Big Odds Winners

[rbtblb82]

Depending on who's stats you look at, favourites win around 1/3 races (33%). From the stats I collected for part of last year, it was 32.2%. I've always seen that as being 2/3 of races aren't won by the favourite, so you're far better off backing something with bigger odds.

There's always plenty of big odds winners, every day, here's a quick look at yesterday's results:-

• Catterick - 7 Races - 1 Favourite -- 8/1 + 11/1 + 33/1 -- 3,564/1 treble
• Lingfield - 7 Races - 2 Favourites -- 18/1 + 66/1 -- 1,254/1 double
• Ludlow - 8 Races - 4 Favourites -- 15/2 + 25/1 -- 212/1 double
• Perth - 7 Races - 3 Favourites -- 13/2 + 10/1 + 80/1 -- 6,600/1 treble
• Punchestown - 8 Races - 4 Favourites -- 17/2 + 14/1 + 20/1 -- 2,850/1 treble

Total Races = 37
Total Favourites = 14
Favourites Strike Rate = 38%
Total Big Odds = 11 (or 13 depending on what you consider big odds)
Big Odds Strike Rate = 30%

Percentage wise, I don't why people back favourites. I do occasionally but I think I'm going to put a full ban on it for myself, it's not worth it in any way, by odds & returns, percentages, or satisfaction. Yes you need some luck picking some of these big odds horses, but all you need is one single and you'll no doubt be up for the day, and hit two or three together and your laughing.

Some jockeys come in more often on big odds horses than others too. I have stats on big odds winners from last year on my laptop at home, I'll add them to this thread later. The jockey with the most big odds winners last year was G F Carroll in Ireland (flat).

 

[rbtblb82]

And we've already had a 16/1 and an 18/1 today.

3 favourites in a row at Warwick though, so a big odds winner could be due now - I'm having 2, 3 maybe 4 horses in the 15:15 @ Warwick......

Fat Sam 40/1 + Floating Rock 66/1 + Bucko's Boy 10/1 (was 18/1, market mover) + maybe even Let's Have Another 66/1 (or shorter with Artistic Choice 9/1). Spring Meadow is also a market mover here, was 8/1 now 9/2 but that's too short for me.

EDIT: Well that was disappointing, 4 favourites in a row now, and the two I dismissed because the odds were too short placed - I'll have to keep going for big odds at Warwick now though.

 

[rbtblb82]

20/1 landed @ Warwick but not the horse I'd picked, ha, never mind.

My stats for big odds were collected over 49 days last year.....

No. of Races = 2517
No of Winners @ 12/1 & Above = 339
Strike Rate = 13.5%
(I did 10/1 & Above as well and that was 18.8%, but were both a bit lower from a smaller sample later in the year, so it varies)

Jockeys Hitting @ 12/1 & Above.....

G F Carroll x10
S Foley x8
P J Dobbs x8
Jason Hart x7
Hollie Doyle x6
P T Enright x6
J Fanning x6
Billy Garritty x6
S A Gray x6
Thore Hammer Hansen x6
R Havlin x6
C D Hayes x6
S W Kelly x6
A Mullen x6
B A Curtis x5
Marco Ghiani x5
C T Keane x5
G Lee x5
W M Lordan x5
Tom Marquand x5
Jack Mitchell x5
David Probert x5
Ben Robinson x5
Rowan Scott x5
Harrison Shaw x5
James Best x4
Hector Crouch x4
Jim Crowley x4
S De Sousa x4
David Egan x4
Cieron Fallon x4
P Hanagan x4
Cam Hardie x4
Callum Hutchinson x4
D McMonagle x4
D J O'Keeffe x4
O J Orr x4
Siobahn Rutledge x4
K Shoemark x4
Hayley Turner x4

I did stats on favourites wins as well......

Number of Jump Favourite Wins = 935 races - 348 wins = 37.2%
Number of Flat Favourite Wins = 759 races - 268 wins = 35.3%
Combined Favourite Wins = 1694 races - 616 wins = 36.3%
Clear / Short Priced Favourite Wins = 152 races - 76 wins = 50%

Percentages for winners for different ranges of odds (these were collected later in the year from a total of 892 races)......

