• Hi Guest Just in case you were not aware I wanted to highlight that you can now get a 20% discount on Inform Racing.
    Simply enter the coupon code ukbettingform when subscribing here.
    We have a lot of members who are existing users of Inform Racing so help is always available if needed.
    Best Wishes
    AR
  • Hi Guest Just in case you were not aware I wanted to highlight that you can now get a free 7 day trial of Horseracebase here.
    We have a lot of members who are existing users of Horseracebase so help is always available if needed, as well as dedicated section of the fourm here.
    Best Wishes
    AR

Early days

Hi Mike,
TheBluesBrother TheBluesBrother
Maybe I came to this too late to see the original times, (potential buyer for the house tomorrow so my morning was lost to vacuum cleaning and nailing the odd fence rail back up) _ had to check the 2 Chelmsford times from video as HRB had them at 3:48.28 and 1:22.24, my RP download had them at 3:28.18/`;12.24 which looked correct (as close as I could say) when I checked them.

I couldn't check Timeform, their site is wonky again and I couldn't get the race results to download - I presume this is the source of your slightly faster times?

Dave
 
Here we go again,
no Easter eggs in sight, I'm having a deprived 2nd childhood.... my wife is planning to rustle up an omelette but I don't think that really counts.

Tomorrow, a busy day, is attached - a lot of racing on the go. I've attached race_listing_2apr.csv as a bit of race picking help for tomorrow, as there are so many races on. Personally I prefer to pick my races from the better end of the spectrum, if the prize money is good then you know everyone is likely to turn up with a fit horse that will be trying to win. If you have Excel or similar you can use the sorting and filtering commands to quickly spot the sort of race you want to be in - use the filter option (data menu)in Excel to make each column searchable, with dropdowns to select/deselect values listed - I use this file daily to pick likely races to look at, I will perhaps quic,kly filter to only show 5f and 6f sprints, maybe remove any with too many runners for my liking, or I will perhaps isolate the top three or 4 races by prize money to see where the best class of racing is ocurring.

I'm not posting it daily or anything, I just thought some folk might like to try it out to help with their own race selection, and as tomorrow has 7 meetings on I thought we could all use a bit of help in finding the most likely races to bet in.

Yesterday:
Flat Handicaps
Raw ratings: 2/14 Odds: 2.75 14.0 (Return: 18.75)
Mkt filter 1/6 Odds: 2.75 (Return: 3.75)

Flat Non - Handicaps
Raw ratings 3/12 Odds: 1.5 0.73 1.88 (Return: 7.11)
Mkt filter 3/3 Odds: 1.5 0.73 1.88 (Return: 7.11)

NH Handicaps
Raw ratings: 1/10 Odds: 4.5 (Return: 5.5)
Mkt filter 1/6 Odds: 4.5 (Return: 5.5)

NH Non - Handicaps
Raw ratings 0/3 Odds: (Return: 0)
Mkt filter 0/1 Odds: (Return: 0)

39 out 31.36 back - disappointing, 6 winners is a bit under par, a mix of odds. Selections list (market filtered) about broke even. Today has been okay, 32 out and almost 47 back, which helps to make up for a bit of recent pain.


Dave
 

Attachments

  • 2apr.xlsx
    526.4 KB · Views: 1
  • AWcard_2apr.xlsm
    90.4 KB · Views: 1
  • full_ratings_2apr.xlsx
    71.9 KB · Views: 2
  • selections_2apr.txt
    1.8 KB · Views: 1
  • race_listing_2apr.csv
    3.8 KB · Views: 1
Well, the weather has played a blinder again then, never mind. Not much on tomorrow, a plod round Fairyhouse in the mud and mist or an AW meeting at Lingfield, which I will look at... Heavy going just isn't fun for sorting winners out for me, too many odd things happen, so I'll take a look at the AW but not do much.

Yesterday:
Flat Handicaps
Raw ratings: 2/4 Odds: 3.5 3.0 (Return: 8.5)
Mkt filter 2/3 Odds: 3.5 3.0 (Return: 8.5)

Flat Non - Handicaps
Raw ratings 1/2 Odds: 0.53 (Return: 1.53)
Mkt filter 1/1 Odds: 0.53 (Return: 1.53)

NH Handicaps
Raw ratings: 2/12 Odds: 3.0 5.0 (Return: 10.0)
Mkt filter 1/7 Odds: 5.0 (Return: 6.0)

NH Non - Handicaps
Raw ratings 4/15 Odds: 3.5 5.0 2.25 12.0 (Return: 26.75)
Mkt filter 1/5 Odds: 3.5 (Return: 4.5)

33 out almost 47 back, that'll do. Today was pretty much washed out, from the 2 meetings that survived I think there's a single 7/2 winner so nothing special on Easter Monday I'm afraid.... roll on the flat and some nice weather!

