Forum

General discussion about Bf Bot Manager v3 software for Betfair, Betdaq and Matchbook betting exchanges
By echelon
#28660
Hi Wellsy,

You've raised some good points there. When I was trying out SQ strats, I wasn't using the unsettled markets check most of the time. It's only when DN reviewed one of my strats he mentioned adding that check.

Even if the spread of numbers in our sequence is based on historical performance (which typically should mean more fave1's than fave2's and more fave2's than fave3's etc.) the sequence itself is intended to introduce a sort of randomness into the selections. This means that in theory there should be an equal chance of hitting the sequence correctly whether we use settled check or not.

For the past month I have these numbers for UK races (any number of runners):

F1: 36%
F2: 23%
F3: 18%
F4: 12%
F5: 8%
F6: 4%

I wouldn't bother with F5 and F6 and F4 is dubious. So a possible sequence is:

1,2,1,3,2,1,1,2,1,3

5 chances to hit F1
3 chances to hit F2
2 chances to hit F3

That roughly ties in with the potential of each fave winning.

Whether we wait for settled bets or not, we're still going to have more of a chance of hitting F1 than F2 - hitting the sequence in the right place is just the luck of the draw as I see it. DN probably waited for settled bets, so it's very clear what's happening when analyzing the performance of the strats.

The thing I don't like about waiting for settled bets is that some venues have comms. issues etc. and all the races are left hanging for ages and eventually voided, meaning that we miss out on a lot of betting.

I've not done any comparison between waiting and not waiting for settled bets. If you want to try it...pick a sequence and build 2 strats in live, one with settled check and one without for 10p stakes and see which one wins ;)
By wellsy
#28665
Hi Echelon,

Good points again.

I personally gave up on SQ some time ago, I couldn't keep the variance to a respectable level - that being said, I probably had too many strategies firing, and in what can be a low liquidity market, may have been messing with my own selections inadvertently.

Having played around with a whole bunch of strategies (weight of money, market movers, etc), I do find myself coming back to the likelihood that with greyhounds and how unpredictable they can be, a semi-random (but based on probabilities) strategy may be as likely as any to succeed.

Your comment is also a good reminder that we're not trying to win every bet - but in a sample of 10 or 20 or 50 races, a couple of strategies working together should be able to capture a decent number of winners to yield a 30-40%SR, which should be profitable if hitting the 'right' winners.