Forums

We want to share an important update with our Betfair Community users.

Over the coming months, the Betfair Community platform will be winding down. The final active day for the platform will be December 21st. From December 22nd, users will no longer be able to access the platform.

Your discussions, insights, engagement, and, most importantly, feedback have helped shape the platform and guided innovation. Your feedback continues to be important to us, and we will continue to listen via our dedicated feedback channels, as well as through product testing and innovations.

To share your feedback and experience, visit: https://support.betfair.com/app/home

General Betting

There is currently 1 person viewing this thread.
SHAPESHIFTER
04 Jul 14 14:00
Joined:
Date Joined: 04 Apr 03
| Topic/replies: 3,204 | Blogger: SHAPESHIFTER's blog
Just would like people's perception on the calculation of value.

If a horse is priced up at 3.20 and I see it at 3.00, I would see this as 10% value.

How I got this.

3.20 returns 2.20
I have it at 3.00 which is 2.00

I see this as 10%.

I am aware of someone that calculates this as:

(3.20-1)/(3.00-1)= 110% and says that the 3.20 is 110% value.



If I see arsenal at 1.50 and the match is priced up at 1.70,

They say that this is 140% value.

I see it at 40%

Have my decades of betting and calcuting value been wrong?

Cheers.

Post your reply

Text Format: Table: Smilies:
Forum does not support HTML
Insert Photo
Cancel
sort by:
Show
per page
Replies: 39
By:
DFCIRONMAN
When: 04 Jul 14 14:24
It is the 3.00 calculation is the "interesting" area SHAPESHIFTER . If you get this "wrong", then the amount of VALUE will be "wrong". The 10% calculation part seems ok to me.

Probably no matter how you calculate the priced up part, it is a subjective calculation..............and therefore open to umpteen views thereon.

Interested in others method of arriving at the 3.00 part of pricing upTongue Out.
By:
SHAPESHIFTER
When: 04 Jul 14 14:29
This is someone posting the value that their ratings give.

example:

6.6 BFSP
priced up by us @ 4.80 is 147% value

I get 47%

Look at it from this angle:

If for a pound, you 3.80
Somewhere else, you get 5.60

This is not 147% better
It is 47% better or, in the case of betting, edge.
By:
DFCIRONMAN
When: 04 Jul 14 14:38
47 % appears to be a reasonable way to look at the VALUE......but how do they price up their 4.80?Tongue Out
By:
DFCIRONMAN
When: 04 Jul 14 14:45
I use CRUDE method.......using RANK position. EG

RATINGS for a race on 1st JUNE....just picked at random.

EPSOM DOWNS

14:15

1,Pekan Star   5.5 2ND F   -£100                           IR LOW  5.60           NPNL        NB             -100
2,Ottoman Empire                  2ND 20-1  BFSP 28        IR LOW 0.33
3,Gatewood                                 1ST 9-4 FAV BFSP 2.70 IR LOW 0.00           + 0.4275
4,Cry Fury         8  4TH F
5,Stand To Reason   N/R
6,Rumh            14 5TH F
7,Creme Anglaise  20
8,Haylaman        25
==============================================================Ex £56.20 CSF £58.20 TRICAST £389.93 TRIFECTA £1738.60
9,Ramona Chase    20                                     IR LOW    14             
10,Pivotman        16                                    IR LOW   9.5
11,Borug                                  3RD 10-1       IR LOW    1
12,Black Spirit    20                                    IR LOW  5
13,Licence To Till 33
14,Resurge           6 3RD F



It is a 13 runner race......

The fav won, as it happens, and its SP odds were 2.25 . AS it was RANKED 3 Tp Rated I take this as a 2-1 shot. Therefore 2.25 is VALUE.


Pretty CRUDE but "works" consistently well enough as RATINGS are up to reasonable standard.
By:
DFCIRONMAN
When: 04 Jul 14 14:46
Anything above 2.25 is EXTRA VALUE and should cover commission on here.
By:
DFCIRONMAN
When: 04 Jul 14 14:48
I believe it is better to use bookmakers SP as the guideline, as they are less likely to be "wrong" on their pricing than using BFSP, which can vary wildly.
By:
SHAPESHIFTER
When: 04 Jul 14 14:55

Jul 4, 2014 -- 8:38AM, DFCIRONMAN wrote:


47 % appears to be a reasonable way to look at the VALUE......but how do they price up their 4.80?