Dave
 

Attachments

  • 3apr.xlsx
    192.9 KB · Views: 0
  • AWcard_3apr.xlsx
    71.7 KB · Views: 0
  • full_ratings_3apr.xlsx
    30.7 KB · Views: 0
  • selections_3apr.txt
    474 bytes · Views: 1
Ah good, I hope it's useful - shame about the MR though.
As I was saying a page or two back I also looked at narrowing fields down using VDW type criteria - I couldn't claim to be using his method as I doubt anybody on the planet was ever sure what it actually was - so I wrote myself s similar sort of scraper to collect tomorrow's cards off the RP site, bringing the form lines info with it.... I've attached the one I collected yesterday as an example, columns are name-Master RPR - Date - race info - rpr for race - comment - position/number runners

I use this file as an extra data input into my daily card generating program, output an extra file with data for each runner, and can check things out that I've come across in my reading - some of the VDW ideas in some cases but I can't say I've found anything amazing yet so I'm still pretty much based on speed figures and a bit of pace work. It's handy to be able to test ideas out while you carry on doing the normal stuff.

I made both Irish courses -3.2 for yesterday (I didn't think one was much faster than the other so playing with them would have just been to make it look more accurate rather than actually making it so). The ratings were all over the shop so I can't say I'm very happy with them, but estimating an accurate figure for running through a bog in the mist is probably not an exact science! Frankly I'm glad the racing is cancelled, rather than filling my records with dubious results.

Oh well, off to get tomorrow sorted - hope you aren't too far underwater where you are, the fish will be coming to you rather than vice versa at this rate.

Dave
 

Attachments

  • RP_racecards_2018-04-03.csv
    138.3 KB · Views: 2
Well, it's not a bad day if you like class 4,5,6 on the AW, and at least there's little risk of somebody drowning on the back straight..... but tomorrow doesn't look that inspiring so it's probably just as well that I'll be out most of the day. Leopardstown is still on at the moment, but I'm no fan of flat racing on heavy ground so good luck to anyone trying to find a winner there.

Never mind, cards etc attached as ever.
Yesterday:
Flat Handicaps
Raw ratings: 0/4 Odds: (Return: 0)
Mkt filter 0/2 Odds: (Return: 0)

Flat Non - Handicaps
Raw ratings 0/1 Odds: (Return: 0)
Mkt filter 0/0 Odds: (Return: 0)

NH Handicaps
Raw ratings: 0/5 Odds: (Return: 0)
Mkt filter 0/4 Odds: (Return: 0)

NH Non - Handicaps
Raw ratings 1/7 Odds: 3.5 (Return: 4.5)
Mkt filter 0/1 Odds: (Return: 0)

I'm surprised the ratings found a winner at Fairyhouse to be honest, that they couldn't do the same at Wolverhampton is sad but I'm also not amazed.
Today there's a single winner at Lingfield, 3/1 - I only expect around 1 winner in 5 or 6 so only getting one is not a surprise, it's just a shame when the odds aren't good enough to break even! Never mind, it could all change tomorrow.... roll on the proper start of the flat, let's just fast forward to June shall we? (Maybe stop for a few minutes to watch a few guineas preps mind and the races themselves).

Dave
 

Attachments

  • 4apr.xlsx
    417 KB · Views: 0
  • awcard_4apr.xlsx
    269.3 KB · Views: 0
  • full_ratings_4apr.xlsx
    49.5 KB · Views: 0
  • selections_4apr.txt
    1.1 KB · Views: 0
I made both Irish courses -3.2 for yesterday (I didn't think one was much faster than the other so playing with them would have just been to make it look more accurate rather than actually making it so). The ratings were all over the shop so I can't say I'm very happy with them, but estimating an accurate figure for running through a bog in the mist is probably not an exact science! Frankly I'm glad the racing is cancelled, rather than filling my records with dubious results.

Cork.png

Mike.
 
