I've created a bot that trades in running horse racing only, I've progressed this with forward testing past the practice phase and now onto a live account.
It uses book% as staking and looks to lay (x) horses during a race. On average it will lay anywhere between 2-5 horses. Currently I'm using book% of £5 and I'm looking to start scaling up. I'll probably end up implementing the offset batches rule but just wondering if anyone could give me a general indication of what I'm likely to achieve as a 'maxed out' position.
Naturally I'll end up discovering this for myself over the coming weeks but I'd just like to know if there's a general consesus of a fixed amount? Ideally to make this worthwhile pursing this bot I'd need a book% of £500 to matched across the 4 or so runners, is that beyond the realms of possiblity in running? I hasten to add it only trades over the longer distances, Chase, Hurdles longer than about 2 miles and usually only in the first couple of minutes of the off.
Results have been very promising over the past month but scaling up I understand could render this useless.
**average odds range between 4.0-10.0
Scaling Up - In Running
-
ShaunWhite
- Posts: 10472
- Joined: Sat Sep 03, 2016 3:42 am
Tough question. The better the value your offer represents the more you'll get matched, eg laying at 4 when it's crossing the line vs laying at 4 when something is struggling already. I think the best answer you're likely to get is to keep gradually increasing your stake until your returns start to decline but generally the size of the in-play market would be in proportion to the size of the pre-race market if you wanted to vary it race by race.
-
ShaunWhite
- Posts: 10472
- Joined: Sat Sep 03, 2016 3:42 am
It’s all about your match rate, I record the following:
runner_win_lose | stake | profit | order_match_rate (orders matched / order count) | size_match_rate (size settled / size) | roi (profit / size settled)
Assuming you start with £2 this would give you two rows of data (after 100/200 races) you can then increase the stake and see the impact on your match rate. You will notice that there will come a point where the match rate drops to a point that either profit turns negative or a slightly lower stake is more profitable.
You can then look at trying different execution strategies, a little tip is that smaller stakes get matched quicker and at a better rate. For example it is better to place two £2 bets with a small delay between the two than a single £4 bet.
runner_win_lose | stake | profit | order_match_rate (orders matched / order count) | size_match_rate (size settled / size) | roi (profit / size settled)
Assuming you start with £2 this would give you two rows of data (after 100/200 races) you can then increase the stake and see the impact on your match rate. You will notice that there will come a point where the match rate drops to a point that either profit turns negative or a slightly lower stake is more profitable.
You can then look at trying different execution strategies, a little tip is that smaller stakes get matched quicker and at a better rate. For example it is better to place two £2 bets with a small delay between the two than a single £4 bet.
-
MobiusGrey
- Posts: 294
- Joined: Fri Nov 23, 2018 8:10 pm
Thanks Shaun and Linus, I'd read somewhere that 20-30% of the total turnover of a race will be done in running it just seems when watching it that there's high volatility but no where the kind of liquidity seen pre race. I guess it may be that the prices don't hold for as long and so the money is thinned out across more prices than accumulating around 2 or 3 different odds. I just need to grab a relatively small piece of that pie.
Interesting point about the stages, I already employ something similar in that the bot will fire twice, 5 seconds apart, I had forgotten to mention that part.
I'll certainly starting recording the match rate in more detail, I'm due to double stake sizes next Monday so I think things will become more realistic.
Interesting point about the stages, I already employ something similar in that the bot will fire twice, 5 seconds apart, I had forgotten to mention that part.
I'll certainly starting recording the match rate in more detail, I'm due to double stake sizes next Monday so I think things will become more realistic.