Trading Greyhound racing

[Anbell]

I've been making money on greyhounds!! Ive done 3 long days of trading in it to test over a reasonable volume.

Sadly, its at £1 stakes and my eyeballs hurt.

Seems a lot simpler to me than whats been said on the thread...I assume the catch is that when I try up my stakes to anything significant i'll end up with a lot of partial matches and mess up my strategy...

Three days isnt a long time for most of us.

[Fugazi]

I've been making money on greyhounds!! Ive done 3 long days of trading in it to test over a reasonable volume.

Sadly, its at £1 stakes and my eyeballs hurt.

Seems a lot simpler to me than whats been said on the thread...I assume the catch is that when I try up my stakes to anything significant i'll end up with a lot of partial matches and mess up my strategy...

Three days isnt a long time for most of us.

In terms of variance, maybe not. In poker it can take a million hands to smooth out variance.

But its encouraging so far.

Unless you meant the eyeball strain after just 3 days hah

[jamesedwards]

In all the greyhound models I've done through the years I've always found a huge relationship between ↑stakes and ↓ROI%. Something that works well at £1 can crash and burn at £2.

[Kai]

I've been making money on greyhounds!! Ive done 3 long days of trading in it to test over a reasonable volume.

Sadly, its at £1 stakes and my eyeballs hurt.

Seems a lot simpler to me than whats been said on the thread...I assume the catch is that when I try up my stakes to anything significant i'll end up with a lot of partial matches and mess up my strategy...

Three days isnt a long time for most of us.

In terms of variance, maybe not. In poker it can take a million hands to smooth out variance.

But its encouraging so far.

Unless you meant the eyeball strain after just 3 days hah

He's right, early days as they say. Difficult to curb your enthusiasm when starting but it's extremely likely there will be some false dawns before it dawns on your real etc.

For eye strain however 3 possible suggestions

1) https://justgetflux.com/ (had zero eye strain issues last 15 years or so thx to f.lux)

2) slightly lowering bet refresh rate might help

3) blink more often

[Fugazi]

In all the greyhound models I've done through the years I've always found a huge relationship between ↑stakes and ↓ROI%. Something that works well at £1 can crash and burn at £2.

Thats really useful information. Going to build volume at £1, then start messing with stakes and see what happens if im certain its a working strategy at £1

[Fugazi]

Three days isnt a long time for most of us.

In terms of variance, maybe not. In poker it can take a million hands to smooth out variance.

But its encouraging so far.

Unless you meant the eyeball strain after just 3 days hah

He's right, early days as they say. Difficult to curb your enthusiasm when starting but it's extremely likely there will be some false dawns before it dawns on your real etc.

For eye strain however 3 possible suggestions

1) https://justgetflux.com/ (had zero eye strain issues last 15 years or so thx to f.lux)

2) slightly lowering bet refresh rate might help

3) blink more often

Yep, no stranger to false hope. Must be the hundredth time I've been convinced i've finally discovered a winning strategy in something!

[Crazyskier]

In all the greyhound models I've done through the years I've always found a huge relationship between ↑stakes and ↓ROI%. Something that works well at £1 can crash and burn at £2.

Wise words that mirror my own experiences, and I've been at it daily for almost a decade.

CS

[BWhizz]

Been trying to crack the greyhounds for a year. Scapling, laying and dutching. To no avail. All seems too erratic to me. So moved onto horses and footy.

Any pointers/advice from anyone who is successful with the dogs would be well received.

[sniffer66]

In all the greyhound models I've done through the years I've always found a huge relationship between ↑stakes and ↓ROI%. Something that works well at £1 can crash and burn at £2.

I'd be interested in a discussion on modelling the dogs.
I went through an exercise last year where I downloaded every single race card from The Racing Post each day, for about 6 months (approx 22,000 races), and wanted to test if historical stats had any relationship to performance and results for straight backing and laying. I took the usual card data, with 5 previous race stats, and had my code calculate a few averages on time, split, avg grade, avg 1st bend, avg time, relationship between 1st bend\finish, trap bias etc

Dogs1.JPG

I ended up with about 15 stats that I wanted to assess, and decided to give each of those stats a random weighting out of 100, with all 15 adding up to 100% of the total available. Then I created 5000 random permutations of the weighting amounts and applied those to each race card in turn, giving each dog a score. One permutation may have had split at 80% , avg grade at 20% , with the rest at zero, or all at around 7% to make up to 100

Using the lowest (lay) or highest score (back) I was able to apply all 5000 permutations to the downloaded BF results\odds data and assess results based on every race, and by track and by distance\class. The theory being that by applying the best performing permutation to each days upcoming race\type track\class I would hopefully come up with a winning weighting for each.
I then dropped those into The Staking Machine and was able to analyse possible edge %, p-value, Archie score etc, and also assess the most profitable odds ranges. Results showed I had quite a decent edge, at specific odds ranges

Most tracks\classes were showing a decent profit for the top performing permutation. But, then you get the question re is that just back fitting ? So, using a lot of useful info from Sean's posts I used 80% of the 22k races, regenerated the best permutations from those, then applied the top performers to the unused 20%

Which is the stage when my edge dropped to almost nothing lol. So, the conclusions were that historical stats have no affect on predicting future results for dogs, or my reasoning was flawed

[MemphisFlash]

people rely to much on data from previous races which is irrelevant to the actual days race
same as any other stats for any other sport.
what has gone before is dead and buried and should not be used.
what only matters is what happens on the day of the event.

[sniffer66]

people rely to much on data from previous races which is irrelevant to the actual days race
same as any other stats for any other sport.
what has gone before is dead and buried and should not be used.
what only matters is what happens on the day of the event.

Which was one of my conclusions. But it was an interesting exercise, and I learnt quite a bit doing it. So it wasn't wasted

[The Silk Run]

I concur. Of the four instances I deploy each day none of which rely on historical data !!!

[MemphisFlash]

All historical data does is turns you into a geek that can use excel, and takes away from the real objective - trading.
It's like in football when they talk about number of passes, or time of possession .
It doesn't mean anything
The only stat that matters in a football game is the final score.

[The Silk Run]

I couldn't agree more Tom. Keep it simple, basic fundamentals. Each animal has a 1:6 chance of winning ....

[MemphisFlash]

I couldn't agree more Tom. Keep it simple, basic fundamentals. Each animal has a 1:6 chance of winning ....

yes that is correct, finally someone gets it.