SPEED RATINGS

[MCLARKE]

A bit off topic but a good example of why I tend to use Betfair

Windsor 6.20

ISP 17/2

BSP 20.85

[Zilzalian]

3 hours ago, Zilzalian said:

You are correct to point that out because the second number is a typo and should so should read 127 85 212

totally bolloxed that up now. number should be 127 75 202 the adjustment of SR adds 10 points (85) but i don't use the adjustment.

It may be worth looking at this result sometime in the future and try to understand how we ended up at opposite ends, i would be just as curious if the dragon king came last and where's claire came first. At the moment i am trying to understand the difference in my results between my rating v My rating + Stop watch and i am still undecided which is the better score. @harry_rag might want to give me his opinions if i send him my data for the 2yolds this year and see what he makes of them after all hes got sod all to do for a while. 🤣🤣🤣 now the footy is over.

[MCLARKE]

2 hours ago, harry_rag said:

That feels like random short term variance even allowing for any possible seasonal element.

I can't help wondering if that's a relatively arbitrary filter in terms of why such a narrow band should be more profitable than anything longer or shorter and you're due to see a correction where the winners trend into a different price point. Just keep building up the data and enjoying the returns for as long as they're this good!

There was some logic in that this was the odds band that gave the best returns in the previous years results but I agree I don't feel comfortable that this turned out to be the main reason why the system has been profitable this year.

I expect some reversion to the mean, today's blank could be the start.

[MCLARKE]

I have now ranked each race this year in terms of speed figures achieved. The results are encouraging.

 

[MCLARKE]

Because I have rounded my speed figures to a whole number there are some joint top rated / joint 2nd top rated. If I take this restriction out the results are :-

[Villa Chris]

37 minutes ago, MCLARKE said:

Because I have rounded my speed figures to a whole number there are some joint top rated / joint 2nd top rated. If I take this restriction out the results are :-

You think it’s better to round it off or be more precise ? 

[harry_rag]

@MCLARKE what are PS and PP?

14 hours ago, Zilzalian said:

At the moment i am trying to understand the difference in my results between my rating v My rating + Stop watch and i am still undecided which is the better score. @harry_rag might want to give me his opinions if i send him my data for the 2yolds this year and see what he makes of them after all hes got sod all to do for a while. 🤣🤣🤣 now the footy is over.

Plenty to do, a mixture of taking a welcome breather from full on betting, reviewing past data, tweaking the future pricing approach and doing "real life" stuff. That said, I'll happily take a look at your data if you like. I'll just have to resist any inclination to go too far down the rabbit hole and spending days on it! 

[harry_rag]

4 hours ago, MCLARKE said:

Because I have rounded my speed figures to a whole number there are some joint top rated / joint 2nd top rated. If I take this restriction out the results are :-

Be interested to know the respective performance of backing the one at the shorter/longer forecast SP. I'm assuming the results are to £1 LS? The 2nd rated are trouncing the top rated!

[MCLARKE]

6 hours ago, Villa Chris said:

You think it’s better to round it off or be more precise ? 

Normally I would round it to make it more manageable but when I am looking at ranking then being more precise means there are less joint rankings. In this instance the returns are slightly better.

[MCLARKE]

2 hours ago, harry_rag said:

Be interested to know the respective performance of backing the one at the shorter/longer forecast SP. I'm assuming the results are to £1 LS? The 2nd rated are trouncing the top rated!

I will have a look at that. The numbers would suggest backing the 2nd rated but that doesn't seem logical.

This analysis is only for this season, I need to go back a few years to see what the results are.

[Villa Chris]

9 minutes ago, MCLARKE said:

Normally I would round it to make it more manageable but when I am looking at ranking then being more precise means there are less joint rankings. In this instance the returns are slightly better.

Yeah I think it would be better not rounding it off so you don’t have X amount of horses rated the same . The way I’ve started rating I’ve got that same problem because I’ve stopped allocating a % to certain criteria’s for example distance or going . Just doing a straight points system now but I’m going to look at a way to separate horses with same points if I can. Obviously you’ll get horses that can’t be separated .

[Zilzalian]

Best bet today both from a speed perspective and value is DRAMA in the 730 Kempton 9/2 is 3points over what it should be imho.

[Zilzalian]

9 minutes ago, Villa Chris said:

Yeah I think it would be better not rounding it off so you don’t have X amount of horses rated the same . The way I’ve started rating I’ve got that same problem because I’ve stopped allocating a % to certain criteria’s for example distance or going . Just doing a straight points system now but I’m going to look at a way to separate horses with same points if I can. Obviously you’ll get horses that can’t be separated .

