Tosh

Venue:: FLEMINGTON :: Saturday 30th October, 2010

1st race down 8 to go

I started this thread because I went and stupidly posted on another thread what I
could do with 5K over 3 years.

Because it is unbelievable to 99%. I thought I would give it a go as per other thread

However will only post my lay bets as I don't want to give my prices away on my win bets.

Will explain a bit more later.

Fair play to you CM, I shall follow with interest. Just saw Maluckyday win race 2 at about 6.3 sp, does that make it a roughly break even race for you?

When you stipulate max odds 25.0, min 2.6, does that mean that your bets go in-play, as some of those horses were around 200.0 sp, or are they just a non-bet after the off?

Good luck for the day anyway, most who are brave enough to put their selections up are at the mercy of sod's law!

Trevh

After I set my max limit only the ones that get matched at my price or less qualify
So far 8 lays 7 good  showing a slight loss on my lays showing a profit on win bets as
mr chard won and I had it priced at 2.4 and it won at 5.3 SP (betfair)

You appear to be laying all of the 13 horses in race 4, although you'll probably only get matched on 6 or 7 of them as the other 6 or 7 have odds ranging from 36->140.

I must be having a blonde moment coz I'm not sure how that would be profitable unless one of the rank outsiders wins ?

What am I missing ?

Certainly looks like  a losing race for me
however it wont be a lot. Once I have priced my horses I stick to it .

All are bad value as I post except for the 5 and 6 tab which are over my true odds

All horses that do not have 25.00 next to them are my true price

Horses at 25.00 are not my true price but I use 25.00 to get a bit more on them
Most of the 25.00 are priced at a lot more than 25.00

Star witness I have priced at 12.00 but use 25 to get more on him.

I'm definitely not understanding something

13 horses, you price up one at 9, two at 13.5 and the rest at 25
and offer to lay up at to those prices.

If I could back all the horses at those prices for 1 point each, Worst case I lose 3 points on one horse, make half a point on two horses or win 13 points if one of the other ten horses wins.....
So unless I'm being really stupid ( quite likely ) I'd jump on all those prices as a backer surely ?

'scuse my ignorance, what am I missing ?

I'm amazed at the liquidity, I thought Oz and US races were poor but there was £570k matched on the last race. CM, I don't know if you remember, but you posted a selection process a few years ago, where you used a spreadsheet to determine the top 4 runners in each race, and further narrow them down using other rules, then all the other runners were lays. Do you use a different selection process now?

The 25.00 is not an arbitrary figure it is calculated by a formula
depending on how much a selection is below my calculated price at time of pricing.
Hope that makes sense.

In the above example Star Witness was calculated to start at below 6.00 My price being 12.00
With such an advantage I was prepared to have a bigger lay on it.

Was wrong with that one JUST!

Star Witness was calculated to start at below 6.00 My price being 12.00

But on your table you have it listed as 25 ?

Although I am a pom, the English seem to have lost their b a ll s, turned into real ****.
All I see on this forum is laying horses at low odds. Not much profit in that.

your definitely failing to explain why you listed star witness as a lay @ 25 in your table and then say you think should be 12's

Escapee

Because there was so much value on it increasing from 12.00 to 25.00 increases the multiplier
to get more laid on it.

nope, none the wiser.

Best of luck to you but what you're saying is beyond me at the moment.

Escapee, that was explained earlier if you look. CM, any chance of answering my question 7 posts up ^ please.

1.1m matched on race 5, 870k on the winning favourite. 3 good lay profits for you CM?

Trevh
Have changed my rules to a more valued approach. A bit pressed for time betting at the mo, will answer more questions
in detail later.

