General discussion about Bf Bot Manager v3 software for Betfair, Betdaq and Matchbook betting exchanges
By dn6789
yeah it is obvious we are running up against the limitations of the software itself to some degree when it comes to the UK at least

but for right now I am not losing sleep about not having UK GH and Horse solved yet.

Speaking from a pure AUST perspective

I can build a strategy for runner 1 runner 2 etc or Fav 1 Fav 2 etc --- and I can make it win - because I know how to filter things to get the desired result .. maybe this will be different in the UK and I cannot solve it but I will try.

I would love to see this software expand in the way it works and inputs it can accept plus provide a HUD on the screen with the stats like pokertracker .. but I doubt that will happen any time soon.

Another wishlist item would be able to link Amazon AI into the software and have it make the adjustments rather than me look at it each morning ... but I just get the feeling that this would be some sort of custom software.

At some point I would like to hire 2 people

1. to track stats and create moving averages of everything - to know when to scale up bets and when to scale down bets
2. full time programmer to take what we have now and build all these integrations I am talking about and build out sportsbetting properly etc

but the thing is I have some experience dealing with these sorts of things - and it is alot of responsibility to have full time staff especially when it comes to betting/money sorts of things.
By echelon
I've never used Amazon AI before.

Would you get it to read the BFBM Results and work out %'s and then tweak the strategies automatically to what the new %'s are?

Those jobs sound great, though if you have a decent database of results you possibly only need a programmer. The key thing is that you need to get someone you can trust and have a watertight non-disclosure agreement.

When I was watching the UK horse strategies earlier today the original was streets ahead of the other 3, though at the end of the day interestingly both SQ strategies came good!

We had an SR of 33% and a market SR of 44% on both. The individual bet SR was reduced because those SQ strategies were entering bets 1 minute before the race start, which meant we could have 2 or even 3 bets on the same race as the favourites moved position. I've now set the entry time to 15s which I hope will eliminate that problem.
Horse_Results_Thurs.JPG (49.25 KiB) Viewed 13257 times
By dn6789
amazon can read live data with a plugin - I do not have it set up that way - I downloaded and uploaded to my amazon server etc

but that would be my ideal setup - link everything together

remember I did say - you need to go into the settings and make sure everything is set to that 15 seconds for recording and betting to match everything up and to like you said get the bets right.

SQ is a life saver really -- you can randomize the results -- anyway my gfriend is yelling at me to take her to out somewhere so I gotta go back later.
By dn6789
just another comment on your results

this is exactly why you have to take these ideas and run them right -- because I mention something - you do not think its important -- but then when you run it you see why I mentioned it and then it makes sense

then you have learned something - and ready to keep moving forward -- and the next thing you learn sparks another idea and then you keep moving forward making something better and more robust

also in the other thread we were just discussing how with SQ - you get winning running losing running and mixed runs .. which is something else I mentioned -- you need your staking and winning strike rate to match closely to your take profit and stop loss -- so you can hit your take profit each day ... if you push for too much profit based on your staking - you are open to more randomness .... anyway the more you do this the more you will understand what I mean.

so for me if I hit a FAV 4 during the day that often will reach my target profit -- or I will grind my way there with mixed winners -- all depends how things work out.

Lastly try the various staking methods and see what works best with your capital -- I always intentionally avoided mentioning this as you know why -- but if your doing this everyday - you will find this will help as well.
By echelon
Hi DN,

As ever, thanks for your valuable advice.

I was still getting more than one horse picked up in the same race, so I now enter at 0 mins. This would be cutting it too fine for the dogs I suspect where races can start exactly on time. Horse races tend to start a bit late from what I've seen, or very late in some cases.

