CD Course and Distance

[Matunty]

Hi guy and gals! I am developing a system and was starting to lose. Down to 25%. I started to paper bet on races with less than 13 runners and more than 6 where there was only one horse who had a CD against their name (online Racing Post). The horse had to be in singke figure betting but obviously not the fav or even second fav although sometimes they were. Early uindcations are of an upward trend. Is there anyone else out there who has a system whcih is either based upon, or uses CD (course and distance winner) as a factor. Let's share more notes and make some serious long term dosh!

Matunty

[mathare]

Hi guy and gals! I am developing a system and was starting to lose. Down to 25%.

Of your starting bank? The SR is down to 25%? What?? If it is bank then the system may be fine but the staking could be all wrong.

I started to paper bet on races with less than 13 runners and more than 6 where there was only one horse who had a CD against their name (online Racing Post). The horse had to be in singke figure betting but obviously not the fav or even second fav although sometimes they were. Early uindcations are of an upward trend.

Where did these rules come from? I can see the reasoning for backing CD winners but why in fields of that size? Why does the horse have to be single figure odds? Is that forecast odds or odds close to the off? Why "obviously" not favourite or second fav?

When building a system the rules have to hang together in a sensible way. Look at the combined set of rules and ask yourself why horses that meet these criteria should win more often than horses that don't meet them. If you can't find a reason that all the criteria should apply then your system is probably flawed. In this case what makes you think that CD winners in fields of 7 to 12 runners that aren't fav/second fav going off at single figure odds are going to be profitable? CD winners in such fields I can seee but why the odds filter too?

Is there anyone else out there who has a system whcih is either based upon, or uses CD (course and distance winner) as a factor. Let's share more notes and make some serious long term dosh!

I follow several systems which use CD as a factor but they also use other rules that complement this one to form a set of criteria that all hang together and make sense. I can't give out the rules though as they are not my systems.

If you're after making serious long term dosh I suggest you use the advice and information provided by professionals. Perhaps buy one or two of Keith's eBooks to get you started and get an idea of some of his systems. Read his blog posts on developing a system. And most of all put some serious work in if you plan to develop successful systems. Play around with a lot of rules but for each ask yourself why it should/shouldn't be a factor. Back test systems under different staking plans etc.

[Win2Win]

Using CD without decent filters WILL result in losses, as CD is not that important on most tracks. It usually only has a higher significance at tracks like Chester, Goodwood, Beverley.....

Here's the results for the last 8 years for ALL CD winners, and the prices ranges:

Years 2000__2001__2002__2003__2004__2005__2006__2007__


Odds,_Starting_Price
CATEGORY_____________________WINS____RUNS__STRIKE%_____LSP_____LSP%_____VSP%

<_1/2___________________________9______13__69.23_____-1.51___-11.62____-2.98

1/2_-_20/21____________________34______76__44.74____-16.56___-21.79___-21.69

1/1_-_11/8_____________________37______99__37.37____-16.98___-17.15___-18.33

6/4_-_15/8_____________________58_____164__35.37_____-8.64____-5.27____-4.80

2/1_-_7/2_____________________230____1136__20.25___-244.60___-21.53___-21.16

4/1_-_15/2____________________329____2511__13.10___-390.00___-15.53___-15.78

8/1_-_15/1____________________201____2957___6.80___-678.00___-22.93___-22.18

16/1_-_31/1____________________57____1738___3.28___-545.00___-31.36___-33.28

32/1_-_63/1_____________________8_____603___1.33___-297.00___-49.25___-49.37

64/1+___________________________1_____131___0.76____-64.00___-48.85___-38.57


....and to prove what I say is correct, the tracks here with the higher SR's (and some with profits for CD winners), all have quirks.


CATEGORY_____________________WINS____RUNS__STRIKE%_____LSP_____LSP%_____VSP%
course_not_group_B____________700____7211___9.71__-2090.35___-28.99___-23.92
Ayr____________________________34_____332__10.24____-29.50____-8.89_____3.70
Chester________________________19_____137__13.87____-10.53____-7.69____-4.40
Haydock________________________35_____261__13.41____-12.33____-4.72____10.04
Leicester______________________20_____156__12.82_____-0.11____-0.07____-6.11
Lingfield_(turf)________________8______82___9.76____-49.12___-59.90___-17.17
Newcastle______________________38_____307__12.38_____15.63_____5.09____14.46
Nottingham_____________________26_____237__10.97____-63.32___-26.72____-8.08
Salisbury______________________17_____148__11.49_____74.25____50.17___-13.40
Windsor________________________45_____323__13.93______9.59_____2.97_____1.00
Yarmouth_______________________22_____234___9.40___-106.50___-45.51___-20.22

[Matunty]

Thanks for your response which is always appreciated. The CD factor interests me as I have noticed that it appears to be a factor. The logic of your response is not in dispute. My argument for the string of factors is as follows. Maybe I am an amateur and way off beam and would appreciate any advice even if that advice is to rubbish my system as it has developed so far. My response is

1. races of 7 - 12 reduces the variables and yet still provides decent odds, even with favourites. Whereas races with 20+ runners increases variables and odds but more difficult to find winners. At the other extre, a three horse race may statistically produce the greatest chance (maximum of 3 -1) but the value isn't there.

