Notifications
Clear all

UNIBOOST  

Page 2 / 4
  RSS

Avatar
(@michael-clarke)
Noble Member Customer
Joined: 6 years ago
Posts: 1209
25/10/2020 1:03 pm  

Both selections lost yesterday. Total loss now £433.19.

COURSE TIME  SELECTION ODDS 
HEXH 2.05 RAYNAS WORLD 0.44
WINC 1.42 FAIRWAY FREDDY 2.25

ReplyQuote
Avatar
(@michael-clarke)
Noble Member Customer
Joined: 6 years ago
Posts: 1209
25/10/2020 6:10 pm  

Another odds on winner and a loser today for a loss of £11.11. Total loss now £444.30.

It is fair to say that so far this been very disappointing with 12 consecutive losing days and a ROI of -65%.

There have been 6 winners from 33 selections, expected winners is 13 giving an AE of 0.48.

However I am still confident that the logic remains sound and the results will revert back to the mean.

Based on the over-rounds I would expect to lose about 6% based on the early (un-boosted) price.

However if I then apply the boosted price and apply BOG then I would expect that to generate a ROI of 3%.

In reality I would expect to make a little more than that because of the impact of the favourite / long shot bias and also the additional filters I am applying (although it could be that the filters could potentially make it worse, perhaps I should just stick to the market).

If anybody can point out any flaws in my logic then please let me know as doubts do start to creep in during losing runs like this.

 

 

 


ReplyQuote
Avatar
(@andrewp)
Honorable Member
Joined: 2 years ago
Posts: 540
26/10/2020 12:09 am  

Hi Michael,

This is currently showing quite a high variance from both favourite strike rate and actual/expected winners so I too would expect both to increase over time.

How much does Uniboost boost the price? What price is boosted, SP, current offered price? This is key as the other rules would only get you close to breakeven at best I would imagine.

I make expected winners 13.87 to Betfair SP and 13.9 to your prices shown. Not allowing for commission or rule 4 deductions. If prices shown are the boosted ones, they are not showing an edge over BSP (breakeven) so far. It would be interesting to see if the two measures diverge as things improve.

Odds-on at BSP show a better A/E and more difference.

Basic rules 1 and 2 are fine. I'm not sure 3 and 4 add much

1. Favourite and has best odds or joint best odds on odds checker.

This should ensure they have a good chance of winning and bear as little of the overround as possible.

2. The best book on odds checker is less than 112.

This still leaves a share of up to 11% o/round. O/round per runner is a better measure. A bookmaker loads at least 1 to 1.5% on rank outsiders (100+ on Betfair) so number of runners and shape of race are important. This is why the odds-on are looking better.  

3. LTO placed in 1st 6.

Doesn't add much. Perhaps a quality measure such as minimum lengths beaten (0.2/f?) LTO would help.

4. Class of race less than 6.

Only rules out the really poor unpredictable stuff. Consider less than 5 or a prize money minimum. Also some race types are better for favourites (see Adrian Massey's web site). A 'no unexposed runners in race' rule might help.


ReplyQuote
Avatar
(@andrewp)
Honorable Member
Joined: 2 years ago
Posts: 540
26/10/2020 12:14 am  

Odds-on at BSP show a better A/E and more difference.

By which I meant your prices show a 2.7% advantage over BSP for those horses that were odds-on at BSP.


ReplyQuote
Avatar
(@andrewp)
Honorable Member
Joined: 2 years ago
Posts: 540
26/10/2020 12:18 am  

It goes without saying that my view is you should be scaling your stakes to the odds. :0)


ReplyQuote
Avatar
(@michael-clarke)
Noble Member Customer
Joined: 6 years ago
Posts: 1209
26/10/2020 9:04 pm  

Thanks Andrew.

To explain my logic I will use as an example the 1st bet on here, FARASI LANE.

The early price was 7/2, the boosted price was 4/1 and the best book was 105.

To calculate the true chance I have converted the early price odds to decimal and then divided by the best book, i.e. 1/(7/2+1)=0.222/1.05=0.212.

I have then applied the BOG price to this to calculate an expected value.

In this case the SP was 9/4 so BOG was 4/1 and the expected value was 0.212*4 - (1-.212) = 0.848 - 0.788 = 0.06.

The price boost is on the offered price and the % benefit varies, e.g. the boost for 2/1 is to 85/40 (6.3%), for 7/2 the boost is 4/1 (14.3%).

In theory the best bets should be when the boost increase % is at it's highest and the over-round is at it's lowest.

In this instance because there are only 3 offers per day and a maximum stake of £20 the logic of scaling stakes probably doesn't apply.


ReplyQuote
Avatar
(@michael-clarke)
Noble Member Customer
Joined: 6 years ago
Posts: 1209
27/10/2020 11:08 am  
COURSE TIME  SELECTION ODDS 
BANG 2.00 Good Boy Bobby 0.73
CATT 2.10 LORD OBERON 2.75
SOUT 6.20 Xcelente (5) 1.00

ReplyQuote
Avatar
(@andrewp)
Honorable Member
Joined: 2 years ago
Posts: 540
27/10/2020 6:09 pm  

The early price was 7/2, the boosted price was 4/1 and the best book was 105.