It is their ratings and then they price up the field.

By:
DFCIRONMAN
When: 04 Jul 14 14:55
RESURGE ....RANKED 13 ( due to non-runner0 is a 12-1 shot based on my CRUDE way of arriving at price.....So at 6-1SP it is clearly a VALUE lay.

Statisticians will have a far more accurate method of arriving at "price", but the CRUDE method appears to work well for my purposes.
By:
the man
When: 04 Jul 14 14:55
This is my perception:

If a horse is priced at 3.2 it has a 31.25% chance of winning if odds are correct

If I price the horse at 3 then I think its got a 33.33% chance of winning

therefore I think Ive got a 2.08% edge
By:
SHAPESHIFTER
When: 04 Jul 14 14:57
the man, I believe you are right.  This is 'edge'

So are their ratings giving 'Edge'?

And if so:

edge is not value since value is against the investment.

Correct?
By:
the man
When: 04 Jul 14 15:00
You may well be right, personally I'm only interested in my edge, I don't think in terms of % value, it's not a figure that has any bearing on my staking
By:
SHAPESHIFTER
When: 04 Jul 14 16:45
I think is more that you 'find value' once it has edge. 

Calculating 'value' would come down to how well your edge translated into value.

And this would ultimately be ROI.
By:
kenilworth
When: 04 Jul 14 16:59
Backing 1/1 chances at 2.20 is a 20% margin on one bet
BUT 1/1 chances only win 50% of the time, therefore two 2.20
bets, will give you 2.20 return, for an outlay of 2.00,
which is 10% profit.
By:
SHAPESHIFTER
When: 04 Jul 14 17:14
kenilworth.

You are calculating it as a series of bets rather than the individual bet.

If I bet on one 1/1 bet a month, it has no relevance to the other bets before it or after it.
By:
kenilworth
When: 04 Jul 14 17:30
SS, I agree. What I posted was another poster's
view last year, I can't remember who it was, but
at the end of the day the money in your pocket is
the same! The point is the overall profit on two
bets is .20 for an outlay of 2.00 and that is 10%.
We can call it what we like, but the money is the
same.
By:
SHAPESHIFTER
When: 04 Jul 14 17:50
Yes, talking horse racing.

Value means comparison.

For me, when I find a 'value bet', I haven't put a 'figure' to it.  Simply, as a poster has it, I look for edge thus making it 'value'.

I just haven't thought about it for a long time.

But when i see someone posting "209% value", I thought about my own betting and though, yes, occassionally I will have priced up at, say, 5/1 and the market is at 9/1, I usually only find 2 or 3 bets a meeting that I feel the price is 'value'.

As well, because I price up a horse at 5/1, this price never existed on the market so comparing what I had to 9/1 isn't the correct approach.

What would be correct is, as the man posted above, to calculate edge which is against the percentage rather than the odds of success.
By:
SHAPESHIFTER
When: 04 Jul 14 17:51
...my head hurts.....Sad

And it has been a shtty week of betting on the horses.  My worst week of the year, possibly.
By:
kenilworth
When: 04 Jul 14 17:52
I think we are discussing getting 6/5 about 1/1 chances,
and what the percentage profit is.
By:
kenilworth
When: 04 Jul 14 19:36
SS I thought we were discussing percentages!
By:
DFCIRONMAN
When: 04 Jul 14 20:05
SHAPE - I understand you have a RATINGS SYSTEM , so it is likely you believe, and proven over a large number of results, that your ratings generally indicate VALUE, whether to BACK or LAY.

So trust your system of ratings and the RANK POSITION as the guide on what the odds should be, then compare against SP to see whether VALUE indicated or not.

Generally you will be faced with a number of horses that are VALUE, and this is where you are then faced with what one or more to back or lay.