It seems as though the SQ strategies are quite consistent, so I'll give them a whirl in live soon with small stakes for starters. I think I need to run them in Sim over the weekend before I do that because UK dogs and horses appear to show marked differences between weekdays and the weekends.
Horse_Results_Fri.JPG (45.85 KiB) Viewed 13217 times
Last edited by echelon on Fri Aug 27, 2021 11:03 pm, edited 1 time in total.
By LousyDrunk
I appreciate the conversation you guys are having, it helps motivate me while giving decent direction.

I don't have much to contribute yet, still learning the ropes as I've only got about 5 free hours a week to throw at this. In 3-4 weeks though, Ill have significantly more time.

I have found decent success in locating the longer odds that typically occur 1-5 times a day, constantly betting on those ones and then using a stop once it hits a profit margin because if it keeps going, i'll usually end up at a small loss. During the last week 4/7 it hit the profit, 2/7 it'll broke evenish, (up or down a few $$) and 1/7 it lost money. (I put in a stop loss just incase but it didn't hit it).
By dn6789
I do not know about you guys - but I have put in hours on the PC this week and I am pretty burnt out and need a couple days off - so no posting from me till next week

on mine I have it set at 1 bet per fav position and bet 15 seconds before race - and I do not see it betting more than 1 dog per race -- not sure maybe I have something else that is stopping that

this morning I upgraded all my strategies to step betting because after looking at my results they are very consistent which comes down to the filters - and the slight tweaks I make each day.

not much more I can add other than to say - perhaps if your struggling to build the "perfect" SQ - then try it this way - build many of them and see what the longest win streaks are and longest losing streaks are - make some slight changes and then run with the most consistent one ... because given the right staking and a few filters thinking about it - no matter the SQ you use - you should be able to win.

on horses my SQ is really good for AUST - like I was saying the other day even in the perfect setup the Fav 1 does not win every time - so even a small SQ and staking can win in those situations - and this is what I have found.

anyway I have everything updated and will let it run with the new settings and make some adjustments as the data comes in - and by this I mean monitoring win vs loss streaks - updating the filters as needed and also modifying the staking pattern.

good luck today guys catch you in a few days.
By dn6789
couple things on UK SQ Horse

++ am now only using my custom one - reason just works better
++ changed staking to Step Staking

Here is the thing - when I started these threads I used fixed or bet by price -- reason being I wanted to keep risk low and show you guys you can win with the most basic forms in the software

then I mentioned step staking as something I was adopting. Next I played around with other staking options but now I only have step staking in my strategies.

The NUMBER ONE reason for this - is because I have worked so hard on the selection process my strike rate is high --- with the target being 40% ... with a strike rate like that you want to increase your bet as you move a long because you jag a nice winner after 3-4 losses the return is high.

Also I mentioned the other day my normal run of outs is 4-5 with a maximum of 11 or 12 over months of data -- so my step staking are all built with 15/20 steps and that is it -- after that is resets back to the first one - and you effectively lock in the loss and start over .. so calculate what works for you over 15 bets and have double the capital available for that each day... as really it should not happen that you ever ran 2 cycles without a winner.

so for the strategies developed for the forum that is how I have it.

(also have a take profit for each day - something that you can easily hit and then wait for the next day - the more your account grows the safer it becomes - and sometimes SQ can get stuck where it goes up and down - wins offset losses -- so it is better to achieve daily growth than get stuck)

Lastly something I mentioned before -- do not bother to buy strategies off the net - you are wasting your time - develop your own -- no-one in their right mind is going to sell you something that makes $200 a day for $50 - it is just not happening ..... so take some notes - try some things - fine tune everything and monitor the stats each day and if your deviating too much - make some changes .... as you learn more you will see what your supposed to be doing. Also take note of some of these system settings we have talked about -- does not seem important but when your placing a hundred bets a day -- it sure is.