Where I have "joints" I just add stop watch (RP) on to get a clear top rated. It is rare that they then come out level again.

[MCLARKE]

COURSE	TIME	SELECTION	ODDS	BOOKMAKER
Beverley	3.30	Bay Dream Believer	7.50	LADBROKES
Beverley	4.00	Spanish Angel	12.00	BETFAIR
Beverley	5.00	Ugo Gregory	4.00	BETFAIR

[MCLARKE]

30 minutes ago, Zilzalian said:

Best bet today both from a speed perspective and value is DRAMA in the 730 Kempton 9/2 is 3points over what it should be imho.

It is also my AW selection today

[Villa Chris]

1 hour ago, Zilzalian said:

Best bet today both from a speed perspective and value is DRAMA in the 730 Kempton 9/2 is 3points over what it should be imho.

Was expecting this to be a non handicap race when I looked . 

[harry_rag]

Ok, not speed but still ratings; been doing some analysis on the old anytime data following the discussion with @richard-westwood and @Villa Chris in the daily racing thread.

I think the conclusion here is that it's worth reducing the required edge over my fair odds from the current 8% to 6%. Is it so blindingly obvious that you'd question why I'm even asking or am I missing something? There's the argument for sticking with the current approach and building up more data but I'm inclined to make the change and see how it goes. Here are the figures for 6, 7 and 8%.

6.00%	Bets	£	ROI
A	802	58.94	7.35%
B	555	2.16	0.39%
7.00%	Bets	£	ROI
A	728	37.09	5.09%
B	629	24.01	3.82%
8.00%	Bets	£	ROI
A	687	45.29	6.59%
B	670	15.81	2.36%

• Current approach is 8% leading to 687 A bets (best price meets the required edge) and 670 B bets (best price is better than fair but short of the required edge). The A bets are the ones I'd place real money on. There's no case for increasing the edge above 8% as too many value bets are left unbacked as B bets and too much profit is missed.
• Suggested approach is 6%. A 17% increase in bets and a 30% increase in profit. There's no real case for going below 6% as the returns tail off and the bet volume becomes excessive.
• A quick dismissal of the elephant in the spreadsheet, the worse numbers for 7%. 7.5% looks more like 8 and 6.5% looks more like 6 so I'd put that down to small sample size (just 39 bets between 7 and 7.5%) and ignore it as "noise".

There are 3923 players in the data so far and the 2566 C bets (price is worse than fair) can be safely discarded as the ROI is -9.46%.

 

[richard-westwood]

5 minutes ago, harry_rag said:

Ok, not speed but still ratings; been doing some analysis on the old anytime data following the discussion with @richard-westwood and @Villa Chris in the daily racing thread.

I think the conclusion here is that it's worth reducing the required edge over my fair odds from the current 8% to 6%. Is it so blindingly obvious that you'd question why I'm even asking or am I missing something? There's the argument for sticking with the current approach and building up more data but I'm inclined to make the change and see how it goes. Here are the figures for 6, 7 and 8%.

6.00%	Bets	£	ROI
A	802	58.94	7.35%
B	555	2.16	0.39%
7.00%	Bets	£	ROI
A	728	37.09	5.09%
B	629	24.01	3.82%
8.00%	Bets	£	ROI
A	687	45.29	6.59%
B	670	15.81	2.36%

• Current approach is 8% leading to 687 A bets (best price meets the required edge) and 670 B bets (best price is better than fair but short of the required edge). The A bets are the ones I'd place real money on. There's no case for increasing the edge above 8% as too many value bets are left unbacked as B bets and too much profit is missed.
• Suggested approach is 6%. A 17% increase in bets and a 30% increase in profit. There's no real case for going below 6% as the returns tail off and the bet volume becomes excessive.
• A quick dismissal of the elephant in the spreadsheet, the worse numbers for 7%. 7.5% looks more like 8 and 6.5% looks more like 6 so I'd put that down to small sample size (just 39 bets between 7 and 7.5%) and ignore it as "noise".

There are 3923 players in the data so far and the 2566 C bets (price is worse than fair) can be safely discarded as the ROI is -9.46%.

