Here's what happened on the 30th September 2017
| Match | Prediction | Result | Return |
|---|---|---|---|
| Cardiff v Derby | Cardiff | 0-0 | £0.00 |
| Hull v Birmingham | Hull | 6-1 | £2.00 |
| Preston v Sunderland | Preston | 2-2 | £0.00 |
| Dijon v Strasbourg | Dijon | 1-1 | £0.00 |
Winnings: £2.00
Closing Balance: £5.65
Here are my predictions for the 1st October 2017
| Match | Prediction | Prob% | My Odds | Betfair Odds |
|---|---|---|---|---|
| Everton v Burnley | Everton | 73.89 | 1.35 | 1.7 |
(Lots of very short odds on Betfair, so only one today)
Opening balance: £5.65
Bets: 1 * £1 single
Remaining balance: £4.65
Predicted with some ugly Python code.
I've posted articles explaining how it was done - Part1, Part2, Part3, Part4, Part5.
Code available here.
Also check out my posts on Coding a Poloniex Trading Bot. Here & here
Aaarrggghhhhh!!!!
Hey, how's it going? I've been thinking (uhoh) since all the failed predictions from your programm are draws, what do you think about hedging that result with a smaller stake, so you end up not losing anything significative.
Or including in the program something known as home advantage. Example:
Season home goals average 1,5- season away goals average 1,4 = 0,1
0,1 /2 = 0,05
Inclunding home advantage
Home Team goal expectancy + 0,05
Away Team goal expectancy + 0,05
This is one fo the ways I saw its calculated, not sure if it will make any diference lol (or maybe just add 0,5 to home teams goal expectancy )
When it fails, it does seem to be more often a draw than a loss. The home advantage should already be considered due the the calculations for attack and defence strength but it's certainly worth looking into.
Betfair, and I believe, most other bookies offer the option of betting on home+draw or away+draw, obviously with a reduction in the odds: there's likely still value bets to be found betting that way.
Another thing to try would be calculating the expected goal difference and including that in the consideration if a bet would be worthwhile.
I've almost worked out how to get odds for each scoreline too and will try and add that in soon, if I can.
I'm mostly concentrating on the Crypto trading code at the moment, I've never worked with GUIs before now, apart from simple VB stuff, and I'm finding that quite a trial. As soon as I get some progress there (or get so frustrated with it, I need to give up for a while) I'll get back the the footy.
I will be keeping my predictions going for now, at least until the pot runs dry. It might just be an unlucky spell and it'll be all winners from now on....
Later. :-)
For sure, can't win them all. :)
Btw why did you choose Polioniex for the trading bot? Charts are so poor, only 3 or 4 studies.
Pretty much by accident I think. IIRC it was the first site I found and tried out and I've mostly stuck with it since. Only problem I've every had with them was not being able to withdraw STEEM for so long.