UK is different than AUST/NZ -- make sure you realize that - as even for me so far my UK GH has not come anywhere close to my AUST GH results ... UK Horse SQ seems ok so far but even for me - I bet only very small on it at present - where AUST Horse SQ I am confident there -- my filters are good and results are solid

These are my thoughts - they are free - use them - discard them - or print them out and burn them -- totally up to you.
By dn6789
as you can see last night

6 bets 3 wins .. I am using small bets - just to show you guys that the type of strategy can work and for people that know UK racing better than me - you should be able to outperform me easily.
uksq.PNG (112.27 KiB) Viewed 13186 times
By dn6789
the main point - is that the filters are selecting the right conditions to bet and the Strike Rate stays around 40% over more bets - I constantly talk about that - but it is very important to winning

I have to be conservative as these races are run while I am sleeping so no chance to adjust anything other than during the off time if something goes wrong.

have to look in the software again - but I think there is a feature to only place bets if strategy has a certain strike rate - maybe I need to add that condition here
By echelon
Hi DN,

Your horse strategy appears to be surprisingly simple :shock:

From what I can see, the only filtering criteria is that the race has 6 runners or less and you have a SQ starting with 1,2 (though I've no idea what the rest is).

The weekend effect hit all my Horse Test 4 strats which ended up losing on both Saturday and Sunday. It's a bank holiday in the UK today and when watching the dogs, I've noticed that bank holidays tend to be like weekends. You can have reliable strategies which run great from Monday to Friday and these get seriously hammered on weekends.

This can be quite frustrating because looking at the results, there's no obvious bias on the weekends, the results just seem to be more random. For example Fave 4 trap 2 can be a consistent winner during the week and at the weekend Fave 4/ Trap 2 combos never win.

Surprisingly football is also much trickier to trade at weekends - you've covered a lot with horses and dogs, have you thought about creating any football strategies? There's plenty of liquidity in most games :D
By dn6789
it is not rocket science on those initial SQ - there is more under the hood as things progress but off the bat it is pretty vanilla

I am running it to show it can be done really - I doubt it ever comes close to AUST in terms of profits for me anyway.

No idea on football -- really want sportsline automated for US sports as I have enough to do manually already.

right now I am experimenting with my steps vs SQ -- more on this when I see the results -- so far so good.
By echelon
Good ol' Captain Haddock lol - I got that one too!
Horses SQ2 Mon.JPG
Horses SQ2 Mon.JPG (135.07 KiB) Viewed 13156 times
By dn6789
so easy 7/15 47% +73

everything you want to see basically
ukhorsesq.PNG (20.25 KiB) Viewed 13149 times
By dn6789
its no accident that most times when I post these screenshots - they are all close together with the Strike Rate

sometimes where I get a longer price winner you see 27% 30% 32% -- but most of the time each day it is 40% plus.

this is what drives the profit - no matter which staking you use - but obviously staking is a part of how much you make and what DD you see.
By dn6789
UK stuff lagging for me

the fact is that it works different than AUST and I have to figure out how to adapt better
ukh.PNG (19.81 KiB) Viewed 13108 times
By dn6789
it was a crap day yesterday car servicing - bought new couch - did not really get to see much

but AUST GH and Horse - with the new staking and allowed to just bet - did what they do

UK I stayed awake and watched some races - it was actually winning when I went to bed - but got hurt late in the day .. I think I will tinker with it today and return it to price betting until I see more consistency in the results

step betting is fine when your around 40% - but if you slip to 20-25% then your not going to win.
By dn6789
one change I made to all SQ today is this

I wrote out my SQ lists and went into the step staking part - and made a specific bet based on the number I was betting - and so it also meant my Bet 1 is higher as that is Fav1

some days when you catch those 18/1 on decent size bet that is your whole day but I am just trying to smooth it out .. something else once I finish getting it how I want is to turn these things into graphs so you can see what a day looks like with these things

anyway better get finished before racing starts
By dn6789
changing the stake to match the SQ bet range did not work

today my SQ lost 300 on GH and is up 253 on horse

this is not a normal day and it shows you that a mixed SQ with a rising stake via step is most effective

seems to me the reason is this --

on a normal day where favs win their fair share of races SQ bets these normally with rising stakes and grinds out a profit ... on days where we get variation because I have some longer priced fav in my SQ when those hit on a rising stake - it bails out the SQ losing streak fast .. this is what has given me a very stable SQ

anyway use the info or not up to you cheers

(I now changed my staking back to the way it was) ... the reason SQ horse is winning today is it hit a 20/1 with $16)
By dn6789
so here is GH and Horse since I made the change back

more normal and what we are used to seeing

it just shows you changes you make do make a difference and I went from a loss to a profit pretty quickly.