TheBluesBrother TheBluesBrother
Crikey,
that's a hefty allowance and no mistake - to be honest whilst a spot on figure is the best outcome, when the going is as dire as this I'm not convinced that anything remotely accurate is likely to come out of it.... I had enough problems with figures from a year back being OTT. I don't suppose it's going to make a huge difference, the jumps is nearly done and I'm really just waiting for June!

Dave
 
Well, that was an exciting day - off at 10 am to hospital for the Mrs to have a check up following the surgery she had last Nov/Dec, home for 5.30, total time actually in room with surgeon chappy having a butchers.... 5 minutes. I discovered three things - 1) Reading a kindle for over 2 hours as you travel is a good way to make yourself very sleepy, 2) My Coccyx is not built to support me for the whole day travelling and aches most uncomfortably to tell me to stop doing it, 3) You can tart an NHS cafe up all you like, they still make some bloody strange sandwiches at times.

I am taking a popular homeopathic remedy for my aching Coccyx, called Lager, and the pain is gradually easing.

Right, another epic day tomorrow - let's look at the bad bit first, yesterday's results:
Flat Handicaps
Raw ratings: 1/5 Odds: 3.0 (Return: 4.0)
Mkt filter 0/2 Odds: (Return: 0)

Flat Non - Handicaps
Raw ratings 0/2 Odds: (Return: 0)
Mkt filter 0/1 Odds: (Return: 0)

NH Handicaps
Raw ratings: 0/0 Odds: (Return: 0)
Mkt filter 0/0 Odds: (Return: 0)

NH Non - Handicaps
Raw ratings 0/0 Odds: (Return: 0)
Mkt filter 0/0 Odds: (Return: 0)

7 races, 1 winner at 3/1 - to be fair there's not a lot to be done when you only have one racecard, the strike rate of the speed ratings is only 1 in 5 or 6 usually and so a single winner on a card is not unusual - in fact zero winners wouldn't have been a major surprise - and 3/1 is better than 4/6.....

However, more races today and there should end up being 22 runners I think, there are still 6 to go at Kempton, and so far there have been 3 winners - at 7/1, 12/1, and 10/1, so today kinda makes up for things a bit. (Thankfully). I note that there was a nice 20/1 2nd (Bungee Jump) too, wish I'd been here, sen that, and had an EW on it.... (I wish all sorts of things that don't happen though, it's part of being a man I think).

Well, the sun has actually just appeared - suddenly it looks nice outside, you probably still need a big thick coat and wooly hat like, but at least it looks nice.... me? I'm keeping the heating on....

Dave
 

Attachments

  • 5apr.xlsx
    173.6 KB · Views: 0
  • AWcard_5apr.xlsx
    140.9 KB · Views: 0
  • full_ratings_5apr.xlsx
    26.6 KB · Views: 0
  • selections_5apr.txt
    568 bytes · Views: 0
Tomorrow as ever is attached....
Yesterday:
Raw ratings: 4/18 Odds: 7.0 12.0 10.0 1.5 (Return: 34.5)
Mkt filter 2/8 Odds: 7.0 1.5 (Return: 10.5)

Flat Non - Handicaps
Raw ratings 0/4 Odds: (Return: 0)
Mkt filter 0/1 Odds: (Return: 0)

NH Handicaps
Raw ratings: 0/0 Odds: (Return: 0)
Mkt filter 0/0 Odds: (Return: 0)

NH Non - Handicaps
Raw ratings 0/0 Odds: (Return: 0)
Mkt filter 0/0 Odds: (Return: 0)

As mentioned yesterday that's 22 out and 34.5 back - all from handicaps on the AW. I must say I'm a little surprised that the handicaps find winners but the non handicaps don't, I'd have thought the handicapper's efforts would make stakes races appear the easy option! (To be honest the flat last summer definitely gave a much better strike rate in flat non handicaps).

Today was looking a bit iffy, a single 13/8 winner at Wolverhampton and the 4/7 winner of the first at Chelmsford didn't look that promising, but the 20/1 winner of the 6.45 has just ensured a level stakes profit for the day and made things look much rosier. I've also been pleased at how some of my more recent work has performed today. I said a while back that I was looking at some of the VDW bits to see if they could help thin the field down a bit, and although I wouldn't say that most of the ideas I checked out proved much use (in the short term) I tend to go off things if they perform dismally over say 50 races on the trot. I originally came up with a mix of recent form numbers (ie did they come 1st - 4th last time out, 2nd last, and so forth), a simple 'add up the form of the last 3 runs', the ability rating, class (from race prize money), and two sets of ratings - my own and the RPRs from the Racing Post.