2. I am, on advice gained here, backing 2% of my bank on a daily basis; one win bet per day. My strike rate is now only 25% which must be considered low. At any rate I am losing money increasingly

3. I bet single odds as market strength must be a factor otherwise why don't we all just back the 100/1 shot until it wins! 4. Can you explain what you mean by 'staking plan'. I bet 2% of the bank regardless. Are you suggesting that I increase the % as the bank reduces, say to 2.5, 3%. 3.5% etc?

I really hope that you reespond as the development of a winning system is important to me and, despite your advice about reading up on existing systems, take pleasure in the creativity of developing a (possibly) unique money making system.

Thanks

Matunty

[Win2Win]

Are you reading the 'How to Build a System' Blogs? http://www.win2win.co.uk/forum/blog.php?u=1

1. You will actually find that in non-handicaps, finding the winner in a 20+ race is easier than a 7-12. This usually happens in the last 6-8 weeks of the season as trainers rush to get runs into the horses, and most of these big field races have few with a chance, they are used more for 'schooling'.

2. Staking should usually start at 3%, and adjusted down as profits increase to protect the bank. You say 25%, but over how many selections?

3. You keep saying 'system' now you are talking about using market strength, which would make it a 'method'. You need to do one or the other, otherwise profits will struggle. As for 100/1 shots, that is a bookies price, they are more likely much more than that, hence why you will not make money from them backing blindly. You only have to lool at the exchanges, a Bookie 100/1 shot may be 999/1 on Betfair.

[Matunty]

I am still determined to focus on the races whcih only have one horse with a CD against their name. Presumably if boxed together, this means at the course which the race is on as opposed to separate boxes? I will remove the 7-12 from my system, although you seem to suggest that this is seasonal?

Thanks for your continued rapid responses. This site is BRILLIANT

[Matunty]

I have started with races which only have one horse with CD against them. There may be others with C or D but very few races have only one horse with cd. This appears to be a good starting point to start a study of a hypothesis, including adding other filters as suggested by your blog and other fascinating discussions here. The logic (which my 'friend' Mathare will take apart!) is that a horse with CD has at least this single significant credential. It has won at this course before and over the same distance. Adding other filters, as advised, should reduce the odds again. I just like the primse suggested by CD and am prepared to test it over 100 paper bets. I am sure there are others (in here) and out there who use CD and some may be following a system, either derived or composed which uses the same premise as I am.

Worth a try; and you know I will let you know at each 10 bets the % return (SR).

Cheers guys!

[MattR]

Matunty just a few thoughts here.


Above you say one win bet per day. Does this mean you are just picking one if there is more than one? If so why just the one? If it is a system then a qualifier is a qualifier regardless if you have 1 each day or 12 one day and none for four days.

As you are paper testing this, an idea would be to also log all those that are 'double figure' odds too.


Some stats that might help....

For 2003-2008 Flat races only.

Horses that had won at course and distance (this is not broken down to races where only one horse had a CD though but it may at least give you a starting point)

If we add the following filters

Non handicap races only
Finished 1st or 2nd last time out
Running in the same class as last time

We get the following stats

Bets 549
Wins 125
Pr to SP 114.27
Strike Rate 22.77%

However remember there's no data there as to whether there were 1,2 or more horses with CD in those races. Also be aware that not all years brought a profit and the biggest profit was 6 years ago. The longest losing run was 23 races and longest winning run was 5 races. Also worth noting is that though the strike rate is naturally higher on the lower odds horses, the profit is spread around a range of odds, high and low.

Hope some of that's useful for you and that might give you some ideas to where you can go with your own system.

[Win2Win]

........................................

Hope some of that's useful for you and that might give you some ideas to where you can go with your own system.

