start with?
And is this how you made the 3 gorilla's?

Chrome Prince 19th February 2007 11:21 PM

Mark,

I'm a trader, but I'm not a trader - if you follow.

Betfair is essentially a "trading" platform these days with manual and betting bots trading tiny fluctuations in the market. That approach is not my cup of tea.

Within this barrage of money either way, people have lost sight of the "true" odds of a horse, whatever it might be. The true odds do not matter to the trader, but to me there is vast opportunity to profit.

These bots etc bring some prices to an unrealistic level and then the punters cause a marked fluctuation in or out.

I'm more a bookie trader, but identifying opportunites which are market driven at the same time.
Essentially I am still very much in newbie mode, but gaining information and experience all the time.

I have no set trader alert other than the matched initial bet makes me look at the chart of money or prices.
I do have a certain limit on lays though, a fluctuation out to $4.00 is my limit of exposure, anything more is too much risk for me, although I'm probably costing myself profit in the longrun.

The three gorilla's I made from following that procedure, I didn't follow it on Saturday and copped a hiding.

On post mortem, all the favourites (or most of them that won) should have been back bets and I should only have layed the drifters to $4.00.

It's all information to work with in the future and I'm still way ahead for the year.

There are prices being offered on a regular basis on Betfair that regular traders just couldn't make a longterm profit unless they are trading the fluctuations.

jfc 20th February 2007 04:13 AM

Quote:
Originally Posted by Chrome Prince
....
I did an exercise in the thousands of manually comparing the movements of TAB IAS and Betfair on favourites. I also tried two approaches in real time over a thousand bets (at the start of this thread).
What got me started was the Betfair Advantage Tool, and realising that the price comparison was to a degree flawed, I wanted to compare the best and worst prices and the actual fluctuations rather than just the final price available on Betfair.

As this data is not available in Betfair's historical data, I had to do it all manually (ugh).

But I did only record various prices, not every price (that would be near impossible).
....


http://data.betfair.com/

Assuming you meet their stringent 100 point turnover requirements, this free data includes Australia.

It has every matched price with volume and time range.

While hardly ideal it can be converted to market movements.

Chrome Prince 20th February 2007 04:32 AM

jfc,

I downloaded it and it only had US and UK races.
Can you confirm Australian races are there now?

jfc 20th February 2007 04:36 AM

Quote:
Originally Posted by Chrome Prince
jfc,

I downloaded it and it only had US and UK races.
Can you confirm Australian races are there now?


Yes.

For example the latest file has Randwick February 7 races.

Chrome Prince 20th February 2007 07:23 AM

Thanks jfc,

I downloaded two files and am delighted to find that on almost every race a market lay percentage of 120% exists and is readily available.

Time to get some sleep, and start attacking from a different angle all over again.

Mark had said it was there, and it is ;)

For the punters, a back market percentage of around 80% also exists.

WIN WIN (oops now I sound like Haychee) :D

zeditave 13th March 2007 02:41 AM

there's plenty of data in those historical files... you could be delving through them for months to find an edge unless you have a plan to begin with!

found a few other places for Betfair trading articles:

betfair dot info slash knowledge
laytheodds dot com
racingtraders dot com


All times are GMT +10. The time now is 04:48 PM.

Powered by: vBulletin Version 3.0.3
Copyright ©2000 - 2025, Jelsoft Enterprises Ltd.