Ability rating, for anyone who doesn't recognise it, is the winning prize money a horse has won (not total prize money, you don't include place money), in hundreds divided by number of wins - you can get the figures from the Racing Post racecard info, click on a horse's name and you get profile info (probably need to have at least the RP Club middle tier membership for this?) In there you'll see prize money from wins and number of career wins. So if your horse has won £76,000 with 4 wins in total it's 760/4 = 190 for the ability.

Class is using the value of the race a horse has run in, and his performance in that race, to give a numeric value to his class - the class of a race is again the prize money divided by 100, If today's race has a penalty value of £6800 it's a class 68 race. I simplified things somewhat as I got my program to run through each horse's record of races to find the highest prize money race it had won, and used that AS A STARTING point to decide on the class of my runner. I also have the data for as many runs as I want of course, the normal daily card lists a yea\r of form for all runners, and I have a shorter note to refer to that focusses on the last 4 runs.

So far I've found a mix of the simple form figures, the two ratings, and the class value have been reasonable at pointing out the runners in a field that have the better chances, so I'm hoping to see this continue - it could easily be yet another of those ideas that works for a week then dies a death after all.

Well, let's see how things go on.
Dave
 

Attachments

  • 6apr.xlsx
    383.4 KB · Views: 1
  • AWcard_6apr.xlsx
    225.4 KB · Views: 0
  • full_ratings_6apr.xlsx
    47.8 KB · Views: 1
  • selections_6apr.txt
    1.1 KB · Views: 0
Well the daily results processing has been fairly quick and easy this week, much of the time is taken up checking and amending NH race times usually, so with almost all race times (barring, strangely, a typo in one race per day all week) being okay I've simply run my programs, filed the results, and sat back to drink coffee.

Tomorrow is going to be different of course, today's Fontwell card will take a little fixing no doubt... I'm really looking forward to the flat proper!

Right, yesterday:
Flat Handicaps
Raw ratings: 0/10 Odds: (Return: 0)
Mkt filter 0/6 Odds: (Return: 0)

Flat Non - Handicaps
Raw ratings 3/3 Odds: 1.63 0.57 20.0 (Return: 25.2)
Mkt filter 0/0 Odds: (Return: 0)

NH Handicaps
Raw ratings: 0/0 Odds: (Return: 0)
Mkt filter 0/0 Odds: (Return: 0)

NH Non - Handicaps
Raw ratings 0/0 Odds: (Return: 0)
Mkt filter 0/0 Odds: (Return: 0)

A complete reversal of the preceding day - with all the handicap races going wrong and all the non handicaps winning = try to figure a trend out of that if you can. It helps to think of swings and roundabouts - if one set/group of ratings goes down the tubes one day you can figure it'll probably see a correction sometime to bring it back into check. You can see this in the course win rates, the AW courses all have 200+ races recorded for them now, and the ratings win from 18% to 20% on all of them - over time, basically, the ratings pick one winner in 5, give or take a couple.

Still, 13 out 25.2 back is okay.
Today as I type there's still half th racing to go near enough, there's a non runner and 4 winners so far, pleasingly 2 of the winners are at nice prices of 6/1 and 13/2, so we're almost at breakeven already - we have about 21 pts back so far on 26 outlay, so lt's hop Dundalk and Chelmsford produce a couple of winners to give another day in profit for a nice change.


Anyhow, I thought I'd pop a bit into this thread showing the 'filtering' I mentioned above, this started out as a look at the VDW way of things - I read a few other threads/blogs on here, and a couple of downloaded files where VDW and others discussed the ideas, and then I added a few lines of code to my daily card producing program to collect the relevant data so I could view it easily.

I should point out that the aim wasn't, isn't, to pick a winner from a race, the object of the exercise was to try to reduce each race to a short list of likely contenders, the runners with the best chance. I was expecting to see races that had 2 or 3 contenders identified that I could then use other techniques to decide from - there would also be races where the field could not be reduced sufficiently, and races where nothing looked likely to win, and all sorts of other things too, so there'd be an end result where some of the day's races were identified as being worth a look, with a fw horses highlighted in each. It would then be up to me to pick winners from them.

My expectation was that I could perhaps give myself a good chance in 50% of the races that popped up - my shortlist idea is not expected to do anything extraordinary, I don't expect to find I have the winner in my race shortlist 80% of the time - it'd be nice if that did happen, but I'll settle for a method that tells me where to look for the winner 50% of the time.