You after my job? :D

[MattR]

:D

[Matunty]

MattR thanks for your valued comment. I am starting by selecting ONLY races where there is only ONE CD winner. With other filtersm the best CD match was Wisdom Kiss which was my only bet. It won at 3 -1. Would it have won anyway? Hey, the million dollar question. I would like to think the system is working although it is far too early for any concusions. I want to get to 100 'paper' bets. Can't resist when it looks good, as today and had a 'fun' bet, which was 0.3 rather than 3% of the bank. The best fit tomoroow using the same system and filters is Mister Jingles at Mussleburgh fc is 9 to 1. Will keep this thread going as I want to discuss the filters with you. Cheers Steve

[MattR]

Steve, glad it was of some use, can I just ask on this bit that you wrote above :

With other filtersm the best CD match was Wisdom Kiss which was my only bet


Are you keeping a note of any other qualifiers that meet all the filters if there are more than one? Just I note that you're saying the 'best' match. Does that mean more than one qualified, but you have narrowed it down to one yourself using whatever means?

Good win with todays one anyway. :)

[Matunty]

Matt. Thanks for your interest. Mister Jingles didn't win and came someway down the field. Sometimes I think one can learn as much from working out why it didn't win (after the event) as trying to select winners; just takes twice as long!

The answer to your (insightful) question is:

1. If there are more races with only 1 CD in the race then I apply the following rules (based on other advice I have accepted on these pages) until I am left with one CD and one race!

a) I select the race with less than 15 runners (the Boss suggested that)
b) I then favour non handicaps (again based on advice elsewhere)
c)The draw then is factored in (I have awarded points 1 to 4) for strength of draw based on the Adrian Massey data

If that hasn't eliminated the others and I still have more than one race I will compare:

d) The % wins from last season for the races
e) The % success of the trainer of such races

That usually results in the 'best' one race with one CD runner?

Any comments would be really appreciated Matt!

Cheers

Steve

[mathare]

What happens if you stop after rule c I wonder...

There is nothing to say a system can only yield one selection in a race let alone one selection per day. You may have more luck without rules d and e. It's certainly worth looking back over your data (if possible) to check this

[Matunty]

Having one selection a day is probably far more than most average punters. I enjoy the challenge of 'working it out'. I am sure we all do. d) looks at the % of wins in this type of race and age of horse from the last year. I have awarded points for the range 1-5%, 6-10% etc. I have also thought carefully about your advice about weighting of factors/variables although it is |(virtually) impossible to quantift the relative importance of each factor in order to weight them. This is particularly the case because the variables vary in their degree of relevance toe ach other. Ground vs LTO stats for example are just tow that vary in relation to each other. When you have a system, such as the one I am studyig and developing with 8 or 9 factors then the relative weighting is impossible to manage. However, I would be interested to hear your views and others on this topic of relative weighting. If awarding points for ground vs jockey performance last season or class of race or type of race etc which (of these and others) would you consider to GENERALLY be the most significant?

Regards

Steve (still learning!)

[mathare]

Having one selection a day is probably far more than most average punters.

Not sure I agree with that. It seems to me as though you are forcing the rules somewhat to ensure you have one (and only one) selection each day. I'd rather have a system with set rules that may throw up nothing for 6 days and then 7 selections all on the same day than force it to give me one a day

[Matunty]

Not forcing really. Its about having a system fro which the best can be selected, therefore, theoretically, increasing the odds. I understand your point though. Perhaps you can share the method you use than does not help select anything for 6 days and then throws up 7 winners! Or do I have to pay for that!!

cheers mathare, enjoy reading your responses.

[Win2Win]

Having one selection a day is probably far more than most average punters.

I think you'll find the average is close to every race.

[Matunty]

Confused? The number of bets placed each day divided by the number of adults placing bets (if known) would give the average number of bets per punter. This could only really be calculated from e-betting logins and bets placed; probably unknown? The question remains, how many bets are placed by punters on average per day. I am placing, on average 1 per day, which takes hours to calculate. Steve

[MattR]

1. If there are more races with only 1 CD in the race then I apply the following rules (based on other advice I have accepted on these pages) until I am left with one CD and one race!

a) I select the race with less than 15 runners (the Boss suggested that)
b) I then favour non handicaps (again based on advice elsewhere)
c)The draw then is factored in (I have awarded points 1 to 4) for strength of draw based on the Adrian Massey data

If that hasn't eliminated the others and I still have more than one race I will compare:

d) The % wins from last season for the races
e) The % success of the trainer of such races

That usually results in the 'best' one race with one CD runner?

Any comments would be really appreciated Matt!

Cheers

Steve

If I could throw up a thought here, supposing you get to C and are left with just 1 selection so that is now your bet for the day. What if that selection has terrible stats for D) and E) ? If there were two or more selections left after C then the chances are this selection would 'lose out' and not be bet on when all were moved forward to test D and E, but because it's the only left at C it will be bet on. So if you see what I am saying here is that in one instance you wouldn't bet on it, but if it were the only one left at C you would.

