The BTC Horse Racing Thread
-
@Martin Hey,
Has the latest one (.json file) been updated on the OP of the horse racing megathread page?
-
@Correa my ltbf has done ok, are you using the latest one? mind you I missed Sunday was that the bad day?
-
@Martin best one so far running on BFBM. I can't say the same about laying the beaten favourite though. Not great in the past few days.
-
-
Shared banks are fine for win betting (2-3 strategies), but I never use them for laying. You can put a huge dent in a bank in a short time laying.
-
@Martin said in The BTC Horse Racing Thread:
@Joseph-Henderson ah right then should be same one I would think this is the bot I use Back to lay dobs.gz
I’ll check that out against the one I’m using - cheers
-
@Joseph-Henderson ah right then should be same one I would think this is the bot I use Back to lay dobs.gz
-
@Martin said in The BTC Horse Racing Thread:
@Daniel-Mills is this your own strategy?
My back to lay dobs are -0.52pts and irish dobs are 9.1pts up since last Saturday when I started forward testing.
it's the back to lay 2.0 you posted, I set up the bot myself though. Looks like the only difference with the bot was the hedge which i've got rid of and will keep an eye on for a bit longer.
Cheers to Jonathan-jones and Joseph Henderson too for the input. Will have a proper nose through - missed the original response to David's query about the DOB's.
-
@Martin That is running it as a DOB strategy and without the stop loss, sorry should have made that clear.
-
@Joseph-Henderson I turned it into a dob strategy works much better, although I then turned that into a Irish dob strategy which is doing really well so far, only been a week though so not enough time to truly judge it.
-
@Jonathan-Jones Glad you flagged this issue!
Back to Lay 2.0 is showing a profit of 2.94 for August (1st to 6th inclusive) compared with -6.46 running my bot. One to probably do some more detailed back testing on using the method you outline.
-
@Daniel-Mills said in The BTC Horse Racing Thread:
Noticed some discrepancies in my dobs results this week. ~4pts down instead of 4pts up.
Road testing without the safety hedge I set up in BFBM to see if that puts it in the right direction this week. Software doesn't seem to think it makes too much difference, need to see how that translates live.
Anyone using the back to lay 2.0 checked theirs yet?
I am running the Back to Lay 2.0 on BFBM but haven’t compared to the back test yet. Based on Jonathan’s post sounds like I might need to readjust the settings/back test a bit to compare more accurately!
-
@Daniel-Mills
You will have. This strategy cannot be accurately modelled in this software. With the stop loss its going to be worse but even without it you wont ever get a perfect match.
Without the stop loss you can use the method i described below to see if the strategy is actually likely to even be profitable (because this backtest software alone cant tell you that for certain) in the real world. With the stop loss then its no better than flipping a coin.
-
Noticed some discrepancies in my dobs results this week. ~4pts down instead of 4pts up.
Road testing without the safety hedge I set up in BFBM to see if that puts it in the right direction this week. Software doesn't seem to think it makes too much difference, need to see how that translates live.
Anyone using the back to lay 2.0 checked theirs yet?
-
is this lapsed entry or exit?
This will be an issue with any DOB or trading type strategies that use an exit. The software uses LTP for its entry. In reality LTP might not still be available.
The way to handle un-matched entries is to request a few ticks worse than best price. That way you are pretty sure to get matched. Problem is if your entry isnt what the backtest thinks it should be then you cant be sure your exit even got hit let alone fully matched. So, you will end up with un-matched exits and the software tells you that you won that bet.
You cant model this scenario but you can get an idea of what the worst case for your strategy might be. You can add the IR Low field (and of course the enrty/exit price) to the individual trades output. Then in excel you can easily remove any winners that dont go more than x ticks lower than the exit. Its not a perfect solution but it will tell you if your strategy is actually profitable assuming worse case of bets not matching.
Note:
Wont help if using big stakes and possible low liquidity.
