On the Kelly issue, if Mountains offers sausages at £2 a pound isntead of the usual £2.30 a pound, I will be tempted to buy an extra bag for the freezer. But if they offer them at £1.50 a pound I will assume there is something wrong with them and not buy any.

If you're saying it's a too good to be true factor, surely if on the basis of good form study/stats, the rising odds available would cancel out the rising too good to be true factor (less winners but better payouts).

Getting better,

Whilst the scenario you illustrate is very real, it doesnt actually negate Kelly Criterion at all as that staking application is under the strict assumption that you have calculated the "genuine" chances of something happening. If the price is considerably further from your own price then doubt can creep in but this is doubt expressed over the "genuine" chances, not the stake size. Magician was implying earlier that Kelly Staking isnt suited to horse racing specifically, and Im not sure why that is. If a bettor has a harder time in defining the "genuine chances" then this scuppers the entire bet, regardless of the staking method.

I had £20k at 3.65, cant really afford to lose that so I am giving all the value back to the market by laying slighly lower in running after 10 mins. What a chump I am.

boys it looks like your drawers are holding up

JPG 05 Oct 19:41
Getting better,

Whilst the scenario you illustrate is very real, it doesnt actually negate Kelly Criterion at all as that staking application is under the strict assumption that you have calculated the "genuine" chances of something happening. If the price is considerably further from your own price then doubt can creep in but this is doubt expressed over the "genuine" chances, not the stake size. Magician was implying earlier that Kelly Staking isnt suited to horse racing specifically, and Im not sure why that is. If a bettor has a harder time in defining the "genuine chances" then this scuppers the entire bet, regardless of the staking method.

Looking back at the Baku v Ekranas game back in July, it was becoming ever more likely that this match was fixed as the odds were simply ridiculous. It makes calculating any "genuine chances" very difficult as the degree of accuracy cannot be determined because there are too many unknowns. Either someone with a lot of money that doesn't know what there doing is putting in offers at ridiculous prices, or the match is fixed (which was clearly the case).

I was taking advantage of the odds being 'way out of line' and betting on this game laying overs, when I saw the odds move in the opposite direction of what they should and moving even further out of line. I started trading out at a significant loss. This was the most blatant fixed football match I have ever been involved with.

http://site.forum.betfair.com/jive3/betex/ThreadsFrameset.jsp?forumID=25&threadID=2015418

On another note, anyone use optimal f?

those are not exact probs - but you get the point.

yes you coudl reverse these bets with Kelly staking - but you are committed at the worng price and getting out of them is not easy

with 2 much injury time,not enough injury time,dodgey pens ,last minute goals maybe the reffs have beaten you all to it ,and already cleaned up

The change in the offside rule has been there for a while now, and doesn't really account for the very recent increases in goals.

I think there are two other factors:
1. as has been noted, there is a greater difference now between the top and bottom clubs. There is now such a massive financial gain to be had for clubs in the Premiership (from TV money) and consequent loss for those in the Championship, that the gap has widened between the two divisions. Clubs that are promoted to the Prem are generally much weaker than those that are there season after season. Of course, the presence of the superrich owners of some of the clubs further exacerbates this inequality.

2. The superrich owners are, I would argue, non-purists who like to see lots of attacking football and goals.

The Magician (6) 05 Oct 22:15


Football / Aston Villa v Man City : Match Odds 05-Oct-09 20:00 05-Oct-09 21:58 12,250.10
Football / Chelsea v Liverpool : Match Odds 04-Oct-09 16:00 04-Oct-09 17:56 -5,000.00
Football / Everton v Stoke : Match Odds 04-Oct-09 15:00 04-Oct-09 16:55 14,437.83
Football / West Ham v Fulham : Match Odds 04-Oct-09 15:00 04-Oct-09 16:54 11,518.50
Football / Arsenal v Blackburn : Match Odds 04-Oct-09 13:30 04-Oct-09 15:26 -5,000.00
Football / Man Utd v Sunderland : Match Odds 03-Oct-09 17:30 03-Oct-09 19:26 28,960.54
Football / Hull v Wigan : Match Odds 03-Oct-09 15:00 03-Oct-09 16:55 -5,000.00
Football / Burnley v Birmingham : Match Odds 03-Oct-09 15:00 03-Oct-09 16:55 -5,000.00
Football / Wolves v Portsmouth : Match Odds 03-Oct-09 15:00 03-Oct-09 16:53 -5,000.00
Football / Bolton v Tottenham : Match Odds 03-Oct-09 15:00 03-Oct-09 16:52 12,901.32
Football / Man City v West Ham : Match Odds 28-Sep-09 20:00 28-Sep-09 21:54 -5,000.00
Football / Sunderland v Wolves : Match Odds 27-Sep-09 16:00 27-Sep-09 17:58 -5,000.00

+£45,068.29

nealry half way

.............................
if you are currently on the premium charge threshold

gross winnings £46,702.34
commission paid £1,434.05
commission generated £1,242.02
total charges £1,242.02

total charges 2.66% (well done miles below the premium charge threshold)

premium charge £8,098.45

sorry.....you won't get this back if your next 9 bets are losers

JPG 06 Oct 01:03


Magician,

I have no doubt that that problem when trying to get matched is very real. My point is, its not the Kelly Stake that is the reason or even made any more redundant.