Maybe a way round that would be to set a lower limit % for D and E and then if you have just one selection left after A/B/C then the selection must be over your lower limit % for D and E otherwise it is not a bet.

I do think you should at the very least, whilst paper trading this, make a note of all horses that get through to stage C. You could be missing out on winners.


Steve, one thing I just noted is Mr Jingles was in a Handicap race. So I'm assuming you are doing Handicap races even if there are non handicap qualifiers if they come out better after the draw points are awarded?

[Matunty]

Yes on the last point. I have taken your advice and have entered all the 1 X CD horses into excel and then used the weightings and scorings against the variables previously described. Interestingly, your advice is correct. It has thrown up different and I think potentially logically better selections. The selctio (single bet) for each day will be listed in this thread as long as the system is developed. The best CD+ selection tomorrow is Murrin in the 5.45 at Goodwood. This is bet 9 of 100 in this developing system. The paper bet after 8 bets on a bank of £1000 is £30.12. I adjust the stake up or down, as previously described by increasing the % of bank after a loss and reducing on a sliding scale if the horse wins (in order to protect the profit). Comments please always welcome.
p.s. this site is fantastic! Love the interaction and aulity of conversation. Have you been on this long?

[mathare]

I am placing, on average 1 per day, which takes hours to calculate

I place 30+ on average per day. So as Keith says, more like 1 per race

[Matunty]

Good luck m8. I leave home at 6.30am and get home at 7.30pm. I find this site and my focus on racing and the development of my system therapy! I would prefer to bet 3% of my bank on one horse selected the night b4 and then come home and see how it got on. Sometimes I even have the pleasure of watching it as I have Racing Uk.

Regards

[MattR]

Not sure it would be wise to increase the % stake after a loss (unless it was a very high strike rate system) Even a 50% SR system will produce runs of 7 losers in a row and increasing the stake after a loss would kill the bank in no time I would think on say even a 25% SR system. I may be corrected on this but I wouldn't think it would be a good idea.

A couple of filter thoughts for your system.

Won at same class as current race? It would seem logical that a horse who had won at the same class has a better chance than one who hadn't.

Also could be worth taking into account LTO result. Around 45% of all flat races are won by a horse who came in the first three last time.

[Matunty]

Matt. Thanks again. This is really useful. I had already factored in LTO giving points to 1st last time, less points to 2nd and then 3rd. I have not awarded points for horses outside the top 3, regardless of teh size of the field. Class (drop in) is not yet a factor though I appreciate why! 2 questions here 1. Is there data on the percentage of horses who win when lowered in class? 2. What significance would you give this over other variables? This is where I am short of information. Gow would you rate drop is class against drop in distance. Its not an exact science (otehrwise we swould all be millionaires) but it must be about applying logic and intelligence and known data in order to increase the odds of winning! Comments? Advice?

Cheers

Steve

[MattR]

Just checked on Class and for Flat 9.9% win when previous race was a higher class. That's just as a single filter. If they are a CD winner as well that comes up to 11.34%

As to question 2 I don't really know. I suppose it depends on the difference between classes and that's something I don't know. Keith or Wayne or others on here could probably give you some idea on that but my knowledge is limited on things like that


Using the stas, if you take Flat Races and horses that have won CD, yu get a 10.96% SR. Adding in running at a longer distance brings that SR down to 9.64%. Adding in a shorter or same distance as previous race brings the SR to 11.2% So that would suggest to me that a shorter or same distance doesn't have much effect but a longer distance is definitely detrimental.

As to class, taking the 10.96% figure from before, if we add in curret race is lower or same than previous race class we get 11.54% but if we make it that it's moving up to a higher class it becomes 9.59%.

So the two would seem fairly similar in significance. If the horse is moving up a class and also running a longer distance combined we come in at 6.32% So if you are using points then it would seem to make sense to reward dropping in class or distance more than going up in either. If we use same/lower class and same/lower distance the % moves to 11.62% SR.

So it would seem both definintely have a factor in the results but as to how much I guess is the intangible part!

[counterfeit]

If the horse is moving up a class and also running a longer distance combined we come in at 6.32%

Sounds like the basis for a decent lay system.

[MattR]

Sounds like the basis for a decent lay system.

Actually it does doesn't it? I might explore that a bit further!

[Win2Win]

ROI% will not be good enough for horses below 10/1

[MattR]

ROI% will not be good enough for horses below 10/1

Keith what do you generally do when working for lay systems odds wise? I have used 8/1 SP and under as a base but it's a bit vague really as whether they were under 10.5 on betfair is a different matter.