I've only trialled this for a short while, but it's been encouraging, so I'll carry on with it for a while to see if it's flash in the pan or worth continuing with - it's important to realise that sometimes things work for a while because of the weather, or some other influence, and as the weather breaks so the method stops firing....

Right, a bit later on I'll post a few race cards to show what this looks like in practise - this part of the job is fairly fast, it's picking the winner from the shortlist that takes the largest part of the time.

When I post the cards I'll explain each bit, so you can try it yourself or try fiddling with it to your hearts content!

Dave
 

Attachments

  • 7apr.xlsx
    453 KB · Views: 0
  • AWcard_7apr.xlsx
    148.9 KB · Views: 0
  • selections_7apr.txt
    1.7 KB · Views: 0
  • full_ratings_7apr.xlsx
    59.5 KB · Views: 0
Okay,
I've attached a spreadsheet called 7apr_working which has the race cards I will describe, I've also uploaded a CSV version anyone who wants to follow this but can't open Excel files for whatever reason, it will be harder to read the CSV as the formatting won't be as planned.

Anyhow - I'm going to explain the main points of some of the races in the spreadsheet, there's nothing over clever here or mystic/magical, hopefully you'll see there's a bit of logic involved. You can do the same yourself, if you can't grab data too easily and end up having to input the numbers yourself it's still doable - you just have to limit your efforts to a few races, and need to pick races that are most likely to give you something to work with.... which is my first point:

1) Finding races to work on:
If you run down the listed races you'll see some have lots of highlighting on them, some have little, some have none at all - the highlights show figures we are interested in, so if there is a race with no highlighting we are going to ignore it. The very first race (Kempton 1.30) is one such race - look down the form column, 4 of the 10 runners have a single form figure, 6 runners are previously unraced. Unless you know something the rest of us don't then we'll skip races like this - unexposed horses could do anything, and frequently will!

- So the 1410 at Navan, and the 1430 Uttoxeter are also going - there are too many relatively unexposed runners in each, and we don't NEED to do these races so why take the chance that one of the unraced or lightly raced runners will pop up and win?

2) Find all the runners that have placed 1,2,3 or 4 in either their last race or the one before that.
Simply run down the 'form' column and highlight any horse with a 1,2,3 or 4 in the last or 2nd last time out position. (Note - VDW suggests that if a horse fell, or suffered a similar mishap, you should skip that and go to the next placing - ie a horse reading 561F7 is considered to be a 5617, ignore the F... the idea here is that it may fall, unseat, or whatever through no fault of its own so should be given a bye. My personal opinion is I will do that for occasional falls etc, but if I see a form line reading FFPU6RF I'll still consider it to be failing my check - a horse like that is probably incapable of jumping!)

- I left you a practise race to try out, you can see how I marked up races 2 and 3, but 4 is blank still - have a go at it, answers below!



Race 4 then, you should have marked them all of course.

***The 'Last 3' column totals the last 3 form placings (placed 0 = 10) and is not currently being used - I'll cover this and the Ability column, which is also not being used, later as I finish off.

3) Check the ratings
The SPD column shows the best speed rating (my own ratings) from the last 4 runs for each runner. the RPR column shows the best RPR recorded from the last 4 runs.The 'Added' column is the one we are going to use here. The speed rating alone has about an 18% strike rate, if we stick with top on speed or maybe the top 3 we will probably throw winners out too often. The RPR column however is almost the opposite, RPRs quite often show little between runners, so taking the top rated and those within a point or two will often leave 3/4 of the field there, so we combine the two to get a figure we can use to set a cutoff value -
to save working it out, the added value is the best of last 4 speed figures added to the best of last 4 RPR's AFTER that RPR has been multiplied by 0.85 - ie 85% of the best RPR is added to 100% of the best soeed rating.
A cutoff is calculated - the cutoff is simply 85% of the highest combined rating, rounded down - so in race 2 the highest added rating is 109, 109 x 0.85 = 92.65, round it down and our cutoff value is therefore 92.

- Highlight all of the combined ratings that are equal to or greater than the cutoff for the race.

