Winning Price stats

The sport of kings.
Post Reply
User avatar
SeaHorseRacing
Posts: 2893
Joined: Fri May 20, 2016 7:06 pm

Thanks for all your responses
spreadbetting
Posts: 3140
Joined: Sun Jan 31, 2010 8:06 pm

SeaHorseRacing wrote:
Wed Oct 17, 2018 1:38 pm
spreadbetting wrote:
Wed Oct 17, 2018 12:57 pm
I can do you a dump from my database but there are over 50'000 unique BSP's that have been returned so you'll have to tweak the file yourself if you want it in bands, I can dump out the number of selection, number of winners, number of dead heats so easy enough to work out strie rates from those
Yes please
Here you go hopefully it's all quite obvious, in xlsx and wouldn't let me upload to the forum as csv or xls for some reason probably too large for the forum but only Mb's

http://www.mediafire.com/file/6owe99dlz ... .xlsx/file

Shouldn't be hard for you to query it in bands with a few sumifs, I stuck a simple range checker on it
User avatar
Dallas
Posts: 22673
Joined: Sun Aug 09, 2015 10:57 pm
Location: Working From Home

gstar1975 wrote:
Wed Oct 17, 2018 1:28 pm
What about the BSPs with 2.01 or 2.04, 1.98 etc or 3.04 etc?
I don't have the time to do them all I just picked out 2.0 and 3.0 as examples, you can get all the rest from the dump Spreadbetting has kindly uploaded
User avatar
SeaHorseRacing
Posts: 2893
Joined: Fri May 20, 2016 7:06 pm

spreadbetting wrote:
Wed Oct 17, 2018 2:07 pm
SeaHorseRacing wrote:
Wed Oct 17, 2018 1:38 pm
spreadbetting wrote:
Wed Oct 17, 2018 12:57 pm
I can do you a dump from my database but there are over 50'000 unique BSP's that have been returned so you'll have to tweak the file yourself if you want it in bands, I can dump out the number of selection, number of winners, number of dead heats so easy enough to work out strie rates from those
Yes please
Here you go hopefully it's all quite obvious, in xlsx and wouldn't let me upload to the forum as csv or xls for some reason probably too large for the forum but only Mb's

http://www.mediafire.com/file/6owe99dlz ... .xlsx/file

Shouldn't be hard for you to query it in bands with a few sumifs, I stuck a simple range checker on it

Interesting that there’s little pockets of profits... for example 3.06 to 3.09..

Think this is a coincidence? Maybe just noise?

Just above every crossover point 2 3 6 all showing a profit.
User avatar
Gauss
Posts: 137
Joined: Sun Sep 23, 2018 10:11 am

Dallas wrote:
Wed Oct 17, 2018 1:02 pm
SeaHorseRacing wrote:
Wed Oct 17, 2018 8:52 am

Even money has a 50% implied chance but what is the actual strike rate of even money horses at sp?
Since 2009 there has been 496 with a BSP of 2.0 (50% IP)
243 Won (48.99%)
253 Lost

1103 had a BSP of 3.0 (33.33% IP)
366 Won (33.18%)
737 Lost
Drat, I was 1.01% out on my reckoning, forgot to account for variance.
User avatar
ruthlessimon
Posts: 2094
Joined: Wed Mar 23, 2016 3:54 pm

Looked at generically, sure, BFSP is fair.

But if SP was fair, I shouldn't see a table like this - everything should be 0%

God how I wish Mr. Butler was still around :D
You do not have the required permissions to view the files attached to this post.
User avatar
ruthlessimon
Posts: 2094
Joined: Wed Mar 23, 2016 3:54 pm

Something else that has fascinated me, is BFSP even special? Is it better than 01:37 etc
Atho55
Posts: 637
Joined: Tue Oct 06, 2015 1:37 pm
Location: Home of Triumph Motorcycles

I struggle to see the value of constructing an IP based upon the result of a mechanism set up to settle bets or so it appears.

https://promo.betfair.com/betfairsp/FAQ ... kings.html
spreadbetting
Posts: 3140
Joined: Sun Jan 31, 2010 8:06 pm

ruthlessimon wrote:
Wed Oct 17, 2018 6:44 pm
Something else that has fascinated me, is BFSP even special? Is it better than 01:37 etc
Depends on your idea of special , 01:37 might not be much use to a trader wanting some kind of insurance his closing bet will be matched at the off if he hasn't managed to get closed beforehand, BFSP may be. For a gambler 01:37 may well offer a better price on the winner.

