Ratings and Queries - Suggestions please

[BillyHills]

With all the knowledgeable people that visit these shores nowadays i was wondering if any of you could offer some suggestions to a couple of things i'm working on at the moment.

Basically, without mentioning trade names, i have got use of some software that allows you to calculate your own Ratings and also ask the database a series of 'queries' to help back test systems.
On both sections i would like to know your ideas on a couple of things, bit of a brainstorming session i suppose.

Ratings
1) Not sure if anyone has experience in this area, but on flat racing how many pounds per length would you calculate the ratings, i hear many say 1lb = 1 length and longer on soft ground.

2) I see many horses these days win first time out after a lengthy lay off and quite a few rated races are headed by a horse that hasnt ran for ages, what would you say the cut off point would be to take a horse seriously on his comeback?

3) Horses that run on the AW always seem to be well handicapped when they return to race on turf, a lot of top rated animals have had their last run on the AW, would you discount AW form or not?

4) Some people love speed figures, do you regard this as important, i always think it depends on too many things, pace of race, how many runners etc...to be accurate enough. In the states with the split times its quite different.

Queries
With a massive database its quite possible to enter almost anything and get a result to your query, the usual ones are;

% of winning favourites
Top weights in H'Caps
C & D winners at a track
Record of a trainer in a particular type of race
etc........

All these can be tested to see how they would have done to level stakes over a period of time, very handy if you have devised a system.
Also does 10yr trends, do you think these are important or does each race have no concequence to what has happened before?

Just wondered what you would ask as a 'query' or what would be a good type of race/system to back test.
You can test things out on past races or future decs.

Feel free to ask anything and appreciate any input.

BH

[BillyHills]

Just to give you an example of the 'queries', ran a default one on the following;

Odds On Favs in Handicaps on the flat over 5f/6f since Jan 1st 2006

Runners = 49 (last one Rasaman today)
Winners = 31 (63%)
Longest losing run = 3 (3 times)
£1 level stakes = + £3.83

As you can see, although this is a broad example it could prove whether a system is worth following or not, the losing sequence especially valuable to those using % of bank staking plans.

[Trilobite]

4) Some people love speed figures, do you regard this as important, i always think it depends on too many things, pace of race, how many runners etc...to be accurate enough. In the states with the split times its quite different.

I couldn't make a selection from a sprint without my Speed Rating

But you are correct, pace, draw etc need to be considered.

[whoneedsrio]

Originally Posted by BillyHills

Ratings
1) Not sure if anyone has experience in this area, but on flat racing how many pounds per length would you calculate the ratings, i hear many say 1lb = 1 length and longer on soft ground.

2) I see many horses these days win first time out after a lengthy lay off and quite a few rated races are headed by a horse that hasnt ran for ages, what would you say the cut off point would be to take a horse seriously on his comeback?

3) Horses that run on the AW always seem to be well handicapped when they return to race on turf, a lot of top rated animals have had their last run on the AW, would you discount AW form or not?

4) Some people love speed figures, do you regard this as important, i always think it depends on too many things, pace of race, how many runners etc...to be accurate enough. In the states with the split times its quite different.


BH

1.
The basic weight scale applied to the handicapping of horses for the flat is,

5-6f 3lb per length
7-8f 2&1/2lb per length
9-11f 2lb per length
12-14f 1&1/2lb per length
15f+ 1lb per length

This is for good ground, on fast ground it would need a slight increase & on softer ground a slight decrease, as with most things in this sport it is not an exact science.

2.
Training methods have improved & now most trainers can get a horse fit enough to run well FTO even after a long lay off.
Much depends on the race the horse is aimed at, if it was say a decent prize I would tend to look at the horse closer than one returning in a more modest event where it could be just having a pipe opener for something later on.

3.
The usual reson for a horse being well handicapped on turf compared to aw is it's turf form is not as good, this works the other way as well of course.
I tend to discount aw form when looking at a turf race.

4.
In the States most races are run flat out from the off, over here the only truly run races are at 5 & maybe 6f any distance above this & as you say much depends on the shape & pace of the race.
Just because a horse as not won a race with a fast speed figure dosen't mean that it can't, however a race with a fast time/speed figure & the form is usually pretty reliable.

Rio.

[Carl]

my 2pence worth,

1,I often work about 1lb per length on soft and 1lb per ½L on good/firm-firm

2, I once asked this question and got told the ideal time is 21 days off the track. But like yesterday Id pick a 40 day off the track horse over a 275 day off but as you say some trainers like the runners fresh so really I suppose its down to if the horse has run well fresh before. Bit of a grey area tbh.