Ok, I'm off to bed as I'm upside down to your land of the sun. GL

Days summary ~ Just Lays

Start Bank =    5,000.00
In play bank =  1,000.00    
Gross Profit =    -117.94    
Total Commission =    8.44    
Total Net Profit =    -126.38

Total lays 39, 34 winning lays, 5 losing lays

Detail ~

Race	Tab	Horse	Lay Max Price	Liability        Result
1	3	CONDELAGO	18.5	113.5	7.37
1	4	ACE ACOSTA	10.5	108.5	12.48
1	5	GOLDSTONE	12.5	110.4	12.00
2	1	ALCOPOP	12.0	113.3	16.00
2	2	HERCULIAN PRINCE (NZ)	25.0	118.6	8.12
2	7	ONCE WERE WILD	25.0	118.6	8.47
2	11	CEDARBERG	17.5	116.5	7.51
2	12	MALUCKYDAY (NZ)	8.0	108.1	-108.10
3	2	DIVORCES	25.0	110.8	5.62
3	3	DIZLAGO	9.0	102.6	15.54
3	6	SASA	25.0	110.8	17.87
3	8	ZUTARA	17.5	108.8	8.05
4	1	STAR WITNESS	25.0	116.0	-116.00
4	2	PRESSDAY	25.0	116.0	20.71
4	3	BUFFERING	25.0	116.0	16.11
4	4	SOUL	9.0	107.4	21.96
4	9	SISTINE ANGEL	13.5	111.8	11.52
4	10	SOLAR CHARGED	25.0	116.0	24.16
5	1	SHOCKING	25.0	113.5	14.18
5	7	DESCARADO (NZ)	17.5	111.4	8.02
5	8	SHOOT OUT	25.0	113.5	8.16
6	3	LION TAMER (NZ)	16.5	114.3	-114.30
6	9	DIGITAL FORTRESS (NZ)	11.0	110.6	17.83
6	11	ALMINDORO	14.5	113.3	9.00
7	1	HOT DANISH	25.0	106.7	7.62
7	3	CULMINATE (NZ)	25.0	106.7	8.89
7	6	SACRED CHOICE	25.0	106.7	-106.70
8	1	BLACK PIRANHA	24.0	96.2	5.06
8	5	SWIFT ALLIANCE	12.0	92.0	8.76
8	6	SHELLSCRAPE	10.5	90.8	11.35
8	7	BEADED	15.5	93.9	7.82
8	10	RESPONSE	23.0	96.0	5.05
8	12	DUBLEANNY	25.0	96.3	21.16
9	1	DAO DAO	25.0	102.8	13.04
9	4	JUNGLE RULER	25.0	102.8	10.99
9	6	POOR JUDGE	18.5	101.3	-101.30
9	8	ROCKPECKER	17.5	101.0	10.38
9	10	DEMERIT	25.0	102.8	33.16
9	11	PRINCE BRAEMAN	13.5	99.2	14.50

After reviewing last Saturdays racing and selections as I always do, I have to say
I could find no fault with the selections. Just one of those days that didn't quite pan out.

No real harm done though, laid 5 selections that won from 9 races, SP on those were ~
6.28SP for Maluckyday priced at 8.00
6.79SP for Star Witness priced at 12.00 used calculated max limit price of 25.00
13.17SP for Lion Tamer priced at 16.5
25.00SP for Sacred Choice priced at 33.00 used calculated max limit price of 25.00
16.50SP for Poor Judge priced at 18.50

A 2.53% loss of the bank shows no real harm done. A usual day is 2 to 3 lays winning
which would make a nice profit. Will continue tomorrow ~ Day 2

To save scrolling will begin a new thread for each days racing.

To get how much to stake on a 1K bank (my method). Bear with me for a while.

First, rating a race and getting its winning chance, second, converting to a price in decimal, third
comparing prices to see which is value and finally how much to risk as a % of the bank on each selection


This race example has 8 starters, for typing convenience I will use saddle numbers to name each runner.

Horse	Weight	Limit	Base	P&B	Rating	Diff	Decimal
1	58	51	67	-2	58	2.5	0.5
2	57	51	66	0.5	60.5	0	1
3	56	51	66	-4	57	3.5	0.33
4	54.5	51	59	-3.5	52	8.5	0.03
5	52	51	60	-2.5	56.5	4	0.25
6	51.5	51	60	0	59.5	1	0.8
7	51.5	51	60	-7	52.5	8	0.03
8	51	51	61.5	-1	60.5	0	1

Col. A horse, Col.B weight allotted by handicapper, Col. C  limit weight that handicapper can give, Col D Base run ~ this is from past ratings its best performance in a race, Col E Penalties and bonus's , these are derived from tables on class, distance, form, age improvement, consistency and others  you may wish to include, Col F = final rating which is its base run plus  penalties and bonuses minus the difference from weight allotted to the limit weight, Col G is the difference in weight from the highest rating weight (0), Col H is the decimal derived from a table of advantages to chances.