Jump Winners with Odds >Evens = 14%
Odds between Evens & 7/2 = 42%
Odds between 4/1 & 9/1 = 29%
Odds between 10/1 & 20/1 = 9%
Odds above 20/1 = 6%

Flat Winners with Odds >Evens = 8%
Odds between Evens & 7/2 = 43%
Odds between 4/1 & 9/1 = 31%
Odds between 10/1 & 20/1 = 12%
Odds above 20/1 = 5%

Combined Odds >Evens = 11%
Odds between Evens & 7/2 = 43%
Odds between 4/1 & 9/1 = 30%
Odds between 10/1 & 20/1 = 11%
Odds above 20/1 = 5%

Trainer & Jockeys Combinations.....

122 Jump meetings - 61 Trainer doubles = 49% of meetings had a trainer double (13 trebles 11%)
73 Flat meetings - 29 Trainer doubles = 40% (7 trebles 10%)
195 Combined - 90 Trainer doubles = 45% (20 trebles 10%)

122 Jump meetings - 64 Jockey doubles = 50% of meetings had a jockey double (14 trebles 12%)
73 Flat meetings - 62 Jockey doubles = 85% (7 trebles 6%)
195 Combined - 126 Jockey doubles = 63% (21 trebles 10%)

I know you can get states from all over but I like to collect my own specific stats in my own way sometimes, and the results were why I mostly play bigger odds, and can be a sucker for a combination, but if I stick to my rules and play the percentages it might just work one day.

 

[retriever]

I guess there's no harm in being a 'betting for fun' punter - the type the bookmakers love - but going by the number of responses you've had to your above posts you can probably derive that there's a different breed of bettor on this forum ( in the main ).

For example.....there's one key word missing from all the above and that's probabilities. None of the above facts and figures ( percentages ) have any real meaning or use.

My definition of 'big odds,' as a general rule of thumb, is a price ( to a near 100% book percentage ) equal or above the number of runners.
So a 4 horse race...... 9/2+ ( 5.5+ ) would be big odds.
.

 

[dave58]

Good points there retriever .
R rbtblb82 I think the question that I would ask is how do you actually use these statistics ?
My own thinking is that if there is a way to make a profit just by looking at the odds offered by bookies then surely they would be out of business by now ? A winner at 1/9 is better than a loser at 50/1 , it depends on how often you hit them both.
Example - I posted a winner here today at evens simply because it was the only horse that I was happy about putting up here. I have previously posted winners at double figure odds for the same reason.
I have also posted many losers, but I was ( and am) still happy about doing all of them because I had a logical reason for doing so every time.

 

[pete]

Total Big Odds = 11 (or 13 depending on what you consider big odds)
Big Odds Strike Rate = 30%

Unfortunately the reasoning for this selection method seems somewhat flawed ...

How many horses in total were "big odds" because in order to get the claimed 30% strike rate ... You would have needed to back all horses above whatever price you determined to be "big"

If there are 10 horses in each race, 6 or 7 may have "big odds" ... 6 races per meeting x 7 horses per race = 42 bets that you need to make in order to ensure that you trap the winners of those races.

In order to then get the doubles and trebles ... You have to perm all of those selections with each other in multiples .. doubles / trebles etc ... You are now looking at hundreds of bets per meeting in order to trap the winners that may or may not come.

As has been said before. Plucking numbers out after the events have already happened is very much useless to anyone ...

Your reasoning seems wrong I'm afraid, the numbers that you are using seem wrong, the claims of winners and strike rates are flawed.

Without wanting to come across as negative here the only way to take this forward is for you to post selections pre race and follow up each day with profit/loss ... Do this for a few weeks and see where we are at the end of the process.

I would be the first to be very happy to be proved wrong and would love to see it work for you long term, however I fear that it is far from being as straightforward as you wish it to be.

The first thing you should look at is to use Betfair starting price as the odds you will get at the end of the market that you are hoping to hit will be much, much better than SP.