3, Again I speak to a old bloke who really knows his stuff as he was a bookie etc and he said that a lot of horses dont like going back to turf once they have done well on the AW, Id rather pick a runner going to AW off the turf rather than AW going onto Turf at least until they have had a spin back on the turf. But form wise I would inc it in the form etc specialy if its running on good/firm.

4, I aways look at ATR website for the speed ratings etc but its not the main factor in my own selections but it can swing a 50/50 decision.

Redcar 3-30 Handicap (Class 5) 7f

LTO Speed: Satyricon 81 Fun In The Sun 76 Onatopp 68
SATYRICON (nb) rates top on the clock following his fast-time third to Sweet Gale on the turf at Lingfield last week, and is a proven performer over an easy seven furlongs on good-to-firm ground. A winner over this trip on the all-weather at Kempton in February, he has improved a fair bit since then, and still seems to be on something of an upward curve. Fun In The Sun rates close up, but could find things happening a little to quickly on today's firmer racing surface.

[RussP]

Interesting thread, Graham.
Will come back to it later, as I'm off out shortly

[BillyHills]

Thanks guys for the interest and comments.

Rio, always looked at things on a 1lb/length scale, thinking about it, it makes sense as ive always done NH ratings before which would be about right, thanks for the info, totally overlooked that when dealing with the flat.

Personally hate it when i like the look of a race to sort out only to find one or two of the fancied ones have been on the AW recently, rarely do they translate the form and why should they really, but it just spoils it for me.

Never been interested in speed figures, too volotile in this country with all the different going and different tracks, good tool to just back you up on certain things like Carl says.

The absence from the track is a real puzzler, i know Alan King had one win from 3 years off the track but how can you take their ratings seriously from a run so long ago, again spoils the race make up for me and usually ignore the horse or in fact the race.

Be interested to hear why people think Trends are so important, i know one or two on here swear by them but sometimes you end up with 6 qualifiers, only one wins, can you still call it a successful way to find winners? I also think after a horse has won you can always look back and see why, you can fit info into anything if you want it to.

Anyone got any systems they want to tell us about, i'll try and back test them when i find my way around this database, just let me know.

Thanks again, nice to have an intelligent chat about stategy and stuff rather than just threads with selections in them all the time.

BH

[bowles10]

Originally Posted by BillyHills

Be interested to hear why people think Trends are so important, i know one or two on here swear by them but sometimes you end up with 6 qualifiers, only one wins, can you still call it a successful way to find winners? I also think after a horse has won you can always look back and see why, you can fit info into anything if you want it to.

graham, as you know i like the trends. I try to use them in the larger fields & find i do better picking the winners of 20-30 runners than 10-12 runners. Ok, on occasions i may have 6 running for me but i try to add a few more key trends in to narrow it down even more.

Big meetings are where i like to check the trends & sometimes my trend analysis is different to the RP one for example. 2 years ago i picked 2 50/1 winners at cheltenham, this year i picked another 50/1 shot but failed to put it up as a selection (doh!) even though i only had 1 other selection in the race. I also picked a 28/1 & 22/1 winners.

I will try at Ascot if i get the time to do the big handicaps, Wokingham, Hunt cup, Brittannia etc

Great discussion thread.

I also use the RFO for thier sprint ratings & time figures. For similar reasons, to dismiss certain horses.

[BillyHills]

Rob, thanks mate, interesting stuff.
I'll have a look at the trends at my disposal during Ascot and will add my comments if you do any threads on the big races.
Like i said never really took much notice of trends as such but you cant argue with your success, i know Jon likes them too.

Thanks

[Cassius Marcellus]

For whats it's worth...

I knew a man who bet on handicaps back in th 1960s. He told me that when you have narrowed a race down to 2 or 3 possibilities with the weights, draw, form, etc, etc, always back the one with the weight nearest to the weight of the jockey.

His reasoning was that if a horse had 10 stones and another horse had 8 stones and they were almost level on his ratings, the lower weight could be moved by the jockey being up over the horses neck, while the 10 stoner had 28 pounds of it stuck on his back in the saddle.

I don't bet in handicaps much, so if it works, I have absolutely no idea.

