Staking Question

zuutroy

I'm using betfair for a soccer system at the moment and obviously I started with very low stakes. The question I have is that is there any inherent problems with staking a fixed % of your bank?
Example, say you start of with 100euro and you decide to stake 5% of the bank (backing only), so at the start you're going e5 at a time, but if it's succesful and you get to e200 you're betting e10 at a time.
Seems like it makes sense but I'm not 100% on the statistics of it. Does it have any problems like the Martingale system?

Robin1982

No money management system is perfect, but betting as a percentage of your bankroll is what I'd consider the best method.

Take a look at something like the Kelly Criterion:

(assuming you have a detailed history of your bets)

* Win Probability (W) - probability that the bet will be successful; divide the number of bets that returned a profit by your total number of all bets

* Win/Loss Ratio (R) - total positive bet amounts divided by the total negative bet amounts; average gain of the profitable bets by the average loss of the loss-making bets.

The equation:

Kelly % = W – [(1 – W) / R]

An interesting one I saw recently was the Maximized Asymptotic Return System (MARS), which seems to try to calculate the optimal Kelly bet.

zuutroy

Great link, thanks. Exactly what I was looking for.

zuutroy

Kelly % and MARS both say around 40% for my conditions :eek: They mention staking lower than this for lower volatility. Still 40% seems like a lot, though its a high-confidence low return bet.

ecksor

It seems very very odd that you'd have such high confidence on a bet though that it would come up with a value of 40%.

zuutroy

ecksor wrote:

It seems very very odd that you'd have such high confidence on a bet though that it would come up with a value of 40%.

It's an in-play trade, thus the 90+% confidence

Dodge

I think its best to come up with your own system. For trading purposes I allow myself more freedom that a strict percentage bet.

As long as you keep records you'll be able to get a system that works best for you (eg allow more for football than tennis)

The Kelly one above is a great starting point though

Robin1982

40% may seem high but its the nature of Kelly.

Many professional investors avocate using a maximum of 25% but in your case that would result in sub-optimal growth.

In high-probability (and time-sensitive/in-play situations) situations Kelly tends to become highly sensitive so if becomes more difficult to bet with positive edge (which Kelly assumes you are always doing).

However it is advised to use Kelly on a well-tested system. Before that, flat bets are better.

wb

I pretty much always use a % of my bank when staking (on horses). I work it out based on my strike rate though. For example, if my strike rate was 30% wins, I would use 3% of the bank. If my strike rate was 20% wins, I'd use 2% of the bank as the stake and so on. I've set a maximum of 5% of the bank as the upper limit. Anything with a SR below 10% uses level staking until the bank doubles, then it's 1%.

zuutroy

Have been using the fixed % on this system for a month now (200bets) and its going well...Overall return is ahead of what I expected at ~2% of bank per bet. Plan is continue to a bank level where my bets are limited by liquidity, then I plan to level off and go level stakes and draw out the excess each week as profit.

Had a total shocker yesterday losing 2 in a row....bye bye 36% of bank!!...Spent the weekend battling back though. Glad I stuck with the 20% of the bank option...a little bit easier on the nerves, even though I'd have more money if I used the 40%.

wb

Holy Crap! I never use more than 5% of the bank! I don't bet on soccer, but are you not afraid of a big losing run?

zuutroy

wb wrote:

Holy Crap! I never use more than 5% of the bank! I don't bet on soccer, but are you not afraid of a big losing run?

Well the stats dont lie (or so I'm told). I've trawled through the data for several seasons and several leagues and found an occurence rate of 87% and the markets on average offer odds equivalent to 83% so I can mop up the difference over the long term. The Kelly system discussed above suggests backing with 40%, but I use half that.
When the occurence is so high, a long run of losses is (fingers crossed!) very unlikely

Robin1982

zuutroy wrote:

...found an occurence rate of 87% and the markets on average offer odds equivalent to 83% so I can mop up the difference over the long term. The Kelly system discussed above suggests backing with 40%, but I use half that...

Doesn't the commission effectively cancel out that advantage? Even more so with a sub-optimal betting strategy?

zuutroy

Robin1982 wrote:

Doesn't the commission effectively cancel out that advantage? Even more so with a sub-optimal betting strategy?

No, my winnings would be made up of the gap between the 83% and the 87% and then BF take 5% of those winnings. I've explained it very badly and confused myself now!

hotspur

zuutroy wrote:

I've trawled through the data for several seasons and several leagues and found an occurence rate of 87% and the markets on average offer odds equivalent to 83%

Out of interest what occurence is that?

zuutroy

hotspur wrote:

Out of interest what occurence is that?

It's to do with the time between goals in games. I dont wanna go into detail as a lot of people read this, and the liquidity is poor enough already!

liamdubh

zuutroy wrote:

It's to do with the time between goals in games. I dont wanna go into detail as a lot of people read this, and the liquidity is poor enough already!

Is it a system you purchased or one you came up with yourself?

zuutroy

liamdubh wrote:

Is it a system you purchased or one you came up with yourself?

Came up with myself...dont even remember how at this stage! Terrible weekend though....put in a lot of time and ended up at the same bank after 40 bets. Difficult to keep motivation during the downturns.

Bateman

Time between goals...bad weekend...

You weren't trying to get in and out of "no next goal" in the Spanish football by any chance were you? At this stage of the season, it's wildly unpredictable, even much more so than in the Premiership, from my experience anyway.

When it comes to confusion though, I know how you feel. I have just realised that my tennis betting is roughly 30 times more profitable as my football betting this calendar year so far, despite the fact that [I thought] my football knowledge is infinitely better than my tennis knowledge.

Chucky the tree

Having only two possible outcomes in tennis is a big help though.

Bateman

Chucky the tree wrote:

Having only two possible outcomes in tennis is a big help though.

No doubt about that but I'd be betting more on under/over than match odds when it comes to the football.

Chucky the tree

Hating betting on unders, especially if i am watching a match. It feels wrong to sit there and hoping there isnt alot fo goals.

zuutroy

Bateman wrote:

Time between goals...bad weekend...

You weren't trying to get in and out of "no next goal" in the Spanish football by any chance were you? At this stage of the season, it's wildly unpredictable, even much more so than in the Premiership, from my experience anyway.

Sort of yeah. Spain was a minefield this weekend. Things just aren't as they should be this time of the year. I think I'll avoid anything to do with playoffs, relegation games, champions league spot deciders from now on.

Bateman

Chucky the tree wrote:

Hating betting on unders, especially if i am watching a match. It feels wrong to sit there and hoping there isnt alot fo goals.

Yeah it's normally overs for me. Unders tonight in the Pat's - Derry game though, was a banker.

zuutroy

Bateman wrote:

Yeah it's normally overs for me. Unders tonight in the Pat's - Derry game though, was a banker.

Do you go very deep in your analysis of the overs/unders? I'd imagine thats the only way of making a long term profit

Bateman

Not deep enough zuutroy, it hasn't paid for me this season apart from Irish football. Rethink needed for next season in Spain and England.

liamdubh

Irish football is well worth betting on.

Chucky the tree

MLS and Eircom league to last till august.

Bateman

Some decent MLS picks on this site if you're not already a member

http://www.online-betting-guide.co.uk/?tx11555

Only got into it relatively recently, small stakes for the moment.

Chucky the tree

Red bull are worth a punt against Chicago on thurdays