For a while now I have been pondering the possibility of backing the favourite at the dogs to see if there is any money to be made. Even though the rumours that greyhound racing is not the most virtuous sport you can make a wager on it is certainly the most abundant sport with over 140 races each day and that is just the UK. According to the Oxford Stadium website 35% of all favourites win! This however is across all tracks, distances and, classes of race. To this end I decided that the only true way of sorting the wheat from the chaff is to back every single favourite in every single race and use the power of excel to weed out losing statistics from the more profitable. So without ado I will post day by day findings on this blog
My criteria is simple I will back each race favourite 20 seconds before the post time to a liability of just £1.00 so in effect my average stake should be kept to about 30p per race. This means that should the favourite win then the Profit will be £1.00 and if loses the stake determined by the odds that are available at that moment in time on the betfair exchange. There have been instances where the odds of two of the dogs have been exactly the same at the 20 second point and no bet was struck. This is unfortunate but one of those things. After testing that the software worked as required on March 1st testing began in earnest on March 2nd and showed an overall profit of £13.41 with profits showing from all meetings except Crayford and Hove.
Click to Enlarge
Starting with Central Park, there were a total of 6 winning favourites giving a small profit of £2.66 The rest I will put into a table for easy reference.
Track
Winning Favs
Avg Stake
P/L
Crayford
1
45p
-£3.85
Doncaster
9 (2 Meetings)
69p
£4.73
Hove
6
£1.43 **
-£0.51
Monmore
10 (2 Meetings)
57p
£2.44
Newcastle
6
48p
£1.69
Oxford
8 (2 Meetings
63p
£1.26
Perry Barr
8
47p
£1.48
Romford
10 (2 Meetings)
53p
£3.03
Swindon
4
49p
£0.48
** Quite a few Odds On Favs Lost!!!
Crayford was quite obviously not a track to back the favourite on this day as just one favourite won, Hove had a few odds on favourites running hence the average stake of £1.43 with the shortest priced favourite of the day losing costing a massive £3.12 All in all not a bad day for the first day. There were a total of 160 races of which 64 favourites won giving a 40% win rate.
Lets go on to look at the best performing meeting of the day which was Doncaster. Producing an overall profit of £4.73 with 9 winning favourites at the 2 meetings. Of the 8 winning favourites 4 were in OR class races or Open Class and scored a 100% strike rate as did the OR3 class which accounted for 2 more winning favs. The sheer volume of information I have collected just over the last 20 days is by no means enough and in subsequent posts I will delve a little deeper into which tracks are more profitable than others using this system and which Races should be avoided on all counts after looking at the stats.
Just as a Taster the best performing track using this system between 2 March and 22 march is Nottingham with a profit of £15.89 The worst Track for winning Favourites is Sunderland with a loss of £22.03
My total profit for this period backing every race is £38.28
Betfair Historical Data has been, for some time now, openly available for free but it seems that no one is prepared to share just how you open the damned stuff!!! There are plenty of sites out there willing to share their python knowledge but stop short of a full explanation – very frustrating. At this point I have to give credit to “Trading The Market” guys who have finally given me the Eureka Moment! If you follow the steps below you will soon have all the research information to back test your systems. To begin with you do need to have a betfair account and to access which data you need you will need to be logged in and at the following page https://historicdata.betfair.com/#/home
For the purposes of this post I will be using the basic plan which costs nothing and has a data frequency of 1 minute. If you need more frequent data logs then you will have to purchase your data by the month but there are free plans available for limited months mainly in 2020. In this post I will be analysing the data from the North London Derby – Arsenal v Tottenham on the 1st October 2022 The easy way to handle this data is to “purchase” a month at a time so I will “Buy” the Soccer data from OCT 2022 To do this simply select the plan, sport, and month you need in the drop down boxes and click “Add to Purchases” and then the “purchase” as in the screen shot below
You will then be asked to confirm your purchase – Check that you haven’t included the wrong plan by mistake as the fee will be taken straight from your betfair balance. If everything is ok click confirm and you will see the following screen.