4) Check class
Look at the race information, the prize money for the race is listed there - for race 2 it is £8837,. for race 4 it is £15562, and so on.The class of the race (for want of a better term) is the race prize value divided by 100 (rounded) - so 8837 is class 88, 15562 is class 155, and so on.
The class column shows the highest 'class' race that each runner has won in the past year - again this is from my program, but can be found from the horse information on the Racing Post and others.
We highlight slightly differently here - using two colours - the usual yellow shows a horse has a class rating equal to or better than that of today's race, I use the orange highlight to show the horse has a win but at a lower class.... I think any winning form needs to be considered, we don't want to ignore possible improvers after all,. but the yellow shows us who has already cleared the bar.

5) Now look at it and make your mind up

I'll go over a couple, then let you have a go, then finish off....

Race 2 - everything is highlighted, no help at all then! You can look at this 2 ways, 1 is that everything is pretty exposed and you should check the form to see if one of them is improving - after all none of them has matched the class of the race as yet, or 2 you can say they are too closely matched and skip the race. There IS one obvious point, Teddy Tee's rating of 198 compares to the rest of them who are all on 174 or 175 - that's a big difference in ratings, so I would say the idea here would be to check Teddy out, if he has the form that suggests he could win and no obvious negatives then he may be the bet.

Race 3 Now then, what we want is the runners that have highlights in all the columns, so we'll not be looking for Dartagnan Le Dun to win this despite his rating - we hope - failing that "form of 1-4 in either of last 2 races" check is not something we expect the best chances to do. Hartside is through on all checks so goes into our list, Akula and Flanagan's Field miss on class but have good ratings, and would go on the list as potentially able to take this. When I note these down I use a dash to denote a runner that doesn't quite hit all the marks, so I know to double check him.
From race 3 I have Hartside, -Akula, -Flanagan's Field
(Akula is russian for some type of shark by the way...., also the name of a Russian submarine class)

Race 4: The 1-4 filter does nothing to help, we have 4 obvious possibles, as they match the rating cutoff check and the class check - three of them would need to improve and have orange highlighted class, so I'll mark them as -
Possibles: Tomily, -Tarboosh, -George Dryden, - Dynamo Walt

One more for luck -

we skip race 5 as noted early on,
Race 6: Gregarious (all highlighted), Walsingham Grange (all ok), Ardmayle (all ok), Excellent Team (all ok)

In this race we have 4 full qualifiers, Ashkoul's class is orange and rating is not high enough, Capitoul's rating is okay but has no class rating and the form 1-4 check failed also. I would look at these 4 for the winner, which still leaves half the field of course.

If anyone wants any of the other races covered, or has questions about the ones I've done, then just post a query and I'll try to cover it.

Remember the plan is for the possibles to include the winner 50% of the time, if the winner is in there more often then great - if less often then we keep trying to hone this down until it does work reliably.

*** I said I'd explain last 3 and Ability - if you take the form figures and add the last three together you get a sum that can vary from 3 (111) to 30 (000) - places down to 9 are added as they are, anything 10th or worse is a 10, and you ignore falls etc. If you have only 2 runs then you add the last run on again, ie a 14 is treated as a 144. The idea of this sum is to take advantage of the way that horses with the last three form figures showing as 111, 112, 121 and the like (basically in the first 3 each of the last 3 runs) tend to win a very high percentage of races. Lately they haven't been as great a pointer as they should, so I;m watching this rather than using it right now.

Ability I explained before, but I'll repeat it quickly - Ability is the win prize money (do not include place money) won by a horse, in hundreds of pounds, divided by the number of times it has won. So if a horse won 5 times and the prize money for those 5 wins added up to £96000, th sum would be 9600/5 = 1920 . I am not convinced that this rating is as useful as suggested, a horse that won twice three years ago and hasn't even placed since is not, in my view, the horse it was - ability is, to my view, not sufficiently focussed on good form to be used with confidence. Of course I reserve the right to think differently at the drop of a hat....

Okay, that's it, the next job would be to look over the possibles lists, pick the races that look like the filtering might have produced something worth using, and look up each of the horses in your list to decide which of them you like most - if you cannot make a case for one of them move on to the next race. Don't ignore the other runners - you still want to have an idea of how they will run, as that will affect your pick - but you are not looking to build a solid winners case for them.