Now pricing and how much to bet.

Horse	Decimal	Sum	Book	%	Price
1	0.5	3.94	90%	11.4213198	8.76
2	1	3.94	90%	22.84263959	4.38
3	0.33	3.94	90%	7.538071066	13.27
4	0.03	3.94	90%	0.685279188	145.93
5	0.25	3.94	90%	5.710659898	17.51
6	0.8	3.94	90%	18.27411168	5.47
7	0.03	3.94	90%	0.685279188	145.93
8	1	3.94	90%	22.84263959	4.38

We divide  the decimal chance by the sum of all chances then multiply by a book you wish to have, in this  example a book of 90% to cover commission and variances. Now we have a percentage chance of each horse winning taking into account the built in 10% margin for comm. and variance to ensure a profit. To get the price we divide 1 by its winning %.( 1/11.4% for horse number 1 giving us a decimal price of  ~ 8.76)

We then compare our prices to betfairs price on offer ~

Our price ... Betfairs price
1. 8.76...........11.00
2. 4.38...........5.5
3. 13.2...........4.60
4. 145.93.......88.00
5. 17.50.........15.50
6. 5.47...........5.10
7. 145.93........120.00
8. 4.38............3.70

At this stage of the market we see that Betfairs book is at 1.016 We back them to get a return of a % of our bank. In this case for ease of calculating we will use 10% of our bank which is 100.00. We multiply the 100.00 by our chance of each selection winning, the higher the chance of it winning the more we have on it.

Our chance of winning on horse 1 is 11.4% so the bet would be 11.4%*100 = 11.40 at betfairs offer of 11.00.  To win  114.00 minus our bet of 22.8 on horse 2. if it wins we win 91.2 on the race.
Our chance of winning on horse 2 is 22.8% so the bet would be 22.8%*100 = 22.80 at betfairs offer of 5.5.  To win 102.6 minus our bet of 11.4 on horse 1. if it wins a profit of  91.2 on the race.

It is basically overlay betting, the bigger the margin between your price and Betfairs price, the bigger the profit.

In my tosh post for lays I would show these at these max set odds ~
Number 3, 25.00
Number 4, 25.00
Number 5, 17.50
Number 7, 25.00

Numbers 6 and 8 would be no bets not enough value to either lay or bet on.

Compound Magic
This appears to be a prime example of how some body can post up pretty much exactly how he approaches things, but without giving away his essential edge.
An incredibly well presented post.
And yet no comments and replies, to date at least.
Defies belief.
I would think and hope that any budding HR bettor out there ( which I'm not ) would be onto all this in a flash, with detailed technical HR  questions, probably a lot of which you might not wish to answer understandably.
But the basic maths is pretty clear and shows how systematically and objectively you are going about things.
The main thing I didn't really comprehend is the " decimal" column in chart 1.
Could you explain that a bit more.? It is a rather key element.

I concur with fafh, an outstanding post.


I remember asking some questions  previously in this thread and failing miserably to comprehend the answers so I'm reluctant to further clutter up this thread with questions.


But as I'm here and I find myself in the 'reply box' I'll tentatively fire one off and hope.

As you do Aus racing and I do UK stuff I'm not sure the handicapper works in the same way.
I've never really understood how UK handicapping works, a horse might go from a 'rating' of 68 to say 70 in its next race and yet the weight on the declared card might drop from 9st8lb to 9st4lb. Whats going on here ?

You talk of 'Limit weight' I don't know what this is could you elucidate ?


In fact if anyone wants to chime in with an explanation of how the UK racing handicapping/rating/weight system works I'm sure I'm not the only one who would be greatful.

Thanks in advance.

Escapee only to happy to answer your query

Each race here the handicapper has a maximum and a minimum weight that he has to adhere to.