Bookmaker SP is probably a useful factor to use on say the TOP 3 in rank position. The one that is the lowest SP odds one in T3 is the one in T3 that has the market supporting it and this ,combined with your excellent RATINGS having it in T3 rated, might be one way to select 1 horse to back.


If the RATINGS are any good, then the IR lowest odds should indicate how good the ratings are......as well as number winners obtained this way etc etc

GL SS
By:
U.A.
When: 04 Jul 14 20:58
If something is priced at 3.0 and I believe it should be 3.2 then i believe it's value is 6.67%.
By:
kenilworth
When: 04 Jul 14 22:23
U.A. Should that not be the other way round,
that is priced at 3.2 and your price being
3.0?
By:
U.A.
When: 04 Jul 14 22:43
oops yes sorry
By:
SHAPESHIFTER
When: 05 Jul 14 11:06
DFC....I do have my own ratings but haven't 'calculated' value for years (had to dust off a book).  Essentially for me, value is once I have priced up, if the odds are better than expected, time to consider. 

The only stats I have focused on recently is where I make/lose money (i.e. horse racing, 5 Furlongs have been terrible for two months while 7 to 12 furlongs are doing grea) and, in general, which type of races where my ratings perform better (track, going, distance, etc).

This is mainly to focus my time, prioritize my research.  Come Monday, after a terrible two weeks, will focus tighter until things turn around.

This all came from someone on twitter that offers a ratings service.

ie. 60% of their top 3 win.  A few weeks ago, when they put up the stats for the day, I realized that out of the 30+ or so races that day, over half had 4,5, or 6 runners.  So a bit like shooting fish in a barrel.

As one who strives to find 10, 15 and the occassional "I can't believe my eyes" monsterous edge, seeing someone on a daily basis 'advertise'/'market' 150, 200% value, I questioned how I was calculating since I hadn't thought about it in ages.
By:
SHAPESHIFTER
When: 05 Jul 14 12:09
So, as I said, this was all a bit of a review in my thought processes but let me summarize and see if we have this right.

Bear in mind, it is Saturday after Friday night so perhaps this could be edited but would appreciate your views if this sounds right.


Calculating Edge

This isn't done between prices but, as reminded in the thread, between your perceived % chance of success and implied  % based on the price on offer

So is 'egde' value?

Once the edge is calculated, if it is positive, this is 'value'.

From there, you can decide how much 'edge' constitutes 'value'.  i.e. you might see anything above 5% edge as 'value' while others might feel that anything over 10% is.  So 'value' is more like a line in the sand.

Can you calculate value between a rating and an available price?

I'd say no since one is 'perceived' (the rating) and one is 'actual' (the available price). Sort of like comparing the price of a litre of apple juice to the price of a kilogram of apples.

You can, calculate value between two existing prices.  i.e. two bookmakers or, as betfair do, between SP and BFSP


---------

a bit of repetition now but, in summary:

If one was to quantify on how well their ratings are doing, the clearest is 'edge'

From there, it becomes more individual to decide if this is 'value'.

'Value' is up to individuals who draw their own line.

And the only time you can really calculate 'value' is between two existing prices.
By:
Lex
When: 05 Jul 14 12:10
if you raised something 100% then it stays as is. so 110% is the same as 10%;   90% is the same as -10%

Value is the difference between the long term average price at the start of an event divided by the price you bet the event at.
Events must be derived from a ridgid set of rules so that the characteristics of each event are the same.


apparently
By:
SHAPESHIFTER
When: 05 Jul 14 12:25
Lex, I agree with you.

I think the other point is:

Punters look for 'value' when shopping for prices, whether it be:
- between their perceived probability and the implied probability on offer

or

- between the prices on offer once they move to make a bet

For 'ratings', punters look for 'the edge' they offer.

In my case, the past week, when doing a post mortem, the ratings actually offered edge.

It's my interpretation that burnt a hole in my profits this year Sad
By:
kenilworth
When: 05 Jul 14 12:54
'Value' exists when the odds on offer are/is greater
than the chance of your bet being successful. It's as
simple as that.
By:
Lex
When: 05 Jul 14 13:21
Your Majesty

~and that illustrates why so many fail to make money betting. They can't feckin' read.... Laugh
By:
Lex
When: 05 Jul 14 13:25
actually   "don't"   is a kinder word than   "can't"
By:
SHAPESHIFTER
When: 05 Jul 14 13:56

Jul 5, 2014 -- 7:56AM, King of Dubai wrote:


Question is though Lex...Can I take superfluous jargon to the bank???


