The BTC Horse Racing Thread

Adam

Maybe I'm being a bit harsh actually, it's probably fine for pre-match / pre-race stuff but you're definitely at a disadvantage using it for in-play stuff.

Adam

I wouldn't recommend anything still based on API-NG personally.

The streaming API is like having an inside contact at Betfair shouting the latest prices at you every single time they change, without you asking.

API-NG is like having to phone them to ask what the latest odds are, and waiting for a reply :face_with_tears_of_joy:

? Offline

@martin

Not sure. I only had a short time with bet angel a long time ago. I have used Geeks toy since (but no automation at all in that) and fairbot. Fairbot was working fine withthe older api but i think thats just too slow for automation. I had it set to 0.2 seconds updates (as fast as it goes) but it just want any use. Besides missing about a third of its stop losses (iand i do think thats very much made worse by its speed) the bots dont seem to work as you expect. I think that they probably are following the rules but becauee of its update speed it misses triggers etc so looks like its not doing as it should.

With the streaming api enabled though its noticably faster. The response on the bots is instant. It missed one sl yesterday and always looked like it was executing the rules perfectly. When its on the trading screen there are no issues. Its just the market watch window really doesnt seem to work well at all with the new api and when you use it to switch to a closed race (to check bets placed etc) the race is often mangled. Like its stopped half way (but it obviously executed corrrectly when it ran). I am going to keep going with the streaming api because i dont really see any options. Its just not really usable for anything other than basic bet placement in automation otherwise.

Ian Parry

@john-folan said in The BTC Horse Racing Thread:

@darryll I’m on the Ouzo so no problem there.

Wasn’t any flat pace today. I never looked at anything else. I’m just trolling people with pictures of the pool and my drinks. Think I’m not going to bother for the rest of the holiday and just let the short odds lays run on the old vps.

So queue a really crap 10 days for them and several missed winners on the flat pace and other stuff .

Right. I'm getting on the flat pace for the next ten days then retiring with my fortunes.

Keith Driscoll

@martin said in The BTC Horse Racing Thread:

@jonathan-jones that doesn’t sound great, what do you think is the most reliable bot service?

BF Bot Manager is great for basic strategies, but for speed, flexibility, and reliability, it is Bet Angel, especially with its 20ms updates.

John Folan

@darryll I’m on the Ouzo so no problem there.

Wasn’t any flat pace today. I never looked at anything else. I’m just trolling people with pictures of the pool and my drinks. Think I’m not going to bother for the rest of the holiday and just let the short odds lays run on the old vps.

So queue a really crap 10 days for them and several missed winners on the flat pace and other stuff .

Martin

@jonathan-jones that doesn’t sound great, what do you think is the most reliable bot service?

? Offline

@darryll said in The BTC Horse Racing Thread:

@jonathan-jones I'm interested to hear your findings so please do report back.

Used the streaming api in fairbot yesterday. Its significantly faster. Only had one missed stoploss. I had increased the distance between trigger and placement but these were being hit consistently at the better odds so i guess thats a win.

Its not all good news though. Fairbot is buggy as hell using it. The trades went well but other stuff is totally broken. With the old api as soon as a race is 'closed' the watch window updates. With the streaming api ists hours or sometimes never. You can navigatge to a closed race from the wndow and it will update but as soon as you look at another race its gone again. Closed races dont always say the winners or ltp either. Its a proper mess

Darryll

@john-folan Have fun!

? Offline

@john-folan Have a great holiday!

Ryan

@john-folan YES enjoy it!!

Angus Macleod

@John-Folan Enjoy mate, well earned after all this effort

Martin

@john-folan have a good one mate well deserved!

? Offline

@john-folan enjoy your holiday mate

John Folan

Right today's. I'm off to Greece later, if I can be arsed I might do the selections daily while sitting by the pool drinking a Strawberry Daquri. But don't hold your breath . The filters are all on the megathread anyway

Classic Trobs: Dob, Trob or Win if odds are between 3.5-9
NONE

Flat Pace. Back to Win or Place if Win Price is between 2.0-25.0. 4-14 Runners only. Avoid Soft, Heavy and Firm Going.
0_1658396404049_FLat Pace.JPG

Slow Horses. Dob any over 6.0. Multiple selections per race are what makes this profitable. Just think Pokemon.
0_1658396424094_Slow Horses.JPG

I'm also watching these to see how they perform. They are a variation on my Classic Trobs with less filters.
Trob or Win(Not worth Dobbing) any between 2.0-20.0 Near the Off
0_1658396443304_New Trobs.JPG

? Offline

@paul-foster said in The BTC Horse Racing Thread:

@jonathan-jones just bear in mind that sim modes will tend to match more bets, at better odds, than live mode every seams to.

I tend to use live mode, but with really small stakes rather than sim modes.

This isnt matching at all (and thats in sim mode) so god knows what it would be like with real money. It will give me an easy comparison of the api though.

Nick Allan

@paul-foster same

Nick Allan

@jonathan-jones when I've done L2B in the past l've never used auto its too risky you have to manually trade using geeks toy

Paul Foster

@jonathan-jones just bear in mind that sim modes will tend to match more bets, at better odds, than live mode every seams to.

I tend to use live mode, but with really small stakes rather than sim modes.

? Offline

@darryll said in The BTC Horse Racing Thread:

@jonathan-jones I'm interested to hear your findings so please do report back.

Sure, I have a strat that i am testing on sim mode at the moment. It uses a stoploss that would normally activate inrunning. I give it 10 ticks between trigger and placement (and these are typically odds between 2 and 5 so plenty of liquidity) and its missed at least half a dozen today. Probably more, will know later but the fail rate is so high its going to be pretty easy to tell if the api change makes any difference. If it doesnt help i will bin the strat. Its simple as anything, performs great in the backtester but falls over all the time when i try live because of the stupidly high fail rate on the stop loss.