Your quite right that if the 20/1 should be 40/1 then it will get matched at 20's no matter what the size, but this exactly the same if you were flat staking, betting for a fixed return or whatever method you used.

The problem here is simply weight of money affecting supply & demand. Theres no inherent aspect of the Kelly method that creates or worsens this problem for the bettor then any other staking method would do.


I cant agree with this - Kelly is a function to maximise long term bank - and as such as your bets prove successful your bet size increases.

You ability to win on horses races here is limited to the staking size of the losers - not the size of the pools or the winners.

the probelm with racing is the the only ( well almost only) people that take on big bets in racing know more than you....

and as such kelly will get to a bank size and then stay there for ever making zero profit.

eg, 500K bank grows to 1 million in 2 years - then stays at 2million for 5 years just treading water
but 500K never growing ( ie withdrawing winnings) makes 120K a year for 7 years

it is just an example /figures not exact or modelled but form experience such staking plans/results could be perfectly realised


A smaller

so if you kelly bank is 5000 in horse racing go for your life
but if your bank is 1 million

To pay premium charge on 2% commission would be very impressive indeed.

Magician,

Thanks for the response. Although I disagree here, I have a suspicion a lot of it is over the semantics and not the numbers! But Ill try to make my case all the same. :)

Your absolutely right that for your value bet to get taken in full, there needs to be enough losers to accomodate it elsewhere in the market. My point is that this identically true in the context of each individual bet regardless of stking method. For example:

Two bettors both with a 100K bank. Both are good enough to identify value and both have exactly the same view of every betting event. The only difference being that one does Kelly staking (lets say this ends up with a 1%-5% difference in stake size from one bet to the next.) The other flat stakes at 1% per bet.

On a proposition, the Kelly bettor wants to have 2K on but can only get 1700 on. The flat bettor can get his 1k on in full?

What are you saying here? Because the Kelly bettor cant get the full amount on he shouldnt bet at all? Or only do 1K? Remember, the Kelly Criterion within its assumptions & mechanics, caters for external barriers. In this case (a high barrier) the problem is quite easily overcome (as opposed to the low barrier) whereby the bettor just bets to the max allowed or Kelly stake whatever is the highest.

In the example you mention, the Kelly bettor once getting to 2m, can withdraw 1.5m (or more likely, in regular smaller amounts instead of in one go), start over again, and still outperform the other type of bettor on each individual bet.

Bookmaker Ladbrokes has said profits were down 58% after it suffered a barren start to the English Premier League football season.

The slide in operating profits to £22.4 million for the three months to September 30 comes after a winning start to the season by Chelsea and Manchester United and the low incidence of draws - the bookies' favourite result.

The update on trading emerged as Ladbrokes announced plans to raise £275 million from investors in a bid to reduce its debt position.

Ladbrokes said just four - or 6% - of the first 66 games of the season ended in draws, compared with the five-season average of 25%. Bookies prefer draws because punters tend to place wagers on winning results.

Firms have also been hammered by a run of wins for the league's most heavily-backed teams, as Chelsea and Manchester United secured victories in all but one of their seven league games in the period up to the end of September.

Ladbrokes chief executive Chris Bell said poor margins in horse racing and the weak economic environment also affected the company's profitability.

The tough trading conditions have led the company to step up cost saving efforts, including through a salary freeze until January 2011 and the withdrawal of its full-year dividend payment to shareholders.

Ladbrokes said revenues declined 15% in its UK retail business, although the 3% decline in over-the-counter bets was better than expected. The gross win margin, representing the amount left by losing punters, fell to 14% from 17.5% in the same period a year earlier.

The company, which has more than 2,000 betting shops in the UK, said the proposed issue of new shares would place it in a stronger position when it came to refinancing its remaining debt.

Net debt stood at £962 million at June 30, but this could fall to £687 million as a result of the rights issue, Ladbrokes added.

ttt

What was his bet on the man utd 28k win ?

Hope you recover from marathon tomorrow sufficient to spend some of it :D.....well ....at least buy 1 round ]:)

JPG - isn't the salient point that there are non-triers in football matches once in every 10000+ games but its often said there's (at least) one per race in UK racing? I think his point that you only get filled in volume at poor value is a very valid one (for horse racing) - because there's a LOT more subjective under-the-counter factors that one might have to be aware of and also aware that they are very difficult to quantify and react to (without extracting certain information from the behaviour of the market, which is something that the magician amongst others do with very great success as I understand it).

I couldn't conceptualise how you would operate Full kelly with a bank of 1m+ - surely it would be pointless since I can't imagine many possibilities (the very top divisions it might be possible but would by no means be easy) you would have to actually stake anywhere near the recommended stakes - the markets just wouldn't handle it. You'd end up skewing things badly and getting some small percentage of your recommended stake on each time, and also since (IMO) you are more likely to have edge in more obscure divisions, the problem would be doubly-compounded there.

Has this finished?

is that one million you have won over 2 weeks? congratulations magician.