 

[ARAZI91]

retriever - Your spot on here although one better way of defining it statistically is to use FS (Field Size) as a guide where for example if you were using BFSP you can use this metric by converting both Odds and FS to Probabilities ,simply by dividing both into 1 , so ...1/Odds and 1/FS.... then dividing the Odds (as a prob) into FS (as a prob) which spits out a numerical value.
((BFSP(p)/(FS(p)) where small (p)= Probability - spits out a value where 1.00 = Odds = FS as integers (when odds and actual Field Size are converted back from probabilities to numbers) , >=1.00 =Odds are equal or greater than FS as a probability and <1.00 =Odds are less than FS as a probability.
I call it simply (P/p)Ratio and it is very predictive as a metric when used from cumulative past races for horses , trainers, jockeys etc as it is a guide to past expectancy (price) whilst taking an important part (Field Size) into account. There is also a way of compounding "class" of rival making it even more predictive when using these values from past races...eg L3 races , L5 races , L10 races and also Career.

Example

The values can also be used when looking over a series of data - eg a year, just using the values produced from each race and splitting them.
During a typical season using this metric and splitting those that are greater or equal to 1.00...ie Odds(as a probability) are >=FS probability
and those that are less than 1.00..ie (Odds(as a probability) are <FS probability usually works out like this

UK Flat Turf/AW Season 2022 (01/01/22 -31/12/22 inclusive) - 6338 Races


You can see the difference.
Note that "proportion" is a big factor in statistics (which is why Impact Ratios/Values are such good metrics as they account for this)
Those that were priced greater or equal to Field Size (as probabilities) account for only 37.91% of all runners but 72.20% of all winners over the year whilst their opposite - those that were priced less than Field Size (as probabilities) account for 62.09% of all runners but only a measly 27.80% of all winners - this is simple proportion of both runs and wins
Using the Impact Ratio Delta(a simple subtraction of Impact Ratio 1 - Impact Ratio 2) a horse priced(as a Prob) >= Field Size Probability actually has a 147.50% chance of winning than it's opposite and a 90.5% more chance of winning than a random runner (All Runners)


R rbtblb82 - Whilst i admire anybody attempting to record their own statistics as it is hard graft and tedious work - what is missing from your statistics is basic "Proportion" which is what pete and retriever are explaining - what you are doing here is only counting winners and forgetting about the most important of the "strike rate/win%" equation ......runners...... in most of your statistics .
I would also add that a 49 day sample with this sort of stuff is open to plenty of noise and is no guide to the future, especially just counting wins. Most serious statistic users use maybe 5 years, some 10 and even then we have to account for margin of error, the game adapting and changing over that time,new trainer/jockey/horse/sire/ pools, tracks changing etc. There are techniques involving splitting a large sample into 3 parts mostly involving TTV(Test, Training & Validation) eg a 10 year sample could be split into 4 year "Test" , 3 year "Training" and 3 year "Validation" periods - ideally each part should show some correlation to the other parts.

"proportion" is missing in this part of your post with no account of runners



I don't know your dates of your 49 day sample range but i'll use BFSP(where odds are decimal) but use your same cut-off points in odds ranges (as if they were fractional ISP) BUT also include runners and a few other metrics to show you how the further away the market your playing the more "proportion" counts as the % of runners start to "outweigh" the % of winners for a full years UK Flat Turf/AW racing.
So using the same sample as above, the full year of 2022 UK Flat Turf/AW racing and a sample of 6338 races which includes EVERY runner, EVERY race
It would look like this



Green is where the odds-range beats "par" showed in All runners..........Red - NOT!

And if we graphed % of Runs To % of Wins it would look like this


Note that as both lines approach the >10-1 to <=20-1 range the % of wins dips under the % of runs and the gap between the two lines increases exponentially as the price gets larger - it's simply ......"PROPORTION"

Note* This is not to say there is not opportunity in the "red" odds-ranges as remember this a full years results as a collective (Btw NH racing shows the same trend as i seen you have included jumps racing in your study) - but caution is advised and indeed i would say extreme selectivity - those who play in this sort of odds-range expect a lowish strike rate, longer losing runs and probably demand an extra premium on top for "Margin of Error" from a "Value" point of view.
Collecting your own statistics is great but most of this stuff is basic and very public knowledge and easily found in most commercial software and what i have showed here is factored into the market and has been for a number of decades now.
What was interesting about your post was where you ventured to look for "Longshot jockeys" ....but again just "counting" wins is meaningless without the other side of the equation - RUNS - having done some of this stuff in the past i reckon Trainers would be a better source to record in this area -"jockey" has more of a random factor to it when it comes to "longshots" but a real study of "Longshot Trainers" can be worthwhile - i will post up a recent study of "LS trainers" but beware when you introduce an odds-range to a factor it is basically a form of "backfitting" so you also need to check for significance - CHI (the racing version of it developed by Steve Tilley) or better Standard Error(SE) are decent metrics to use.