What I look at for my short l,ist:
Trainer - 14 day win rate, AE ratio, also do I know they are good/poor at this track (easily found in RP)
Jockey - 14 day win rate and AE ratio. Also, has he/she ridden the horse before - if so did they place or win on it? If ridden before did the horse run well for them compared to when other jockeys rode it?
Distance - Distance wins and places, preferably without negative comments, I want to see that they were able to challenge at the end of the races they ran over this distance.
Going - similar to distance really, win or place form on the going
C/D wins - nice to see
track - check the formlines, is there anything blazingly obvious there, such as the horse clearly only goes on Left or Right hand tracks.... don't forget the shorter sprints tend to be straights mind!
Draw - use a stalls check if you can to check that runners are well/badly drawn - not an issue over 2 miles of course, but important for sprints at least.
Comments - you don't want anything overtly negative, 'reluctant to race' is not a good sign. Just a simple check - if you are good at putting the picture together for a race then great, but you can at least check that the horse appears to race fairly willingly.

Pace info (AW racing only) - for sprints and up to about 8f I want to know where the early pace is coming from, a sprint ought to have at least one obvious early leader but by the time you get to 7 or 8f you can find that nobody wants to go on - an 8f race where they dawdle for 2f because they all want to be covered up is a 6f sprint with a lengthy preamble! As distances increase I look more at the AP ratings. If a sprint and a bend is involved before the final run in then I like to check speed around the turn - all these figures are in the AWcard each day for AW races.

The list is not absolutely complete, you may have all sorts of ideas too, but this is what I look at - you'll hopefully have noticed that I do not simply look at my ratings and call it a day. Racing is a very complex puzzle that constantly changes, you have to work hard at it to get anywhere. (Or be very very lucky).

One other thing - assuming I have now got the race sorted and picked my main choice for the race and possibly a backup for if the number one choice doesn't run, or perhaps I want more than one in the race for some reason, only then do I check the forecast price and compare it to the idea I will have of what I think are fair odds.... I will bet on a horse if I think the odds are a bit tight, or if I think the horse is better than the odds, but I won't back a horse if I think the odds are noticeably shorter than they should be and I hardly ever bet odds on.

If anyone has questions, ask
Dave
I forgot to mention OR - especailly in handicaps it's useful to see if a horse is going up or down in the ratings - as horses move from class of race to class their alloted weight will change, but the OR will tell you if they have been penalised or given an easier time of it - ie it isn't the weight that matters (other than lump a piano on it's back and it isn't going to jump too well) but what the OR is doing. It is often useful to compare its OR today to its OR the last time it won.
 

Attachments

  • 7apr_working.xlsx
    19.6 KB · Views: 13
  • 7apr_working.csv
    6.9 KB · Views: 1
Last edited:
Okay then, Sunday tomorrow so not much on the cards - see attached.,

Yesterday:
Flat Handicaps
Raw ratings: 3/12 Odds: 6.5 3.0 2.75 (Return: 15.25)
Mkt filter 3/6 Odds: 6.5 3.0 2.75 (Return: 15.25)

Flat Non - Handicaps
Raw ratings 0/4 Odds: (Return: 0)
Mkt filter 0/0 Odds: (Return: 0)

NH Handicaps
Raw ratings: 1/5 Odds: 6.0 (Return: 7.0)
Mkt filter 0/2 Odds: (Return: 0)

NH Non - Handicaps
Raw ratings 1/2 Odds: 1.63 (Return: 2.63)
Mkt filter 1/2 Odds: 1.63 (Return: 2.63)

23 out 24,88 back - this was very annoying because it went from well up \t the end of the afternoon to little better than even, as the evening racing produced one short odds winner. Today is worse, as only the fakenham results have thrown winners up really - 4/7 at Fakenham, 2/7 at Uttoxeter is at least reasonable for strike rate, but zeroes at Kelso, Kempton and Navan... and Wolverhampton is also about to be a dead loss by the look of things. What winners there are have been at short odds, so it's a bit of a caning I'm afraid.

Anyone who followed my 'weeding out' of the card yesterday should have seen that around half the races produced a winner from the highlighted runners - while this wasn't a case of winners jumping off the page it did as planned - 50% of races with a winner in the shortlist, and around half of the runners eliminated from each race automatically. Given the odds they were coming in at (Keyboard Gangster was at 28/1, the longest priced, others were also at long odds though) the point is that winning long shots were not filtered out - they were in the list of runners to look more carefully at. You'll also have seen why I use two colours for class, yellow for those that have won at this level or higher - these runners will always be of interest - but the orange to show a step up is needed (based on 1 year's form remember) but, crucially - as I said yesterday -

- ' I think any winning form needs to be considered, we don't want to ignore possible improvers after all,. but the yellow shows us who has already cleared the bar.'

