The BTC Horse Racing Thread
-
@jonathan-jones Yes, that's still the main update to the horse racing that we are currently working on.
-
@Adam
I remeber a while ago there was talk about reworking how some of the criteria works so they are evaluated at the time a bet was placed. So for example the Favourite condition would be evaluated based on what was favourite at the time the bet was placed rather than just at BSP. Did anything progress with that? Is anything going to progress with that? Some of the presets are using that condition ten mins out so in reality you could end up seeing the order swap several times. -
@karl-pick
i am usin v3 and didnt get any yesterday or today
-
Just a quickie. I have had no qualifiers for Back the Favourite for 2 days, is this the same for everyone else or is my system playing up?
-
First thing that jumps out is horse weight rank.
Dont think the software knows that untill it gets the race results the next day. -
Something went well wrong yesterday in my test of all the strats (but not pace backs). The software says i should of been up 1 point, the reality is down almost 7 points. The only strat that actually matched up was back the fav since that had no trades.
Back C&D: software says - 2, actual was -3
Lay Beaten Fav: software says + 1, actual was -1
Low Lays: software says +2, actual was +0.5
Pace Lays: software says no trades, actual was -3.5 with 2 trades.Unfortunatly i dont have the selections sheet from yesterday to confirm it all went well so will make sure i keep todays. On that i am double checking the automation rules that my little app has created for fairbot against the selections for today. Half way through so far and they are 100% correct
EDIT~
All differences sorted.
I had changed my filter in the software for the low lays to not have the number of runners condition. My automation rules though still had it. That does account for that difference so thats sorted now.
The Lay beaten favourite my automation rule was totally wrong. It didnt have the not 1st/2nd fav rule. Backtesting without that rule matches my results so thats sorted now.
The Pace Lays software backtest was reporting no trades. Its now reporting trades and matches my results. Not sure why that one was late updating but thats sorted now.
Back C&D Fav is still one point adrift but there is a problem with these ones that specifiy favourites. They do swap places (even in the last 10 seconds) so that is going to be something to watch. On the counter of that i did gain a point monday because of that little oddity so all square. -
0_1676024794465_rules_export (1).json not sure if that will work but thats all the conditions
-
@paul-scott THere are a load of conditions that sont really work well for qualifiers. Can you post a screnshot of your criteria?
-
Hi guys, ive created a new strategy on the horse racing software but i have a problem with it. everytime i go on upcoming qualifiers it shows as none but when you click on individual trades it shows that there was some selections? has anyone else came across this? ive tried running it at various times during the day but always shows as none upcoming but there always trades?
-
@philip-brogan said in The BTC Horse Racing Thread:
0_1675875336127_rules_export(4).json
Hi Martin,
I have made another revision to the Back the Favourite Strategy. Can you tell me what you think?
Many thanks
Philip Brogan
great roi! less selections is the trade off, quality over quantity strategy
-
Thanks for your comments David and Jonathan.
It was just a thought. If Back the Favourite looks back fitted now I am sorry
Many thanks
Philip Brogan
-
@david-milligan
This one "Back The Pace Horse - At Odds 2-25 " in the presets. It was created on the 6th October. Prior to its creation every month was positive, since its creation its only had 2 positive months and still hasnt reached its pre creation peak. Its entirly possible its just a blip but without a compelling reason for the staggering performance difference that chart screams that its back fit. -
@jonathan-jones the Flat Pace Backs has done OK on forward testing though, do you mean a different strategy?
-
@philip-brogan
Not overly keen on pushing these things too much in terms of optimization. Its a better than fair chance it will end up too back fitted. Look at the Pace Backs to see the problem. That strategy has bombed since its creation date. The difference in performance after its creation to before is so stark there is no possible way thats just variance. -
@philip-brogan said in The BTC Horse Racing Thread:
0_1675875336127_rules_export(4).json
Hi Philip, your adjustment seems to have improved ROI a fair bit but the overall profitability looks to have gone down due to the lower number of selections! Interested to see what the more knowledgeable guys think but I'm going to stick to 3.0.
-
I'm in the process of optimizing my Bet Angel bots and have a quick question.
Just to be certain... The above means, that starting from 10 minutes before official start time and right up until official start time, that if a selection falls within the odds range it will be backed/layed. Is this correct?
-
0_1675875336127_rules_export(4).json
Hi Martin,
I have made another revision to the Back the Favourite Strategy. Can you tell me what you think?
Many thanks
Philip Brogan
-
@dan-mackinnon said in The BTC Horse Racing Thread:
@john-folan haha! I’m forever reminding myself to be patient.
Whenever I check my results against the software and find one that’s different I turn into the dejected ex “was it me? Did I do something wrong? I can change!”
This made me laugh
so relatable -
@jonathan-jones keep us updated!
-
I am experimenting trying ALL the strategies at once. Obviously its automated (god knows how you would manage to do it otherwise). Even automated it would normally be difficult enough on fairbot as you can only allocate a single strategy file to each race. Fortunatly i can code so hacked up a little tool that reads the qualifiers.xls and creates a rule for each of the selected strategies and horses then writes a fairbot strategy file for each race with all the individual rules. Its turning out damn difficult to track results though.
Only in sim mode this week to see how well it conforms to what 'should' happen. Then i will throw 50 quid at it and let it play @ 1%. Analysis of the results of all the strategies amalgamated suggests it shouldnt blow the account.
A couple of tips if you are using fairbot automation:
Use streaming API. I noticed a while back that no matter what you set the APING settings to you would miss a couple of entries a week entirely.
trigger odds in conditions you need to use LTP (as thats what this software is using).
1333/6373