As you can see from the screen shot there are over 250000 markets! for our research we only need 3 or 4 of these so we need to find them. I want to know how the under/over Markets behaved when each goal went in during the match. To do this click on the “My Data” Tab as shown below and you will see your most recent purchases as well as plans you have bought before.
We now need to populate the down load fields with the information that we need so if we look at the match stats we can see that there were 4 goals scored in total.
I want to see the following market odds O/U 1.5 O/U 2.5 O/U 3.5 O/U 4.5 O/U 5.5 O/U 6.5 and the Match. To do this we need to populate the download files panel as follows:
In the Plan Box (which is at the top but hidden in this screen shot) enter Basic In the Sport Box enter Soccer In the From and Upto Boxes enter 1 Oct 2022 Leave the event ID Blank In the Event Name box enter Arsenal In the Market Type box select the markets you need and in this case its Match_Odds Over_Under_25 Over_Under_35 Over_Under_45 Over_Under_55 Over_Under_65 In the Country Box select GB (There is no UK) and in the File Type Box select M Click the Add to Downloads button You now should have 12 files ready to down load. Click the Yellow Download button
You will find this file in your downloads folder as a Data “tar” file
Double click the file to un pack it in your unzipping program
Keep double clicking until you get to the actual data files
You should now see your 6 “bz2” files To make them easy to find Create a new folder on your desktop and name it ARS v TOT
You can now drag and drop the files into this folder from your unpacking software
We now need to convert these files into a readable format and Betfair provide one such program easily available from their website.
The first file we want to see is the Match Odds and this will usually be the largest file as most volume would have passed through that market. In this case we are looking at File 1.203213491
These are not very user friendly file names so I am going to rename them. to do this we simply select a file from the list using the betfair Historical Data Processor as below and note the file number and which market it contains. In this case File 1.203213498 is the O/U 6.5 goals Market. We dont want to do anything else yet. We are just identifying the files for our convenience
We can then repeat the process with all the files noting the file numbers to the markets and then rename all the files as below
We are now going to convert these files into meaningful excel ranges where we can look at them in more detail. To do this we again select a file from the folder using the Historical Data Processor and click the download settings button as shown in the screenshot below
This will open the settings dialog box as shown below
We now need to populate our spreadsheet with any relevant information we will need by clicking the drop down box arrow and selecting the fields by ticking the relevant tick boxes (dont worry if you put too much info into the spreadsheet as we can delete it later)
Once you have selected what you need close the drop down box by clicking the “up” arrow and because we are just interested in the in play data slide the “preplay” button to off as shown below and click save
We are now ready to download the file as a CSV file into Excel we do this by simply clicking the “Download CSV” button as shown and let the wonders of better brains than our own do its stuff.
This will put a zipped file into your downloads folder with the original file number
If we double click this file it will unpack and be available as a csv file which we can open with excel.
It may be a good idea to move this file to a new folder on your desktop and rename it as I have done here
If we open this in excel this is what we get.
We can now start to manipulate the data to best suit our needs. We can start by removing some of the columns we dont need. I have removed the following columns to be left with the following spreadsheet. Market ID Selection ID
The next task is to make the published time more reader friendly. as you can see this is total jibberish, it means nothing. If you select cell A2, in the address window you will see that the full date and time is shown
So that column A shows the time in hrs and minutes we need to select hh:mm from the custom cell formats as shown below.
When we click ok then cell A2 will show the actual time of the published data for that row
To convert all the row simply select Column A and repeat
When you click ok your spreadsheet should look like this
you will notice that the times are repeated 3 times this is because there are 3 separate outcomes within this betfair market – Arsenal to win The Draw and Tottenham to win. If we now draw our attention to Column F this is the last traded price column and we could do with tidying up this so that all the figures are shown to 2 decimal places. To do this simply select Column F and using the cell number format dropdown box, select “Number” as shown below.
Your spreadsheet should now look something like this
A bit more tidying up by changing the column headers and this is what you should be looking at.
This is your basic data spreadsheet for the entire match for the Match Odds Market You can now save it into a folder and repeat the process with the other 5 markets
In my next post I will show you how to manipulate the data so that we can see what happened in each market as each goal went in
Please like and subscribe to my blog so that I can get a feel for what content can be most useful to you.
Betornot 