Trading What I see !?

Learn sports betting strategies and discuss key factors to consider when placing a bet.
Post Reply
User avatar
ShaunWhite
Posts: 9731
Joined: Sat Sep 03, 2016 3:42 am

goat68 wrote:
Wed Nov 24, 2021 4:44 pm
Paper testing lets me do that with reasonable accuracy
No it doesn't.

Here's a brief thread with comments from Euler, LinusP, Dallas and me all basically saying just that viewtopic.php?f=37&t=19982&p=205043&hilit=
User avatar
goat68
Posts: 2019
Joined: Tue Jun 30, 2020 3:53 pm
Location: Hampshire, UK

ShaunWhite wrote:
Thu Nov 25, 2021 2:36 am
goat68 wrote:
Wed Nov 24, 2021 4:44 pm
Paper testing lets me do that with reasonable accuracy
No it doesn't.

Here's a brief thread with comments from Euler, LinusP, Dallas and me all basically saying just that viewtopic.php?f=37&t=19982&p=205043&hilit=
I stand by what I said, and already understand what is in that thread...and is within my expression of reasonable...
Backtesting v paper trading is the same emulation, but what is different is the psychological approach to strategy development.
Bactesting is all theory
Paper i can watch the live dog race, evaluate prices, watch dogs run, compare market view vs my model, and examine a single race in detail
User avatar
goat68
Posts: 2019
Joined: Tue Jun 30, 2020 3:53 pm
Location: Hampshire, UK

The other psychological affect of paper trading, is it slows you down... which probably sounds negative to you... but it makes you think, rather than zooming down avenues of data mining...

My approach will probably actually be in reverse, develop the strategy paper trading, then backtest once I have the framework to do so. I need to setup a retrospective "web scrape", ie.get dog form data as if it was scraped on Wed 24th Sept 14:36pm...!
Cardano
Posts: 215
Joined: Mon Aug 21, 2017 12:02 pm

goat68 wrote:
Thu Nov 25, 2021 7:48 am
ShaunWhite wrote:
Thu Nov 25, 2021 2:36 am
goat68 wrote:
Wed Nov 24, 2021 4:44 pm
Paper testing lets me do that with reasonable accuracy
No it doesn't.

Here's a brief thread with comments from Euler, LinusP, Dallas and me all basically saying just that viewtopic.php?f=37&t=19982&p=205043&hilit=
I stand by what I said, and already understand what is in that thread...and is within my expression of reasonable...
Backtesting v paper trading is the same emulation, but what is different is the psychological approach to strategy development.
Bactesting is all theory
Paper i can watch the live dog race, evaluate prices, watch dogs run, compare market view vs my model, and examine a single race in detail
I completely agree with you Goat regarding the psychological approach to strategy development - back testing is all theory
Reading last years newspapers isn't going to tell you what the news will be tomorrow.
Trader Pat
Posts: 4327
Joined: Tue Oct 25, 2016 12:50 pm

Cardano wrote:
Wed Nov 24, 2021 10:04 pm
Trader Pat wrote:
Wed Nov 24, 2021 9:48 pm
Straight backing or laying is only gambling if you don't know what you're doing. There are plenty of people who reduce their commission by straight backing and laying.

Just like back testing is only pointless if you don't know what you're looking for. If you spot something when trading live you can then back test the data you've collected to see if it repeats. However collecting data and then looking for patterns in it is probably not so worthwhile.
"There are plenty of people who reduce their commission by straight backing and laying."
Can you explain that comment a bit more
Sorry that was a mistype... I meant increase commission to help reduce PC
User avatar
goat68
Posts: 2019
Joined: Tue Jun 30, 2020 3:53 pm
Location: Hampshire, UK

So i've done as Shaun suggested and run a BSP efficiency test on UK greyhounds for August markets:
4400 markets, £1 on every dog, BSP adjusted to 100% book, net -£185 for the month, efficient or not ??:
Screenshot 2021-11-25 at 21.40.05.png
I'm going to run it on 5 months of data now...
You do not have the required permissions to view the files attached to this post.
User avatar
goat68
Posts: 2019
Joined: Tue Jun 30, 2020 3:53 pm
Location: Hampshire, UK

Hmm, just trying to figure, why did I adjust for book% ? can't get my head around it?
User avatar
goat68
Posts: 2019
Joined: Tue Jun 30, 2020 3:53 pm
Location: Hampshire, UK