 

[ARAZI91]

Unfortunately the reasoning for this selection method seems somewhat flawed ...

How many horses in total were "big odds" because in order to get the claimed 30% strike rate ... You would have needed to back all horses above whatever price you determined to be "big"

If there are 10 horses in each race, 6 or 7 may have "big odds" ... 6 races per meeting x 7 horses per race = 42 bets that you need to make in order to ensure that you trap the winners of those races.

In order to then get the doubles and trebles ... You have to perm all of those selections with each other in multiples .. doubles / trebles etc ... You are now looking at hundreds of bets per meeting in order to trap the winners that may or may not come.

As has been said before. Plucking numbers out after the events have already happened is very much useless to anyone ...

Your reasoning seems wrong I'm afraid, the numbers that you are using seem wrong, the claims of winners and strike rates are flawed.

Without wanting to come across as negative here the only way to take this forward is for you to post selections pre race and follow up each day with profit/loss ... Do this for a few weeks and see where we are at the end of the process.

I would be the first to be very happy to be proved wrong and would love to see it work for you long term, however I fear that it is far from being as straightforward as you wish it to be.

The first thing you should look at is to use Betfair starting price as the odds you will get at the end of the market that you are hoping to hit will be much, much better than SP.

Bang on about the "permutations" here pete ,even just for doubles or trebles as my study above showed there will be tons more horses at higher prices than lower prices, Imagine a card at the Cheltenham Festival where the avg FS since 2011 is around 17 runners per race and using >10-1 as a cut-off point -indeed since 2011, those runners actually account for approx 73% of all runners - the perms cumulatively would indeed be astronomical. Sadly with an astronomical loss.

 

[mick]

When describing my own thinking the mantra i will have used more than any would be " Its all about the odds " but even those who agree will have there own interpretations and uses of same. ? Re bigger odds while its never wise or fair to generalize lets use 10/1 as the benchmark any horse at this price will have negatives to chance in its form or circumstances with the option for backers to attempt to find valid excuses or reasons which justify still backing it to beat those negatives.

There is an arse about face version of the above when those at the front of the market are being over bet for arguably the wrong reasons which then creates bigger than anticipated odds against others. But imo a lot of work and experience will be needed in order to identify and get the best from either situation.

When i start working a race of interest my attitude is i am not going to find any horse i wish to back regardless of its odds and prob 80% of the time this proves so, but when i do identify a possible bet which progresses into a horse i want to back then its odds become the final part of the process. I set a minimum value price and will only bet if this or higher is offered.

I abbreviate the above with MAP Minimum Acceptable Price and as an example if i have set mine at 8/1 and the market is saying its a 14/1 chance then my attitude is i have not started out by seeking a big priced selection but rather it has found me. Stats and numbers can prove useful but unless you put in the effort to understand what they may be saying and why then i suspect the Bookmakers will be pleased that you are betting off them.

 

[retriever]

Hi R rbtblb82
I don't wish to be rude - just trying to help along with others on this forum - but even your stat collecting appears to be flawed as well as your general method.

Take the first on your list - G F Carroll. You say over 49 days last year he had 10 winners at 12/1 or above (?)

Well I count 7 in total and that was between June and October ( 5 months ).







I would suggest you avail yourself of a stats program, like HorseRaceBase, and do some proper research. You may enlighten yourself and no doubt save some money in the long run. Best of luck.
.

 

[dave58]

R rbtblb82 maybe an easier way forward with this might be to find favourites that you have your doubts about and lay them on Betfair - probably more practical than backing loads of horses.?
Just a thought.

 

[mick]

R rbtblb82 maybe an easier way forward with this might be to find favourites that you have your doubts about and lay them on Betfair - probably more practical than backing loads of horses.?
Just a thought.

Or using the above situation as the starting point for backing some long shots in the race. ? Another version i saw on this forum was a member who started a thread called Taking on Hugh Taylor. With part of his thinking being the market following this tipster has created value opportunities in the races he tipped in. Often when a horse drifts it has nothing to do with its actual chance of winning.

