help with expected losing run sequence, please

Depends on what you mean by ''a losing run''. I would imagine around 30 consecutive, but the losing run interspersed by the occasional winner can be anything.

If the losing run is anywhere near 30, it is clear they are not genuine 9/4 chances and you should be laying them.

Could get a losing run 30 on toss of a coin. Sounds a very wise expectation for 9/4 chances

There's too much attention to losing sequences anyway, the best test always should be maximum deficit

http://www.probabilitytheory.info/topics/winning_losing_runs.htm

I have created a little simulation program for these purposes.

With my simulation, I tested a sequence of 2,000 bets with a theoretical 9/4 probability. The distribution was an eye opener. The longest losing run was 45, so beware! ;)

It never hurts to err on the side of pessimism when it comes to these things. One (profitable) strategy of mine yields a long-term strike rate of 34% on UK horse racing, but last year I still managed a longest losing run of 34!

And as Kenilworth alludes, it's rarely one single losing run that decimates a punter's strategy. On most occasions punters slowly 'bleed to death' over time from a long-term downward cycle of losses interspersed with the occasional winner.

When I used to go to the casino I would wait for a losing sequence and then use a staking plan to cover the maximum amount of losers.

This was reasonably successful until one day I came across a sequence that blew me out.

The maths I used to calculate the longest losing sequence were the true odds of a given result happening multiplied by 12.
E.G.when punting a10/1 shot the max. losing run would be 120(12x10). This was a pretty good guideline, although obviously not infallible.

if i lay 3 winners in a row ill cut my throat

ive layed 5 horses on the trot and none were in the top 2 in the betting,was very painfull and took along time to get losses back,wont happen again once bitten

The average longest losing streak is a shade over 14.

"Average longest'" .. Hmmmmm

You need the strike rate for an accurate answer. If you assume 3.25 is about break even then the strike rate is about 28.57%. For 400 bets the expected longest is 15.

In a simulation of 100,000 trials for 400 bets the highest was 43.

My guess from simulating it is E(x) =14.2, Var(X)=3.5

Sorry, Var(X) = 3.5^2

These figures mean over 400 trials you are unlikely (3%) to see a losing run of greater than 21.

no one biting mate

thanks chaps. my football bets have been at an average price of 9/4 and have already hit a losing run of 13 and had their been an equaliser in one other match would have been a run of 25 losers.

i currently bet 5% of my bank on each selection and have come to the conclusion that i'm betting too much as a %. will drop to 2 or 3% i reckon. all bets are returning approx 15% POT so I don't think i should really be laying them as someone suggested!

thanks for your help

I make it 15

good for for you MWDS 2.5% is about the right level, you can see how punters underestimate losing runs and maximum deficits, good luck in the future

Did a small simulation for average longest losing run as well, came to 13.77 but only 155 observations.

Bin Frequency
7 1
8 0
9 4
10 13
11 18
12 15
13 28
14 24
15 17
16 12
17 7
18 6
19 1
20 3
21 1
22 0
23 4
24 1
More 0

RacingCert 21 Feb 13:44
http://www.probabilitytheory.info/topics/winning_losing_runs.htm

"In 550 throws you will get a maximum run of nine heads in a row."

What nonsense is that?

Search me.

45,3256

You might be interested in how we have solved the age old problem of losing runs in horserace betting, we have devised a software program that can be used for our phone line bets or for your own bets.
This coverbet software requires a horse to run well to get your full stake money back on a loser, running well can also sometimes mean falling, going well, pulled up, running on.
We no longer get losing runs in a traditional way and it now means full confidence when betting without requiring large betting banks.
The software is fully automated and works on the api.

Try this.............

http://home.iprimus.com.au/jfc2000/drawdown.htm

Not sure if it will help you but it's quick and easy!