He can give less than the maximum weight and can give more than the minimum weight but can't give
more than the maximum or less than the minimum. The minimum is the limit weight, not what the lowest
weighted horse might carry.

The base run rating is my own rating. I allot each horse at the end of each race what it should have carried
to win, this becomes its last rating. If the rating is its best rating so far it becomes the base rating.

If the rating is less than its base rating it usually signifies that the horse is on a decline and the next time
I do future ratings this is a penalty that will come of its base rating.

It's unfortunate that here we are all metric so it's a little difficult to relate to UK.


FAFH will get back to you on the decimals when I have a little more time.

Escapee I get most of my form from risa. This URL if you open it will show race 2 of today.
For Four-Years-Old and Upwards - 0 - 95. 53 kg min. Apprentice Can't Claim
In this case it shows the limit weight of 53KG


http://materials.risa.com.au/PUBLISHING/DOWNLOADS/FREEFORM/RD0611FLEM_A.asp?racenum=02&StateDesc=VIC

Cheers

Escapee

You might find this helpful, 

http://www.britishhorseracing.com/inside_horseracing/about/whatwedo/handicapping.asp

in particular the Guide to Handicapping which you'll see if you scroll down the page.

Thanks Compound & Top2

FINE AS FROG HAIR ~ 

Decimal explanation

It is obvious that if a horse is penalized with extra weight it will slow it down. How much is the eternal question.

The most common theory is that an increase of 1.5Kgs (3.3lbs) will slow a horse down by 1 length over all distances and conditions.

This is disputed by many, myself included. I compensate for this when I do the penalties and bonus's.
A horse when reaching a critical weight will be more affected by a weight increase than if the weight increase is well below the critical point. Another factor is the class of the horse, a group 1 or 2 horse usually is less affected by a weight increase than the inferior brigade. Lets say these factors are compensated for before we consult our table of advantages to chances.

My table ~
Column "A" Expected margins between top rated horse and other horses expressed as weight.
Column "B" Number of times top rated horse wins
Column "C" Corresponding number of times other horses win
Column "D" Converted to decimal

"A"	"B"	"C"	"D"
0.0kg	100	100	1.00
0.5kg	100	90	0.90
1.0kg	100	80	0.80
1.5kg	100	67	0.67
2.0kg	100	57	0.57
2.5kg	100	50	0.50
3.0kg	100	40	0.40
3.5kg	100	33	0.33
4.0kg	100	25	0.25
4.5kg	100	20	0.20
5.0kg	100	14	0.14
5.5kg	100	11	0.11
6.0kg	100	8	0.08
6.5kg	100	6	0.06
7.0kg	100	5	0.05
7.5kg	100	4	0.04
8.0kg	100	3	0.03
8.5kg	100	3	0.03
9.0kg	100	2	0.02
9.5kg	100	1	0.01
10.0kg	100	1	0.01

This means if your top rated horse (a)in a 2 horse race had an advantage of 2.5 kgs over the other horse(b) you look up the table and see that the likelihood of horse (a )winning is 100 and the likelihood of horse (b) is 50. Add those figures together and you get 150 so from 150 events the likelihood is that Horse (a) will win 100 of those races and that horse(b) will win 50 half as many times which is 50% which is decimal 0.5

By adding all the decimals you can compute the winning chance of each horse and derive a price.

A 3 horse race A =1.00 (top rated):  B =.67 at a disadvantage of 1.5Kgs:  C =.25 at a disadvantage of 4kgs add the decimals up = 1.92 from 192 races at these same advantages ~ A would win 100, B 67 and C 25

To get the price for the 3 horse race we divide each horses decimal by the sum of all decimals  ~
Horse A = 1/1.92 = 0.5208 : B 0.67/1.92 = 0.349 : C  0.25/1.92 = 0.1302 that has given us so far the likelihood in % terms of each horses chance,  then we divide  1 by those chances to convert to a price ~
1/.5208 = 1.9201 for horse A, 1/.349 = 2.8653 for horse B, 1/.1302 = 7.6805 for horse C.
All the chances must make a book of 1 decimal.

