Annoyingly forgot to save the data where I took out NZL/AUS dogs from the win sample summary below (they have different commission rate). However it had similar negative ROI. Here's the results including 2% commission and including NZL and AUS dogs.
Below is using £1 back stake, or £1 liability lay stake. Remember that when laying at BSP you would actually have had to use £10 liability as a minimum!
Back bets (win markets)
1st Favourite: Profit/Loss = £-7882.80, Races Included = 950468
2nd Favourite: Profit/Loss = £-24294.59, Races Included = 946077
3rd Favourite: Profit/Loss = £-20425.02, Races Included = 947482
4th Favourite: Profit/Loss = £-12279.41, Races Included = 943301
5th Favourite: Profit/Loss = £-9146.38, Races Included = 934853
6th Favourite: Profit/Loss = £-15818.07, Races Included = 872546
Lay Bets (win markets):
1th Favourite: Profit/Loss = £-139557.52, Races Included = 950468
2th Favourite: Profit/Loss = £-49825.08, Races Included = 946077
3th Favourite: Profit/Loss = £-25298.94, Races Included = 947482
4th Favourite: Profit/Loss = £-14648.78, Races Included = 943301
5th Favourite: Profit/Loss = £-8454.90, Races Included = 934853
6th Favourite: Profit/Loss = £-4286.84, Races Included = 872546
I have also attached a spreadsheet for win and place markets which had a further breakdown into year and odds for win AND place markets including commission. This spreadsheet I did remember to remove AUS/NZL. At odds 30+ it shows a profit in place markets. That is because there have been some historic crashes and the BSP got recorded into the system incorrectly at 1000-1 . These are incorrect. You could not get the odds of 1000-1. I did sum everything together and as per win markets, there are no profitable strategies
Data used for backtest:
https://promo.betfair.com/betfairsp/prices
.
Full backtest of every Greyhound at BSP EVER!
Glad someone found the thread!
To get a balanced view on this data you need to equalise the staking.
Laying to a liability will get you that on the lay side, but backing to a fixed stake will not achieve that on the back side.
You should back to a target profit to achieve that.
Laying to a liability will get you that on the lay side, but backing to a fixed stake will not achieve that on the back side.
You should back to a target profit to achieve that.
This does remind me. Theres quite a few strategies that are profitable when layed at bsp to a fixed stake. Unfortunately cant lay bsp at fixed stake
-
ShaunWhite
- Posts: 10353
- Joined: Sat Sep 03, 2016 3:42 am
You can if you keep cancelling/replacing your sp bet as the price moves (bsp will be approx ltp). But no staking method will turn a losing strategy into a winning one. And the only one that might enhance it is Kelly. That was devised in 1956 and believe me, people have looked for better
And you'll move bsp as soon as you bet at it.
Must have been a calculation error from gpt, This was around a year ago, think it was laying at BSP in certain odds ranges to my memory. GPT thought it was profitable in certain odds ranges at fixed lay stake but not fixed liability.ShaunWhite wrote: ↑Mon Feb 17, 2025 6:07 pmYou can if you keep cancelling/replacing your sp bet as the price moves (bsp will be approx ltp). But no staking method will turn a losing strategy into a winning one. And the only one that might enhance it is Kelly. That was devised in 1956 and believe me, people have looked for better
And you'll move bsp as soon as you bet at it.
-
ShaunWhite
- Posts: 10353
- Joined: Sat Sep 03, 2016 3:42 am
Unusual streaks are the danger with fixed stake at longer odds. Getting to a theoretical profit can be a long rough journey. %ev is the best measure, and the p-value.
Yep, they were at longer odds ..ShaunWhite wrote: ↑Mon Feb 17, 2025 10:30 pmUnusual streaks are the danger with fixed stake at longer odds. Getting to a theoretical profit can be a long rough journey. %ev is the best measure, and the p-value.
Youve seen this pattern before
