Greyhound Mystique

[murdok]

I also like to work with this date https://www.gbgb.org.uk/ , there is only one problem, they load the new races for the next day too late

[Archery1969]

I also like to work with this date https://www.gbgb.org.uk/ , there is only one problem, they load the new races for the next day too late

What ?

[murdok]

I also like to work with this date https://www.gbgb.org.uk/ , there is only one problem, they load the new races for the next day too late

What ?

sorry wrong link
e like to get data from here https://form.dog/

[Archery1969]

I also like to work with this date https://www.gbgb.org.uk/ , there is only one problem, they load the new races for the next day too late

What ?

sorry wrong link
e like to get data from here https://form.dog/

The data on that link is for today being 18/06/2020, so why is the data too late ?

[towelfox]

I also like to work with this date https://www.gbgb.org.uk/ , there is only one problem, they load the new races for the next day too late

What ?

sorry wrong link
e like to get data from here https://form.dog/

Race data is loaded at 1am and re-checked at 7am.

[murdok]

Race data is loaded at 1am and re-checked at 7am.

Archery1969 was what I was talking about ,

if i work the next day i have to be up late to make selections

[sniffer66]

Bit more playing around scraping the SL Site yesterday. Managed to add the Grade and Distance easily

Slightly harder to get the form table per dog but managed to get that into another array, adding the dog name as a primary key so I can relate the two tables.

Now I can start querying both tables to return various averages etc over x races/time periods.

Todays data attached. Form table is only a subset unfortunately as the full file is too big too upload (14k rows)

Output.xlsx

form_subset.xlsx

[MD7890]

Am I misunderstanding what "Best" time means? I assumed it was the dog's fastest recorded race time at the race distance...yet I have seen many cases where the detailed form shows faster times than the one shown in the headline numbers...
Example below: Best time = 28.83, yet the 2nd-5th row (and more below) show race times below this number.

[towelfox]

Am I misunderstanding what "Best" time means? I assumed it was the dog's fastest recorded race time at the race distance...yet I have seen many cases where the detailed form shows faster times than the one shown in the headline numbers...
Example below: Best time = 28.83, yet the 2nd-5th row (and more below) show race times below this number.

You're right. Their 'best' is just wrong on this occasion.

You have to dig into the full form for accurate stats unfortunately. It's not always wrong but you don't know until you check.

[sniffer66]

Am I misunderstanding what "Best" time means? I assumed it was the dog's fastest recorded race time at the race distance...yet I have seen many cases where the detailed form shows faster times than the one shown in the headline numbers...
Example below: Best time = 28.83, yet the 2nd-5th row (and more below) show race times below this number.

You're right. Their 'best' is just wrong on this occasion.

You have to dig into the full form for accurate stats unfortunately. It's not always wrong but you don't know until you check.

and the posted last/best times in the header often show the last time faster than the best

[Archery1969]

Right, lets get this show on the road and help the new comers who must be trying to walk in a mine field....

The original post has transformed into the following.

Data Gathering

We need access to information on each Dog being:

- Venue
- Date
- Time
- Grade
- Distance
- Selection
- Best time of last 30 days
- Races at grade
- Races at distance
- Wins at grade
- Wins at distance

Strategy

Now we have our data we then calculate a percentage chance for each dog based on their best time, races and wins. From that then you work out what you think their odds should be. There are a number of ways of doing this and their are better people than me on the forum who will help you with this. The key is getting the correct data above. You can get this from SportingLife, RacingPost or TimeForm to name a few. Just makesure the data is relevant as you can get some errors in the information being provided.

- If the Betfair odds for the selection are 5% or more higher than our calculation then we BTL or straight back the selection and close out at 40s to post time or let the bet stand. You decide.
- If the Betfair odds for the selection are 5% or more lower than our calculation then we LTB or straight lay the selection and close out at 40s to post time or let the bet stand. You decide.

What can go wrong

The market disagrees with all our hard work and thinks another Dog should be the favourite.

- A dog is stepping down in grade and the market thinks it will do much better at the new grade.
- A dog is stepping up in grade and the market thinks it will still blow the field away at the new grade.
- A kennel or trainer or owner or pissed rich person is chucking money on another dog.
- A dog is called 'Big Dick' and pissed women and men start jumping on. Forgive the pun.

Further advice

- Ignore the race if it contains a dog stepping up or down and has no data for the new grade and/or distance. Sometimes they win and sometimes their lose.
- Wait to see if our selection starts to get an increase in volume and then bet or trade accordingly. You can always cut your losses and switch to another dog if ours looks a dead duck.
- Don't bet on anything with a big dick.

I suggest you paper trade the strategy for awhile until your happy with the results. Do not risk money you can't afford to chuck away.

That's it. Good luck and stay safe.