If we take BSP and 01:37 as a whole, both are likely to be returning to a 101% book so theoretically they'd be no advantage pricewise of taking one over the other.
User avatar
ShaunWhite
Posts: 9731
Joined: Sat Sep 03, 2016 3:42 am

ruthlessimon wrote:
Wed Oct 17, 2018 6:37 pm
Looked at generically, sure, BFSP is fair.

But if SP was fair, I shouldn't see a table like this - everything should be 0%
They would all be zero (ish) on a large enough sample? If there were specific points of value that were so easy to find, somebody would be taking them and the price would be affected. Remember even Mr king of the manual traders runs multiple servers (7+) all keeping everything in check. Btw How many horses are on each of those chart price points? I'm guessing the middle section which is iro zero gain represent more runners than the outlying ones. Bigger sample, closer to zero?
User avatar
ruthlessimon
Posts: 2094
Joined: Wed Mar 23, 2016 3:54 pm

ShaunWhite wrote:
Wed Oct 17, 2018 9:49 pm
They would all be zero (ish) on a large enough sample?
Could be. This is one of those awkward dilemmas - a sample that's too big is called generic; & one that''s too small gets called insignificant.

The bit that's "interesting" was a sample of 564 trades over a 4mth period.
You do not have the required permissions to view the files attached to this post.
Last edited by ruthlessimon on Wed Oct 17, 2018 11:27 pm, edited 1 time in total.
elofan0
Posts: 316
Joined: Fri Jan 13, 2017 4:44 pm

price stats would change from different events

https://www.flatstats.co.uk/blog/turf/nursery1.html
ColinB
Posts: 2
Joined: Thu Oct 18, 2018 2:51 am

Big thanks Spreadbetting, these figures more or less confirm an idea that has been floating around in my head for a while.
If I could ask 1 more favour .... Could you confirm the date range that these figures were compiled in.
This will enable me to keep my preferred price range up to date from here on in.
Much appreciated,
spreadbetting
Posts: 3140
Joined: Sun Jan 31, 2010 8:06 pm

ColinB wrote:
Thu Oct 18, 2018 2:59 am
Big thanks Spreadbetting, these figures more or less confirm an idea that has been floating around in my head for a while.
If I could ask 1 more favour .... Could you confirm the date range that these figures were compiled in.
This will enable me to keep my preferred price range up to date from here on in.
Much appreciated,
27-03-2008 to 16-10-2018 , think it's all the data files since Betfair started releasing them. Think that's why some of the data has been reported as 1001, rather than 1000, in the early files. I'd also imagine the BSP prices are a lot closer to 100% books these days but the older data shouldn't skew the final figures much
ps1967
Posts: 36
Joined: Tue Mar 23, 2010 11:39 pm

spreadbetting wrote:
Wed Oct 17, 2018 2:07 pm
SeaHorseRacing wrote:
Wed Oct 17, 2018 1:38 pm
spreadbetting wrote:
Wed Oct 17, 2018 12:57 pm
I can do you a dump from my database but there are over 50'000 unique BSP's that have been returned so you'll have to tweak the file yourself if you want it in bands, I can dump out the number of selection, number of winners, number of dead heats so easy enough to work out strie rates from those
Yes please
Here you go hopefully it's all quite obvious, in xlsx and wouldn't let me upload to the forum as csv or xls for some reason probably too large for the forum but only Mb's

http://www.mediafire.com/file/6owe99dlz ... .xlsx/file

Shouldn't be hard for you to query it in bands with a few sumifs, I stuck a simple range checker on it
You wouldn't happen to have something like this for greyhounds would you? Or any idea where I could possibly get something similar?
Post Reply

Return to “Trading Horse racing”