If your ratings and calculations are right and you backed the three horses at those prices ~
A wins 100 races = 100*1.9201 = 192.01
B wins  67 races = 67 *2.8653 = 191.9751
C wins  25 races = 25*7.6805 = 192.0125


In this way you can determine the price on any amount of starters in a race.

Is this type of analysis something you have struck upon yourself.?
Or is it all accepted as a fairly not well known, correct, if not necessarily practiced, approach to things.?
Of course, always give or take the inherent,subjective decision making that will presumably make all those following this type of approach come up with ( slightly) varying results ?
Remember I don't bet on horses, never have, and so maybe my question is perhaps totally naive.

FINE AS FROG HAIR ~ 
Fairly well known and documented practice
The real difference is how each person applies the penalties and bonuses which will affect the winning chances
derived at and therefore the prices.

Not much to worry about being copied as 99% could not be bothered with the amount of effort required, they would
rather just guess or use a simple subjective system.
I feel rather dumb making my original post on your thread knowing full well what I outlined would not be thought possible.

Then my ego said "I'll prove it,I'll show em" big mistake or what.

From now on I will stick to trying to answer excel queries. No one really wants to know if what you do is a success
or not unless it benefits them. Lets face it, most won't lay the longer prices even though they have a bigger edge
long term than the shorter prices, that is why I was prepared to post the lay side of my opp.

If I posted my win bets and they saw the profits from those some would start backing them as they represent lower
liabilities, that would then effect prices and my profits.

Cheers, party on.

Compound Magic

https://spreadsheets.google.com/ccc?key=0AjriKHhwpRQQdFFLemtUUGk1VUlVaFBmWjZ2YVh2UHc&authkey=COiIzZUP&hl=en#gid=0

Our route may be different but the end product is similar don't you think?

The race is the November Handicap, the last big handicap of the UK Flat Turf season, which took place at Doncaster on 06/11/2010.  The winner, incidentally, was Times Up.

Compound
As a general rule, what type of races do you not even bother to spend a second looking at ?

top2rated

Very interesting, pricing by rank. I see you are a speed rater. Tried that with no success maybe my input
information on track surfaces and the going was not accurate enough.

Of the three types of rating systems that I have tried, most success for me has come with the limit weight method.
The popular American rating method of giving each horse a par value certainly has merit but is a bit more difficult
to price.
It seems common in UK to rate by speed but as said I could not come to grips with it.

I think any objective method using known parameters is the best way to profit on horse racing if you are a position
player.

It is interesting how other punters decide what to bet on and how much to stake. I agree with lori on "Kelly criterion"
but when backing and laying multiple selections in the same race have not been able to create a formula to cover all.

You once asked if I would be willing to help with an excel spreadsheet and I never gave you a proper answer, so yes to that.

Hope you are making a quid or two. Putting in the effort that you do, deserves reward.

What does 'P?' indicate in your spreadsheet Top2?

FINE AS FROG HAIR

As a rule I stick to races where the total prize money is 70k or greater.

Several reasons.
1. Being liquidity, these races almost always have 300k or more bet on them.
2. Acceptance fees to these races are quite expensive so majority are triers.
3. 4th place getter usually can win more prize money in these events than the provincials pay for the win.
4. Stewards are more on the ball and are less corruptible, less chance of race fixing.
5. City tracks are a lot better maintained, they are not as tight and small as country tracks which means
horses are not inconvenienced as much in the running.
6. They are mostly classier runners, more consistent in their form and make it easier to handicap.

Lastly, I don't have time to do more than two meets a week. After the races finish there is about 4 hours work
rating all horses on how they ran and before the next meet when acceptances and weights are known about 6 hours
work working out penalties and pricing etc..

I bet on two year old races that have not raced before even though there is no ratings to be done. There are some
direct pointers to how these will go but that's a topic for another time.

cheers.

Compound Magic

From where I'm sitting, an Excel query is never far away, so I'm sure I'll be taking you up on your offer.

tobermory

P? column

The number 1 in this column identifies a 'progressive' horse.  To qualify for the 'progressive' tag the horse must have recorded its best form or best speed figure last time out.

thanks

ttt for BlindProbe

Thanks for bumping this Tobermory, very much appreciated.