- Sun Dec 03, 2023 10:38 pm
#28658
Hi Nigel,
Unfortunately it appears that when DN deleted his ID all his posts were deleted too - which makes reading threads like this one a bit tricky!
The key thing DN was trying to get across by using SQ is that it's one of the simplest ways to create a strategy which has a good chance of working. It helps avoid the sort of variance which sets your balance plunging into the red on a simple strategy like Back fave 1 or lay fave 2 etc when it hits a bad run.
As far as I recall, we need to forward load the sequence with the faves which are most likely to win and scatter a few random faves here and there.
For example, don't have a sequence like 4,3,2,1,3,5,2,,, it's better to have 1,1,2,3,1,2,4,1,3,2,1,1...
I've looked at dogs history over a long period before and you may not find it as useful as you anticipate. What I find better is to filter out AUS and NZL markets in the Results tab and use the view statistics button to get a flavour of the winning % of each fave. DN implied that he doesn't look at any distant history and also mentioned that he tweaks his strategies quite regularly when their SR drops - my view is that the previous month should be enough data.
You can create sequences using traps too - though you're going to see more variance that way because sometimes traps "go cold" and hardly win at all one day, then the next day they're winning all the time! I mainly stuck with the fave sequences.
It's worth having the market condition added to ensure that there are no unsettled bets before making a new one - that was one of the things DN mentioned after reviewing one of my strats - I think the rest of the things he mentioned were fairly standard and he didn't give away any juicy secrets lol.
One thing which is crucial is the field size - for AU dogs, DN always ran different strats for 4,5,6,7 and 8 dog races (again a market condition setting) - the same applies to horses and SQ can be used in the same way, though there's a much greater volume of dogs races each day, so it's easier to test strategies.
On testing - don't use Sim for anything other than making sure that the logic in your strategy is working correctly. Delayed odds in Sim will mean that you can end up picking up different faves to those you see in live when there's a lot of steaming/drifting of odds going on.
Have a close look in your data at times of day where the faves are most likely to win. I've noticed a lot of medium odds non-fave dogs winning in the early morning UK races in particular, so running a strat with sequence 1,1,2,1,2... could pick up a load of losers.
I've found AU dogs to run hot and cold, some days the fave wins all the time and other days they're never in the running - in my view, UK dogs are bit more reliable when it comes to faves winning - plus some races in AU have very poor liquidity. I always ignore NZ races because they're too few and far between for me.
One further thing is that timing can count for a lot. You could have 2 different people loading the same strat into live with a difference of 5 mins between them and one person has a bunch of winners hitting the sequence perfectly and the other has a bunch of losers because they missed the optimum start of the sequence.
The length of the sequence doesn't really matter, though I think DN may have mentioned that his are normally 10-20 numbers.
As for odds ranges, you could download the winners from the Results tab into Excel. As a simple view look at the integer odds value 1,2,3,4,5 etc. and put these into a pivot table against each fave.
I've included some of the strats I messed around with as a starting point - though you'll need to add unsettled markets and field size checks.
Attachments
(4.03 KiB) Downloaded 30 times
Horses MM_2_n_3.png (58.56 KiB) Viewed 15330 times