I think today was a difficult day to pick winners for the average punter, the odds being returned rather suggests that anyway, and if you can erase half the field in 30 seconds with decent reliability then it's worth watching to see how this gets on in the longer term.

Just to show my fallibility is well established I took a closer look at a few of the TV races and managed to pick:
Baywing (2nd, 11/2) Takingrisks (3rd at 9/2) Tomily (5/2 3rd), Boundsy (nowhere, 13/2) and Shenanigans (6/1 2nd) = I was particularly peeved when Baywing clouted the last 2 fences when he looked to be motoring to an easy win, and also by the fact I pretty much flipped a coin to decide between Boundsy and the actual winner Raven's Lady..... it's fair to say that I managed to hit the crossbar with just about every horse I picked.

Oh well, if it was easy (a) Everyone would be doing it, (b) Bookies wouldn't take my bets. The fact that they still welcome me with open arms tells you all you really need to know.....!

Dave
 

Attachments

  • 8apr.xlsx
    178.2 KB · Views: 1
  • full_ratings_8apr.xlsx
    30.9 KB · Views: 0
  • selections_8apr.txt
    609 bytes · Views: 0
I thought I'd put up today's 'winnowing' spreadsheet, to see how it does again -
I've called it the VDWsheet but it's really an output from my card writing program that I am feeding various VDW style items out to, rather than any attempt to follow the chap's method (I've yet to find anyone who knows what 'the method' actually is!)

As before I highlighted potential contenders - the aim is to have a shortlist for each race which contains the winner, and to do that for 50% of races covered. A runner gets added to the shortlist if they meet the 3 criteria of having (A) 1-4 in their last or 2nd last run (where a runner shows a fall, brought down, pulled up etc I will skip this and go to the 3rd lto, but if a runner's form is chock full of F's and P's I'll leave it off the list as being unable to complete a race due to problems with the horse, rather than calling it a bit of bad luck. (B) The combined rating calculation must equal or better the 85% cutoff value (seen below each race), and (C) It should have won in the past 12 months, preferably at the level of today's race. Today's race prize money divided by 100 is the class of the race, in the final 'class' column you can see the most valuable win the horse has managed since last April, also in 100's of pounds. So if the horse has a class value equal to or better than today's race class value then it passes the class check - a win at lower value than today means the horse needs to step up a bit today, so when evaluating form you want to see some suggestion that the horse will be able to do that - a runner with a class value below today's is still in the shortlist, it just means you need to double check whether you think it's up to stepping up a bit or not.

Several races have runners with insufficient data to place any faith in this fitering, these races are skipped - on today's cards they include the 1405 at Exeter, and the 1715. At Carlisle the 1730 is skipped.

Also, the 1450 at Carlisle has a shortlist of 3 of the 4 runners - it';s fair to say that the filtering process is pretty immaterial in that race, the 1555 is pushing things a bit with 5 of 9 left in - I'd take the help but would skip this as a betting proposition unless, when checking the from of the runners, I could quickly remove at least 2 of the list. (I won't count one of the 5 winning as a success for the method, in other words, as cutting a race filed in half isn't that spectacular!)

Exeter 1510 is similar, 4 of 6 still in the list.

For the Carlisle 1450, 1555 and Exeter 1510 my thinking would be that I'd probably skip these anyway - this method of reducing a field to a shortlist is based on a statistical approach to the problem, so I'm taking the view that if it is in the shortlist then it has a chance of winning - a race with 6 live contenders in is not as good as a race with 3 in, in my view, so I'll take the easier problem to solve thanks - there are enough races left with a small number to choose between so I don't need to tackle the more difficult ones.

Anyhow, there should be a couple of results there already, for fairness only count the times this works from 3pm onwards!

Dave
 

Attachments

  • VDWSheet_8apr.csv
    7.3 KB · Views: 9
That's another VDW bit if I recall correctly :) the progression of speed figures v class is certainly worth looking at but I see that as the sort of job that needs to be done in the second stage when the full card info is used to look at the shortlist in greater depth to decide if a bet is present.

I'm probably not going to touch this for a bit, I want to have a good feel for what's working before I change things - my original setup made use of the last 3 places and ability and I decided they weren't helping, and I've only been running this version for about 5 days. I'm very deliberately not trying to analyse too much - if I can cut a race down to 3 or 4 runners often enough then I'm achieving my aim, I distrust races where this only produces 1 or 2 runners from the whole field.

Dave
 
Back
Top