- This topic has 16 replies, 11 voices, and was last updated 17 years, 2 months ago by
.
-
Posts
-
Just wondering how many forumites actually go to the trouble of rating each horse in a race and then compile their own tissue to see where the value lies.
Personally I usually try to narrow the field down to a few contenders and then guestimate as to which ones offer value or not but that is very subjective.
As a result of the subjectivity, I feel that maybe I should have a more objective approach to that particular aspect of betting.
Any thoughts?
I have voted NO as I don’t actually compile a tissue, rather preferring to be subjective at that stage of the process.
Well sort of. I narrow it down to anything between 2-8 runners and then compile a tissue based on a data table. Assuming I can be bothered of course 😉
It takes time and effort but, if you are at all serious, compiling a tissue is an essential step before considering any betting back , lay or swerve opportunity. Otherwise, others are controlling your betting and not you. It has to relate to your own selection methods and success rates.
I disagree Robert. I can’t tell what price a horse should be, but I can tell what it shouldn’t be.
Yep! "my tissue" allows me to establish a proportional hierarchy of probabilities in a given race, from most likely to least likely winner based on my analysis. It then allows me to match/benchmark my opinion against the opinion of the crowd. If I’m wrong I lose.. If I’m right I make money.. Simple for me and the best move I ever made (thanks Tooting 😉 )
I make a book, after a kind .. I would agree with slipperytoad on this, although not the tooting bit .. 😎
Just giving credit where credits due Dave.. It was Tooting who pointed me toward the inclusion of a tissue as part of my handicapping process.. Other inferences implied by my comments are purely innuendo and speculation.. :biggrin:
(Edited by slipperytoad at 8:51 am on Feb. 17, 2007)
Agree absolutely with Robert and Slipperytoad.
One way or another to gain an edge you need to be better at assessing the probabilities of the outcome than the market. While there are other ways of doing it (such as DJ’s method which I’m not knocking – some people instinctively know when a price is wrong) I find the method which works best for me is a rigid and systematic compilation of probabilities based on a variety of inputs (ratings, stats, factors such as trip/ground and current form).
TDK – I’m going to do a little exercise later today to examine (very crudely) the performance variation of low grade handicappers against the variations in performance of Group 1 horses.
Quote: from davidjohnson on 6:43 pm on Feb. 16, 2007[br]I disagree Robert. I can’t tell what price a horse should be, but I can tell what it shouldn’t be.<br>
david,
That means you should at least be trying to agree – it is one key area that should not really be an issue of differing opinion. You can prove or disprove it yourself very easily.
A 4/1 horse offered at evens is an obvious no no for many.<br>A 3/1 horse offered at 5/2 is a no no for me but something you might consider as reasonable, without a prepared tissue of your own.<br>A horse at 3/1 which is 5/2 on my tissue is a definite consideration. You might back that as well as me but then you are only backing half the value horses I might back.
Not backing a poor value horse is the next best thing to missing out on backing a losing bet and both can turn the same amount of form study effort from negative to positive.
Give it a go, perhaps, you have nothing to lose.
Absolutely TDK – although it obviously helps to have an idea of what is a backable price for whatever you’re punting on.
It also helps to find out what areas you’re best in I find – for example I tend to concentrate more on nervous chases & better class open/handicap chases.
Of course it can be a time consuming exercise & quite often I can’t be arsed with it but I always have a general idea of what price I’d back a horse at & at what price I’d leave it be. It’s bit of a no-brainer otherwise really – if you truly think a horse is a 2/1 shot, why would you want to unload at evs?
Ok – very crude analysis but of interest all the same.
I looked at 22 horses who all raced on King George day last July. 11 of them made up the fields for the KG and the Gr2 race at York same day. The other 11 ran in a Selling Handicap at Nottingham.
I looked at the last 6 races run by each horse with the following results. (All these results are based on their RP rating for last 6 runs)
Group horses – average rating 111 median rating 115<br>Seller horses – average rating 43 median rating 46
Average standard deviation for the 11 Group horses = 11.71 (this is an average of the standard deviation for each horses last 6 rating figures)<br>Average standard deviation for the 11 seller horses = 9.46
Median St Dev for the Group horses = 9.15<br>Median St Dev for the Sellers = 10.13
All of which suggests, on this meagre evidence, that there is as much variation in the race to race performaces of Group horses as there is in low grade racers.
However, I realise this small sample is pretty meaningless but perhaps food for thought and maybe a project for someone?