[sniffer66]

Bit more work done on the screen scrape from UK Sporting Life

I have 2 arrays being created, first is the current race data, second is the form table for each dog (available from the "Form" dropdown under each listed dog/race)
Using those I can query and calc a few averages and key stats per dog\race by linking to the form data and returning the results to the main array.
Not set up the last 30 days query as yet so averages are worked out over latest 3 races for the distance - may extend that as more come in from post lockdown, or just calc based on date

Next step is to add in finishing position over last x races

Col. I - Best Time for Dist
Col J - Last time for Dist
Col K - Best\Last Averaged time ( I+J/2)
Col L - Average Time over last 3 Races (For current race distance only)
Col M - Average Split over last 3 Races (For current race distance only)
Col N - Is Dog with best time for distance > 0.1s over 2nd best timed dog Y/N
Col 0 - Is Dog Best average time over distance (from Col L) Y/N
Col P - Is Dog Best Split time over distance (from Col M) Y/N
Col Q - is Dog Fastest Best (from Col I) Y/N

I can't 100% guarantee the data but all my checks so far match out. Still getting the odd zero result in a cell so need to track those down.
Results attached for todays UK races (that are on SL) - enjoy

Output.xlsx

[sniffer66]

Bit more work done on the screen scrape from UK Sporting Life

I have 2 arrays being created, first is the current race data, second is the form table for each dog (available from the "Form" dropdown under each listed dog/race)
Using those I can query and calc a few averages and key stats per dog\race by linking to the form data and returning the results to the main array.
Not set up the last 30 days query as yet so averages are worked out over latest 3 races for the distance - may extend that as more come in from post lockdown, or just calc based on date

Next step is to add in finishing position over last x races

Col. I - Best Time for Dist
Col J - Last time for Dist
Col K - Best\Last Averaged time ( I+J/2)
Col L - Average Time over last 3 Races (For current race distance only)
Col M - Average Split over last 3 Races (For current race distance only)
Col N - Is Dog with best time for distance > 0.1s over 2nd best timed dog Y/N
Col 0 - Is Dog Best average time over distance (from Col L) Y/N
Col P - Is Dog Best Split time over distance (from Col M) Y/N
Col Q - is Dog Fastest Best (from Col I) Y/N

I can't 100% guarantee the data but all my checks so far match out. Still getting the odd zero result in a cell so need to track those down.
Results attached for todays UK races (that are on SL) - enjoy




Output.xlsx

Update to the sheet. Added wins over race distance.
Originally I went for a straight number of wins but some dogs have 50+ races in the form table, some only 2-5. On that basis I chose a wins over distance/races over distance ratio - added columns should be self explanatory

I've also subbed some zero cell figures with a "-1" to prevent incorrrect Yes/No rankings

None of the data used in anger as yet, so no idea how it reflects movements\results in the real world

Output.xlsx

[wearthefoxhat]

Bit more work done on the screen scrape from UK Sporting Life

I have 2 arrays being created, first is the current race data, second is the form table for each dog (available from the "Form" dropdown under each listed dog/race)
Using those I can query and calc a few averages and key stats per dog\race by linking to the form data and returning the results to the main array.
Not set up the last 30 days query as yet so averages are worked out over latest 3 races for the distance - may extend that as more come in from post lockdown, or just calc based on date

Next step is to add in finishing position over last x races

Col. I - Best Time for Dist
Col J - Last time for Dist
Col K - Best\Last Averaged time ( I+J/2)
Col L - Average Time over last 3 Races (For current race distance only)
Col M - Average Split over last 3 Races (For current race distance only)
Col N - Is Dog with best time for distance > 0.1s over 2nd best timed dog Y/N
Col 0 - Is Dog Best average time over distance (from Col L) Y/N
Col P - Is Dog Best Split time over distance (from Col M) Y/N
Col Q - is Dog Fastest Best (from Col I) Y/N

I can't 100% guarantee the data but all my checks so far match out. Still getting the odd zero result in a cell so need to track those down.
Results attached for todays UK races (that are on SL) - enjoy




Output.xlsx

Nice work.

Found a nice pattern on the Nottingham ones.

Re;- The rogue times are best ignored. ie : Nott 10.07; T5 Alfies Lane in (Column K; 30.355 )

11.47 Test - Dutch highlighted and Italic selected

11.47test.jpg

[sniffer66]

Bit more work done on the screen scrape from UK Sporting Life

I have 2 arrays being created, first is the current race data, second is the form table for each dog (available from the "Form" dropdown under each listed dog/race)
Using those I can query and calc a few averages and key stats per dog\race by linking to the form data and returning the results to the main array.
Not set up the last 30 days query as yet so averages are worked out over latest 3 races for the distance - may extend that as more come in from post lockdown, or just calc based on date

Next step is to add in finishing position over last x races

Col. I - Best Time for Dist
Col J - Last time for Dist
Col K - Best\Last Averaged time ( I+J/2)
Col L - Average Time over last 3 Races (For current race distance only)
Col M - Average Split over last 3 Races (For current race distance only)
Col N - Is Dog with best time for distance > 0.1s over 2nd best timed dog Y/N
Col 0 - Is Dog Best average time over distance (from Col L) Y/N
Col P - Is Dog Best Split time over distance (from Col M) Y/N
Col Q - is Dog Fastest Best (from Col I) Y/N

I can't 100% guarantee the data but all my checks so far match out. Still getting the odd zero result in a cell so need to track those down.
Results attached for todays UK races (that are on SL) - enjoy




Output.xlsx

Nice work.

Found a nice pattern on the Nottingham ones.

Re;- The rogue times are best ignored. ie : Nott 10.07; T5 Alfies Lane in (Column K; 30.355 )

Thanks, still trying to weed out some of the anomalies so I can use it on full auto but not too hard to spot them manually