So to summarize use step staking like 5 6 7.25 8.90 9.95 ... just a gradual increase and let the SQ bail you out of any losing streak you run into - and by that I mean if you take a loss by winning on a FAV 1 6 or 7 into a SQ

anyway if your doing this you know what I mean

on UK Horse I made very slight changes see how it goes tonight.
change.PNG (34.99 KiB) Viewed 13080 times
By echelon
dn6789 wrote: Thu Sep 02, 2021 8:57 am on a normal day where favs win their fair share of races SQ bets these normally with rising stakes and grinds out a profit ... on days where we get variation because I have some longer priced fav in my SQ when those hit on a rising stake - it bails out the SQ losing streak fast .. this is what has given me a very stable SQ
Hi DN,

That's an interesting point - so...if we have step staking of 1,2,3,4,5,6, does that mean that we should have faves 1 and 2 only in the first 6 numbers of the sequence?

I'm thinking that on a normal day when faves are more likely to win, we stand a good chance of getting in on a 1,1,2,1,2,2 type run. As you say though, on a non-normal day when some of the long shots win - perhaps front loading the SQ with faves too much like this means that we have a longer losing run with no chance of hitting longer shots until step 7 and further on in the SQ.

I've not been looking a sequences too much recently as my first focus has been filtering down the bets before applying the sequence. Your earlier hint on looking at shorter term stats. and adjusting them regularly (e.g. daily) if required is helping a lot - trying to work with monthly or YTD stats. is tricky and not something which I'm doing any more.
By dn6789
there is no right answer for the correct SQ - without figuring out how many races a day the SQ is going to bet and what is your target area

lets take AUST GH - there are like 100-140 races a day

now if you bet 70-80 then you need a balanced SQ -- if you bet 20 - then you need an SQ aimed at that target 20 races and your going to need to have your staking right

I have tried a lot of things to alter how SQ bets and when it bets and how to stake it best .. step makes more money because the filtering controls the losing streaks and like I said when you hit a Fav 4 or Fav 5 - any losing streaks is quickly reversed no matter what your actual SQ is.

If I go to the results tab and look how AUST GH ended last night it was this

6 4 1 2 4 2 1 2 2 1 2 1 5 2 with no filtering

if I look earlier it was 1 5 2 5 4 3 2 3 2 2 2

so that tells you really the first example - you need a grind strategy - where in the second period Fav 1 was just a drag on profits and something more balanced could take advantage of Fav 3/4/5 as often those pay quite well - but again we do not know when you apply some filters what would come out exactly - those are assuming your betting every race.

if your betting a lot of races - then perhaps you would bias your first 6 to higher Favs - and then say OK past that I need more balance in the SQ

my personal SQ length is 15 before it resets on end or on win
By dn6789
also look in the support forum for what I am asking them to build into the software

lets take a longer term period of 300 races -- but have a shorter period like 40 races

if the SR of a runner over 40 is above that of 300 - then either allow the bet or bet more ... if the short term strike rate is bad then not allow the bet or decrease the stake

this would make a lot of sense for us right --- we filter the race - we SQ the race -- but we only bet when our runners are performing well .... this sort of extra filter could be using many ways
By dn6789
as some days tracks favor the inside or outside for one reason or another -- our SQ is using a more normal distribution based on our filtering - but conditions may vary

so this would be another way or protecting capital or making more money - if the races are biasing our selections

reduce our losses and increase our wins.