The majority of Bookmakers and there employees are not good judges of form, which is why they have to make a living from laying not taking the odds. There early odds have to include an opinion but they cover there backsides via working to piddle take defensive over rounds. Sooner or later they revert to betting to supply and demand market forces and pure numbers, and this can create opportunities for us backers.

 

[dave58]

Another way to use ARAZI91 suggestion about looking at the trainer is to look at trainer/jockey combinations.
I've got one system that I can only describe as nepotism in action - sometimes you can see patterns where a trainer uses a certain jockey under certain conditions

 

[Jackform]

dave58, 'nepotism' is the key word I picked out. You should be aware that I am just a pen-and-paper form compiler and not good with todays search engines. However, years ago I used to follow hcap runners at long odds with some success that had close family connections - trainer/owner, jockey.
Some other selections bet were from 'upside down hcaps', that is starting from the lowest weight upwards. Selections were the first one or two runners ridden by (3) and (5) lb claimers.
A similar system was to take to take the lowest win or place runner from last time ridden by a current top twenty jockey.
Some of the above may still be worth looking into?

A mate got taken on as a steward at Uttoxeter and asked if I could possibly give him a selection on his first day to impress people with.
I gave him Verasi that had dropped over a stone from the previous race but had some reasonable form previously. He gave it to the Clerk of the Scales and Tommo, who was commentating that day, and it came in at 66/1.

 

[OnTheNod]

I think their is a way with big odds horses but be prepared for some long losing runs and maybe only back them once in most cases, another thing its not very easy to find a false favourites from my experience i tried laying them for a while and draw about level.

 

[dave58]

dave58, 'nepotism' is the key word I picked out. You should be aware that I am just a pen-and-paper form compiler and not good with todays search engines. However, years ago I used to follow hcap runners at long odds with some success that had close family connections - trainer/owner, jockey.
Some other selections bet were from 'upside down hcaps', that is starting from the lowest weight upwards. Selections were the first one or two runners ridden by (3) and (5) lb claimers.
A similar system was to take to take the lowest win or place runner from last time ridden by a current top twenty jockey.
Some of the above may still be worth looking into?

Jackform exactly the sort of thing I was referring to.
Great minds think alike !!!

 

[Jackform]

I seldom do back-fitting and you should always put your money where your mouth is so I did a quick check of the Perth card.

Perth (going forecast GS) Market odds from around 10.15. L Russell top track trainer.
1.30 Inferno Sacree (9/1 mkt) drop in weight and (3) claimer
2.05 If Not For Dylan (22/1) drop in weight and (5) claimer.
4.25 Syltezar (28/1) drop in weight
5.00 Nights Of Doyen (50/1) drop in weight.

 

[Jackform]

This is my last word and I will b*gger off , Just to show that I am getting to grips with this online thing I queried the A1 robot about our problem and it knows less than me .

There is no definitive answer to how to select horses at long odds that have a chance of winning, but there are some tips and strategies that might help you improve your chances. Here are some of them:

• Look for value rather than just the most likely winners. Value is when the odds are higher than they should be based on the horse’s chances of winning.
• Use past trends of the race to identify the type of horse that has performed well in similar conditions. For example, the distance, the ground, the course, the trainer, the jockey, etc.
• Lay against the horses that have the least chance of winning in the race, especially if they have odds between 3.0 and 5.8. This means you bet that they will not win, rather than that they will win.
• Understand the basic probability of horse racing and how to calculate the expected value of a bet. This will help you avoid bets that are too risky or too low-rewarding.
• Avoid betting on long shots that are too unlikely to win, such as those with odds over 20.0. These bets are often foolish and rarely pay off (talking b*ll*cks!).

I hope this helps you make better decisions when betting on horse racing. Good luck!

 

[jeansten]

 

[Sandhog]

Today at Doncaster may be a good meeting for R rbtblb82 to show his way of betting before the races are over, as fields are big and I would expect some big-priced winners.

I'd be especially keen to see how he uses his jockeys on 10/1 shots, which I think he mentioned.
It wasn't clear t me if he just does them all when they fit the bill, or, what?

And, though it may limit the amount of bets he has, the effect on the percentages from backing several in a race may not be what he wants.

I don't suppose anybody will mind if he puts up his selections here, so I can see his work in action.