 

See ...the racing is diff...if I reduce the % to below 10 you get too much chaff ....lots of outsiders that look value but are inflated really so the 10% seems to cull the crappier bets and for a horse to achieve 10%+ it must have half decent form today 

[harry_rag]

6 minutes ago, richard-westwood said:

See ...the racing is diff...if I reduce the % to below 10 you get too much chaff ....lots of outsiders that look value but are inflated really so the 10% seems to cull the crappier bets and for a horse to achieve 10%+ it must have half decent form today 

Yeah I focus on the top end of the market so my fair odds are never bigger than 4.09 for any of the selections that I capture in the data. I do bet at bigger prices applying essentially the same methodology but I only record the bets not every player. 

Unless anyone makes a good case to the contrary I'm inclined to just reduce the edge from 8% to 6% and see how it goes. It will be quite easy to track the additional bets and see how they perform. They're probably selections that I look at where time allows and end up backing half of based on stats anyway! Anything that increases the ratio of automatic bets to subjective ones gets my vote!

[richard-westwood]

Just now, harry_rag said:

Yeah I focus on the top end of the market so my fair odds are never bigger than 4.09 for any of the selections that I capture in the data. I do bet at bigger prices applying essentially the same methodology but I only record the bets not every player. 

Unless anyone makes a good case to the contrary I'm inclined to just reduce the edge from 8% to 6% and see how it goes. It will be quite easy to track the additional bets and see how they perform. They're probably selections that I look at where time allows and end up backing half of based on stats anyway! Anything that increases the ratio of automatic bets to subjective ones gets my vote!

I think that's the grail .....if you can get to a point where the bets become automatic then it takes a lot of hassle out and betting becomes easier 

[MCLARKE]

I suppose the key is how you calculate the fair odds, in racing there are so many variables I find it almost impossible

[harry_rag]

1 minute ago, MCLARKE said:

I suppose the key is how you calculate the fair odds, in racing there are so many variables I find it almost impossible

With football it's "easy", relatively at least. As long as you can arrive at an expected number for team, match or player goals you can just apply the poisson distribution to it to calculate indicative odds for over/under/exactly any number of goals. My approach is reassuringly similar to that used by the firm that compiles the odds for most of the major firms whose prices are on Oddschecker. I'm more likely to win taking on the "maverick" firms who don't use their services than I am taking on their clients but I still manage to beat them once in a while!

I just take my expected numbers from the spread firm's prices and apply my calculation to that for fair odds. Might as well get some use out of the buggers even if I can't bet with them!

[Torque]

What are the average odds. A sample means nothing without the odds to go along with it.

[harry_rag]

1 hour ago, Torque said:

What are the average odds. A sample means nothing without the odds to go along with it.

Without wishing to channel my inner Robert De Niro I’ll assume that remark is addressed to me! I’ll check tomorrow but pretty sure it’s a sliver either side of 3.

[Zilzalian]

Me thinks some of these buggers are more interested in maths than the horses...🤣🤣😂

[harry_rag]

13 minutes ago, Zilzalian said:

Me thinks some of these buggers are more interested in maths than the horses...🤣🤣😂

Definitely true in my case! 

[Torque]

Betting is maths. Whether it's overt via spreadsheets or covert via 'feel' 'experience' or any other word to describe the innate ability to pick prices that represent value - betting is maths.

[Torque]

24 minutes ago, harry_rag said:

Without wishing to channel my inner Robert De Niro I’ll assume that remark is addressed to me! I’ll check tomorrow but pretty sure it’s a sliver either side of 3.

Not aimed at you in particular @harry_rag rather anyone that has the numbers. Using the average odds you mention and a 5% yield, you'd need a sample of about 4,500 bets to suggest strongly that luck wasn't at play. That is to say that if you had a yield of 5% after about 4,500 bets at average odds of 3, there'd be a less than 1 in a 100 chance that those results were down to luck.

[harry_rag]

6 minutes ago, Torque said:

Betting is maths.

Probably not going to be adopted by the bookies as their new advertising angle to get the hipsters opening accounts! 

Though there is that advert where the bloke in the Chinese with Cantona is wittering in about the height of the centre backs and betting on corners! 

[Torque]

1 minute ago, harry_rag said:

Probably not going to be adopted by the bookies as their new advertising angle to get the hipsters opening accounts! 

Though there is that advert where the bloke in the Chinese with Cantona is wittering in about the height of the centre backs and betting on corners! 

Definitely not. There's no glamour in the maths angle plus people love to think they know better than others and there's no better way to do that than to give betting a mystical sheen where there's a secret alchemy to success which only a select few know about and which can't be quantified.