The question to be answered is : Do the figures support the notion that the better the horse the less the variation in performance? <br>
That is an interesting piece of research. Maybe there is only a perception that there is a wider variety in performance lower down the scale.
Then again, perhaps it is not so much the extent of the variation in performance, but the circumstances in which that variation takes place and the possible reasons for it.
I have been using ’tissues’ to compare with market odds since 1978. A tissue is just the opinion of the compiler and has no other merit than that, no different than a private handicap but expressed as prices. You can either compile your own or use one of the commercial services. I tend to use RPR these days as it is free, incorporates foreign form and seems to be as good as any.
I only consider non-hcaps with not too many runners and those runners where the RPR tissue price is no greater than 10/1 (outside that I don’t believe anybody knows what the odds really are). Several runners can be backed if the odds are viable, coupled to obtain a return of 100 points. Selections are supported at their tissue odds and not market odds.<br>The best thing to do is cherry-pick a good example the 2.10 Navan today:<br>  RPR<br>1 140  11/2 15pts SP 3/1 No bet<br>2 138  8/1  11pts SP 11/2 No bet<br>3 143  4/1  20pts SP 13/2 Bet lost = -20 pts Justpourit<br>4 151  7/4  36pts SP 11/8 No bet<br>5 133  50/1 (not considered over 10/1)<br>6 140  11/2  15pts SP 10/1 Bet Won = Return 165 pts Young Deperado
That’s how I currently employ RPR as a tissue using a simple arithmetical calculation changing the ratings into percentages, then into fractional odds and finally the points to return 100. I should add that if the points/price of the viable selections total more than about 80 points that race is eliminated as a betting <br>medium.<br>Nearly forgot, there is a problem with this system if not more than about three-quarters of the field have a rating as it skews the odds.
(Edited by Formath at 9:05 pm on Feb. 18, 2007)
Would be interested to know how you convert the RPR into percentages (though of course I wouldn’t necessarily expect you to tell me). There was a thread on this subject a few months back (started by EC I believe) and one of the things that came up was that the simpler methods of turning ratings into percentages (sum and divide) tended to produce an unrealistically cramped range of odds. Yours seem a more realistic spread.
Aranalde –
I developed a simple spreadsheet which converts ratings into odds and has a constant which you can vary which widens or narrows the ‘range’ of odds. It works pretty well and is a useful tool. UYou can also enter the actual odds and it calculates the differences (in % chance terms) between the ‘ratings tissue’ and the actual odds.
I have the constant set where I think it works best but haven’t carried out any meaningful analysis as to whether it is the best constant to use.
If you want a copy of the spreadsheet pm me your email.
Quote: from cormack15 on 4:50 pm on Feb. 18, 2007[br]<br>All of which suggests, on this meagre evidence, that there is as much variation in the race to race performaces of Group horses as there is in low grade racers.
However, I realise this small sample is pretty meaningless but perhaps food for thought and maybe a project for someone?
The question to be answered is : Do the figures support the notion that the better the horse the less the variation in performance?
If you think of a class pyramid then there is relatively little competition at the top, and getting on for hundreds of equal competitors nearer the bottom.
Just like in the Premiership there are only 3 likely winners in September, and at this stage only one, but in the Championship there could still be any one of six still in with a chance of coming top.
So if this is allowed for in the tissue the Grade 1 race has short priced favourites which tend to win and the seller has often short priced favourites which tend to lose. The selling horses then get falsely accused of being inconsistent when in fact their chances/ odds should be longer as any one of six can often be the winner, as the competition is greater and the difference in ability less.
<br>When I was writing Inside Track, I did some research on long priced winners. I looked at two full season of flat racing and found that 50/1 winners were just as likely in Group races as they were in sellers.
In fact over those two years, there were 20 winners with an SP of 50/1 or more and 3 of those were in Group races, and we all know that Group races don’t make up 15% of the program.
My view is that isn’t necessarily evidence of varaiation in performance in the horses involved – but that for whatever reason, punters were not able to correctly predict the likely level of performance of the winner.
Take the best example from Group races last year – Les Arcs winning the Golden Jubilee at 33/1. As subsequent events showed, punters simply hadn’t grasped the amount that he’d improved (or how soft the race was for a Group 1) – it was certainly not a case of Les Arcs being inconsistent. What price would he have been at Ascot if that race had come after the July Cup?
AP
<br>
- You must be logged in to reply to this topic.