I think I might have adjusted BSP the wrong way
User avatar
goat68
Posts: 2019
Joined: Tue Jun 30, 2020 3:53 pm
Location: Hampshire, UK

so yes I did, of course bookie takes a cut of true price, so need to improve BSP by x(book%/100.0)
User avatar
goat68
Posts: 2019
Joined: Tue Jun 30, 2020 3:53 pm
Location: Hampshire, UK

That's a bit better:
Screenshot 2021-11-25 at 22.07.46.png
You do not have the required permissions to view the files attached to this post.
User avatar
ShaunWhite
Posts: 9731
Joined: Sat Sep 03, 2016 3:42 am

goat68 wrote:
Thu Nov 25, 2021 9:44 pm
efficient or not ??:
Depends on whether it's statistically significant because it's never going to be flat. And of course it's of no use going forward and just gives context to previous results. Even if it is statistically significant it won't persist because so many people see it.

The chart actually tells many stories that need to be thought about and investigated. A downward slope could just as easily mean there's simply been a run of results with winners averaging under 6

But it's akin to how a swing trader would look at the market tendency to have steamed or drifted. If it's been steamy, and they have a tendency to back first then they've had a tail wind and probably flattering results. Interpreting this sort of thing is where art meets the science and you only develop the 'feel' for the implications after a few years.

I just thought it was something you might want to add to your 'state of the markets' reporting to flesh out the bigger picture rather than it being something you had the experience to make much use of now.
User avatar
goat68
Posts: 2019
Joined: Tue Jun 30, 2020 3:53 pm
Location: Hampshire, UK

ShaunWhite wrote:
Thu Nov 25, 2021 10:57 pm
goat68 wrote:
Thu Nov 25, 2021 9:44 pm
efficient or not ??:
Depends on whether it's statistically significant because it's never going to be flat. And of course it's of no use going forward and just gives context to previous results. Even if it is statistically significant it won't persist because so many people see it.

The chart actually tells many stories that need to be thought about and investigated. A downward slope could just as easily mean there's simply been a run of results with winners averaging under 6

But it's akin to how a swing trader would look at the market tendency to have steamed or drifted. If it's been steamy, and they have a tendency to back first then they've had a tail wind and probably flattering results. Interpreting this sort of thing is where art meets the science and you only develop the 'feel' for the implications after a few years.

I just thought it was something you might want to add to your 'state of the markets' reporting to flesh out the bigger picture rather than it being something you had the experience to make much use of now.
yep, interesting

The key question I can't answer, is, is the saying "Beating BSP" only relevant if you're betting on every market/selection...?
What's to say your skill of selection means you're selecting the selections on the positive slopes above...?
User avatar
goat68
Posts: 2019
Joined: Tue Jun 30, 2020 3:53 pm
Location: Hampshire, UK

5 months, very interesting!
Screenshot 2021-11-25 at 23.43.55.png
You do not have the required permissions to view the files attached to this post.
User avatar
ShaunWhite
Posts: 9731
Joined: Sat Sep 03, 2016 3:42 am

goat68 wrote:
Thu Nov 25, 2021 11:16 pm
The key question I can't answer, is, is the saying "Beating BSP" only relevant if you're betting on every market/selection...?
It comes down to statistical significance. Beat sp on one runner and it means nothing. Beat it on them all and it means everything, beat it on some of them and it means something.

I realise that's hopelessly vague but let's just say that in the absence of any other comparison then it's the best worse choice. It's certainly better in the short/med term than looking at your pl. You might not know that a couple of years ago there was a run of 26 (I think) losing favorites. If I'd backed them all at 5 ticks above sp then despite them all losing I'd gladly do the same again.

Value and sp and all that aside, if we were manually trading, and closing for a profit at the start (ie 'beating' sp with a prior bet) then we'd be happy doing that, and proving that beating sp makes money even if you do it on 7 horses a day instead of 700. Maybe that's worth thinking about?
User avatar
goat68
Posts: 2019
Joined: Tue Jun 30, 2020 3:53 pm
Location: Hampshire, UK

So updated model strategy still positive, and ahead of previous model running in parallel:

New model:
Week 1 Back : 192.40
Week 1 Lay : -158.40
Week 1 TOTAL : 34.00 for 979 bets

Previous model:
Week 1 Back : -193.12
Week 1 Lay : 163.87
Week 1 TOTAL : -29.25 for 1148 bets
Post Reply

Return to “Betfair trading strategies”