Trading What I see !?

The sport of kings.
Post Reply
Anbell
Posts: 1209
Joined: Fri Apr 05, 2019 2:31 am

goat68 wrote:
Sat Sep 25, 2021 10:01 am
ShaunWhite wrote:
Sat Sep 25, 2021 5:11 am
Anbell wrote:
Sat Sep 25, 2021 4:10 am


Ive never heard that you need to be running BA.
They explain how to do it in the API FAQ so it's not a big deal, so long as it's used reasonably rather than habitually.
So exactly, i would be doing it on every single bet, so I'd call that habitually!
I dont think that'd be a problem, tho I dont know for sure. I do it 500 times an hour without a problem.
User avatar
goat68
Posts: 1434
Joined: Tue Jun 30, 2020 3:53 pm
Location: Hampshire, UK

Anbell wrote:
Sat Sep 25, 2021 10:24 am
goat68 wrote:
Sat Sep 25, 2021 10:01 am
ShaunWhite wrote:
Sat Sep 25, 2021 5:11 am

They explain how to do it in the API FAQ so it's not a big deal, so long as it's used reasonably rather than habitually.
So exactly, i would be doing it on every single bet, so I'd call that habitually!
I dont think that'd be a problem, tho I dont know for sure. I do it 500 times an hour without a problem.
Do you use BA though?
Im just hypothesising they would be more lenient to the BA subscription... since they know it's a standard feature of BA and BA is no doubt a significant volume them..
Anbell
Posts: 1209
Joined: Fri Apr 05, 2019 2:31 am

goat68 wrote:
Sat Sep 25, 2021 10:34 am
Anbell wrote:
Sat Sep 25, 2021 10:24 am
goat68 wrote:
Sat Sep 25, 2021 10:01 am


So exactly, i would be doing it on every single bet, so I'd call that habitually!
I dont think that'd be a problem, tho I dont know for sure. I do it 500 times an hour without a problem.
Do you use BA though?
Im just hypothesising they would be more lenient to the BA subscription... since they know it's a standard feature of BA and BA is no doubt a significant volume them..
I'm not a normal user, but I suspect that you're seeing ghosts where there arent any. (I dont know though)
User avatar
goat68
Posts: 1434
Joined: Tue Jun 30, 2020 3:53 pm
Location: Hampshire, UK

I follow Peter's twitter feed, just noticed this link today: https://www.betfairtradingblog.com/valu ... r-trading/
Of course I am doing "value betting", and this is a great article summarising my joys and woes, this paragraph is great and VERY true:
"
When you are doing well with a result dependent strategy, it’s easy to think you are a god and when you are doing badly it’s easy to think you are an idiot. This is one key reason that people have trouble replicating or even creating successful strategies.
"
Just wondering if Peter's reason for re-posting this was my recent Ups and Downs...!?!
CallumPerry
Posts: 561
Joined: Wed Apr 19, 2017 5:12 pm
Location: Wolverhampton

I’ve been thinking about some of the analysis comments from earlier in this tread for about 10 days now to the point where I feel inclined to ask more because I feel it may be really useful for my personal analysis.

I’m intrigued by the taking of the cumulative PnL and comparing it to an indicator such as the price-vwap. I believe it was said this checks the effectiveness of whatever you use as the x-axis i.e. price-vwap? I caught something about putting results into buckets and then doing this cumulative thing, I have no idea?

I’m mainly referring to Shaun’s explanation here at: Sat Sep 18, 2021 9:50 am

I’m theorising capturing a bunch of variables with each trade and creating one of these graphs for each to check their effectiveness simultaneously, not sure if this is how it works. Shaun or others, could anybody help me out please?

For (a bad) example, say I want to check whether my bot performs better at the point the markets has a high overround or a tight one, could I capture the book % at the point each of my trades close along with that trade’s PnL and group this data into discrete chunks i.e. book% = 100%-101%, 101.1%-102%, 102.1%-103% … 110%+

Then, map out the cumulative frequency or something and that will show where book% plays the most significant part in my system? If the results are similar with both halves of my sample, then it may be significant? Is that the rough gist of what’s going on? Apologies if I have misunderstood.
User avatar
ShaunWhite
Posts: 6941
Joined: Sat Sep 03, 2016 3:42 am

That's pretty much it Callum. I won't do a long reply with examples etc cos my pc is dead, and there's obviously nuance, but sorting results by things other than date can show where you're losing/winning. Grouping by bands is useful if you don't have a very big sample but you can usually get a feel for it not grouping. Your bet history and all the circumstances at the time the bets are placed is a really useful source of info.
CallumPerry
Posts: 561
Joined: Wed Apr 19, 2017 5:12 pm
Location: Wolverhampton

Cumulative Frequency Book %.xlsx
Is this the right kind of way to set it up? I'm selfishly taking over the thread for my own development but I feel this could help Goat68 too. I hypothetically created a book% figure which would be captured as the trade opens in column A and a hypothetical PnL result -£2 to £3, over 1,000 trades, in column B. Columns G:L are the buckets and columns P:U contains the cumulative frequency.

As all of the results have a +ve expectancy, the graph shows all lines rising nicely. Is this how you would use this? Any series that goes -ve is a weakness of that strategy? Just before sending this message, i've changed the last series to be -ve, so this would suggest that this hypothetical strategy doesn't perform well when the book% is >110%?

If I am correct, this looks like something I would love to do more regularly. To throw in some ideas, instead of Book%, could bucket: market volume, race course, selection volume %, a custom indicator etc? I hope I have not missed the mark here and that it sparks some ideas for others.
You do not have the required permissions to view the files attached to this post.
User avatar
goat68
Posts: 1434
Joined: Tue Jun 30, 2020 3:53 pm
Location: Hampshire, UK

Well AUS horse bot doing rubbish today, every one a loser, arghhh variance!
Screenshot 2021-09-29 at 10.55.23.png
You do not have the required permissions to view the files attached to this post.
arch4672
Posts: 15
Joined: Sat Feb 20, 2021 8:40 pm

goat68 wrote:
Wed Sep 29, 2021 10:56 am
Well AUS horse bot doing rubbish today, every one a loser, arghhh variance!
Screenshot 2021-09-29 at 10.55.23.png
Are you sure it's variance, or it's just that your backfitted strategy isn't very good?
User avatar
goat68
Posts: 1434
Joined: Tue Jun 30, 2020 3:53 pm
Location: Hampshire, UK

arch4672 wrote:
Wed Sep 29, 2021 12:59 pm
goat68 wrote:
Wed Sep 29, 2021 10:56 am
Well AUS horse bot doing rubbish today, every one a loser, arghhh variance!
Screenshot 2021-09-29 at 10.55.23.png
Are you sure it's variance, or it's just that your backfitted strategy isn't very good?
quite possibly!
I think I might have to start not relying on backtesting so much, it's far too easy to just "fit" things it would seem...
Post Reply

Return to “Trading Horse racing”