To calculate the true chance I have converted the early price odds to decimal and then divided by the best book, i.e. 1/(7/2+1)=0.222/1.05=0.212.

That's a good a way as any to estimate horses' 'true' chance from bookmaker prices Michael. If you've a sizeable record of these final figures (e.g. 0.212 above) you should have a pretty good idea how accurate the method is.

A/E will tell you. If you take just those that make their price in the market but use the calculated 'true %' as the 'expected' rather than the higher market price % it will measure the accuracy of your calculated prices.

I didn't realise you get BOG as well as 'boost'. Effectively, you have five prices you can record and compare, early, 'true', boosted, BOG, SP, BSP. It will be interesting to see how much effect is due to BOG and how much is due to boost at the end of the day. I suspect BOG may be the bigger player.

The boosts are reasonable though. What are the boosts on the odds-on, one bookmaker 'tick' higher (e.g. 1/5 to 2/9)?

I would hope that the £20 limit means they won't close the Uniboost and BOG offers to you.


ReplyQuote
Avatar
(@michael-clarke)
Noble Member Customer
Joined: 6 years ago
Posts: 1209
28/10/2020 12:28 pm  

A winner and 2 losers yesterday for a loss of £24. Total loss £457.19.

COURSE TIME  SELECTION ODDS  STAKE 
FAKE 1.02 PLENTY IN THE TANK 1.63 4.00
KEMP 7.15 INDIGO TIMES 2.00 4.00
KEMP 7.45 ALMIGHWAR 1.63 4.00

ReplyQuote
Avatar
(@michael-clarke)
Noble Member Customer
Joined: 6 years ago
Posts: 1209
29/10/2020 9:39 am  

Another 3 losers yesterday. Total loss now £517.19.

COURSE TIME  SELECTION ODDS 
NEWT 1.28 MEMPHIS BELL 2.00
NEWT 3.28 MOMELLA 2.13

ReplyQuote
Avatar
(@michael-clarke)
Noble Member Customer
Joined: 6 years ago
Posts: 1209
29/10/2020 5:55 pm  

@andrewp

Posted by: @andrewp

I didn't realise you get BOG as well as 'boost'. Effectively, you have five prices you can record and compare, early, 'true', boosted, BOG, SP, BSP. It will be interesting to see how much effect is due to BOG and how much is due to boost at the end of the day. I suspect BOG may be the bigger player.

BOG certainly makes a big difference. At the early prices the expected loss would be 6.4%, at boosted prices it would be 1.3% and at BOG it is a profit of 3.1%. At SP it would be a loss of 4.9%.

Posted by: @andrewp

The boosts are reasonable though. What are the boosts on the odds-on, one bookmaker 'tick' higher (e.g. 1/5 to 2/9)?

Yes, this is basically how it works.

Posted by: @andrewp

I would hope that the £20 limit means they won't close the Uniboost and BOG offers to you.

Given my current record I don't think there is any danger of them closing the offer anytime soon !


ReplyQuote
Avatar
(@michael-clarke)
Noble Member Customer
Joined: 6 years ago
Posts: 1209
29/10/2020 5:58 pm  
Posted by: @andrewp

. LTO placed in 1st 6.

Doesn't add much. Perhaps a quality measure such as minimum lengths beaten (0.2/f?) LTO would help.

I think I'll take this subjectivity out of the equation and just concentrate on those selections where the boost % increase is greater than the over-round.


ReplyQuote
Avatar
(@michael-clarke)
Noble Member Customer
Joined: 6 years ago
Posts: 1209
30/10/2020 12:07 pm  

A winner and a loser yesterday for a profit of £20. Total loss now £497.19.

COURSE TIME  SELECTION ODDS 
WETH 12.25 APPLE ROCK 1.00
NEWM 1.10 AINSDALE 3.00
NEWM 3.30 WATERS EDGE 2.00

ReplyQuote
Avatar
(@michael-clarke)
Noble Member Customer
Joined: 6 years ago
Posts: 1209
31/10/2020 11:27 am  

A winner and 2 losers yesterday for a loss of £10. Total loss now £507.19.

NEWM 2.40 HEADMAN 1.38
NEWM 3.50 ASPIRATION 3.33
ASCO 1.55 IBLEO 2.75

 


ReplyQuote
Avatar
(@michael-clarke)
Noble Member Customer
Joined: 6 years ago
Posts: 1209
01/11/2020 11:54 am  

3 losers yesterday. Total loss now £567.19.

COURSE TIME  SELECTION ODDS 
HUNT 1.18 PERFECT MYTH 1.50
CORK 2.15 RAGNAR LODBROK 1.63
CARL 2.40 IMPERIAL AURA 1.88
CARL 2.05 INSTANT REPLY 5.00

ReplyQuote
Page 2 / 4
Share:
Back to top button
Close