Twitter and facebook did Cool

By:
SHAPESHIFTER
When: 05 Jul 14 14:00

Jul 5, 2014 -- 6:54AM, kenilworth wrote:


'Value' exists when the odds on offer are/is greaterthan the chance of your bet being successful. It's assimple as that.


So, kenilworth, doe my 'summary' make sense?

Value can only be given a figure between two 'existing' prices (i.e. bookmaker comparision)

and

A price only becomes value when the 'perceived probability' of a rating is greater than the 'implied probability' of the price at which point you have two figures to calculate the value.

By:
SHAPESHIFTER
When: 05 Jul 14 14:23
I'm getting more into comparing my perception against the odds on offer then moving value value from there.  More pen and paper than spreadsheets.

I've been rating horses on their 'abilities' though I do have my speed ratings as well.

So, for me, if I have
Horse A: 65
B: 65
C: 64
D onwards, 59 or less


So, for me, if A is 5/1 and B is 7/1 and C is the fave at 11/8, I will look closely at A and B since I have them of equal or slightly better than the fave so, at a glance, the value route is A, B.

I will focus on the top three and notch them against going, distance, jockey, etc.  From there, it is a matter of notching their rating up and down depending on going, jockey, distance, etc.

Right now, going over the last three months, I am starting to eliminate races that I can't seem to read.

I'm also going through having to reprogram myself a bit.  re: I have confidence in my abilities yet recently, a couple of times when I had horses that were clearly better rated (ie.  from topweight downwards, 56, 59, 55, 57, 55, 54, 63, 58, 53) I would find the top rated horse priced, in one case, at 30/1 and drifting.  I said "I must be wrong" and then looked at some other horse (or passed on the race).

the horse drifted to 42/1 and won.

There were five I passed on in June that were priced 24/1 or better. 

So my gut is I am overthinking and will work towards the balance that it is more important to get the ratings right after an event then having faith as I did before to notch up/down or keep the ratings the same on the horse once I watch the race.
By:
SHAPESHIFTER
When: 05 Jul 14 14:24
Which is what I am doing today.  182 races on the hard disc now to be chopped through
By:
Lex
When: 05 Jul 14 14:30
If you keep records of your betting, could you filter your results where only the 10% edge bets remained.



Unlikely. Everything reverts to the mean so you will still get peaks and troughs reflecting in winners and losers.


Too much jargon?

ok. then - "nope.can't be done" Laugh
By:
SHAPESHIFTER
When: 05 Jul 14 15:26
...for me, a pint and a stiff joint Grin
By:
SHAPESHIFTER
When: 05 Jul 14 15:28
If you can say, what were your findings here???

Right now, five furlongs though if I take the time to put more weight on the draw bias, it seems to be better.

Translated: ratings are good but fragile at 5 furlongs.  Need all the boxes to be ticked which, if everyone else is looking at draw bias, can be weaker prices.

I used to be able to find strong horses who could beat the bias which made for some good prices as the market followed the stats.
By:
Lex
When: 05 Jul 14 15:57
At my age I can do without stiff joints...
By:
CoinFlip
When: 07 Jul 14 16:52
I agree with Kenilworth's first post.  That's how I do it normally (2 below) although I think there can reasonably be suggested to be two versions:

1) Actual potential winnings / winnings of expected price.  You won't go that far wrong with this and you'll probably manage your bankroll better but  I don't like it because it doesn't seem to tie with an investment model.

2) Expected % chance of winning / odds implied % chance.  Seems to cope better with a range of odds and makes more intuitive sense.  Seems to prioritise bets better based on a small amount of modelling but probably won't manage risk as well as #1.

Not sure about the comment above about regression to mean in a sports context.
sort by:
Show
per page

Post your reply

Text Format: Table: Smilies:
Forum does not support HTML
Insert Photo
Cancel
‹ back to topics
www.betfair.com