[RussP]

Graham,
Great thread this. Personally, I don't look at speed ratings at all and am always dubious about horses moving from AW to turf. I used to discount horses that have been off for a while, but have since realised that more and more are coming back from long absences to win. With regards to lbs per length, I think Rio has it spot on for the flat.
Now, I like to concentrate on big field handicaps (11 runners +), and like to bet E/W at double figure odds (if poss). The first thing I look for is a weak favourite. If you can find one of those, you have possibly found a race to bet on. There are too many false favourites in flat handicaps, reasons can vary from:
- the 'trainer' or 'jockey' factor
- up significantly in Class
- up in handicap mark
- winner lto
- wrong trip / going / track
- bad draw
- generally, 'hot gossip'
I then look for something to oppose that fav with. One major consideration has to be given to the draw, even in races over 1m+. Never discount the strength of this. I look for a horse with proven form over trip / ground / grade. One such example is given below, taken from Sundays racings, resulting in a 11/1 winner:

5.10 Bath
Class 4 handicap with 13 runners over a trip just short of 6f:
Roman Maze (11/1 Totesport) - I originally liked Mine Behind (as per AK) but lack of recent form puts me off a little so I have settled on the Ryan Moore ridden Roman Maze. Conditions seem right for this horse and a very good jockey booking to boot. Best at around 6f, loves firm going and very capable in this grade. Some stats to back this up: 4 wins & 5 places from 16 attempts at around this trip, as opposed to 3 wins / 6 places from 36 over further. 3 wins / 5 places from 17 on GF or faster, against just 2 places from 16 on good or softer. 6 wins from 31 in Cl 4 or below, 1 from 20 Cl 3 or above. Form is solid too, excuses last twice a badly hampered lto and trip / ground wrong the time before, but previously looked very good when winning at this trip at Windsor 3 starts ago. Excellent EW claims in what is a poor Cl 4 heat.

Looking at those stats AND discounting the previous 2 runs where there were valid excuses, this horse should not have started at anything longer than 5/1 or 6/1.
In handicaps, the better horses usually prevail (ie, those at the top of the handicap) so I always start there. This is another key factor for me, although I like those just below top weight (2nd / 3rd / 4th in the weights). Conditions, though, are key as the horse in form is not always the winner....look for horses with solid, relatively recent, form over todays conditions. C&D form is also important, as is the trends. I usually check the trends as a back up to my selection. Handicap mark is another key stat. Why should a horse raised in the weights because of a good run lto (ie, 2nd / 3rd), win nto? I fail to understand this. Saying this, I am not one for horses dropping in weights rapidly with no recent form. I like to see some traces of ability before investing. I generally discount 3yo's racing against older horses, I think the stats would prove this to be correct in handicaps. Also 2yo and 3yo handicaps are a no-no, too many potential improvers to make a sound form judgement.
Just my thoughts so far
Graham - any backtesting in handicaps that throw up any interesting stats, please let me know. Thanks

[Trilobite]

Interesting and thought provoking thread.

It also shows how we can end up with the same selection but from completely differing angles.

I too was on Roman Maze, as my DHRC pick that day proves.

However, i got to that conclusion based on my own Speed Ratings, for which i can filter for course, going and distance so that all the times for the runners are compared as much as possible like for like.

Like has been said already though, everything else has to be considered too.



Graham - is the Database/Software available for download anywhere ?

[BillyHills]

Russ, must admit i used to swerve big handicaps like the plague, especially sprints, i was in the school of thought that every horse has its day and they all beat each other at various times and it was just lucky if you happen to pick the right day. A horse running today called City For Conquest (4.20 Bev) is a perfect example, i know it will win one day at a good price but how many time do you follow it?
More recently ive had a fair amount of success, especially over hurdles in handicaps, my profit in the EW Fancies has come mainly from these races, and now find them a challenge.
Like the way you methodically go about your work and it obviously works for you, i think people need to realise with theses races the strike rates aint gonna be great but the profit still can be. Thanks for the insight!

Trilobite, spot on with the point that it takes all sorts of different angles when looking at a race and just because you do one method it doesnt mean you'll be on a different horse to Russ or Bowles or anyone else.
Your Own Speed Ratings! must have took ages to set up, admire anyone who devises their own ratings, whether speed or weight based, i did my own ratings on Novice/Maiden Hurdlers, concentrating on class more than anything, tracks and races they were entered were heavily filtered, it was ok but in the end the time i could spend on them wasnt enough. With the database you can put a Rating in for the winner and will work out the rest of the race for you after putting in certain preferences. Which is why asked the question about pounds per length, you can also allow for WFA scale and Jockeys allowance, thats another question do you guys take into account the jockeys allowance, when looking at Ratings?

Trilobite, its not available free, let me put it like that, ive got use of it for the summer courtesy of my Dad and Brother wanting to give it a go and i get access to it. How successful are your Speed Ratings, are you happy with the filtering, how do they compare with the 'trade' speed ratings like Topspeed etc...do you get many that result in the same selection?

Russ, anything in particular with the handicappers, i'll have a play with some different theories later tonight.

[Trilobite]

Trilobite,
Your Own Speed Ratings! must have took ages to set up, admire anyone who devises their own ratings, whether speed or weight based, i did my own ratings on Novice/Maiden Hurdlers, concentrating on class more than anything, tracks and races they were entered were heavily filtered, it was ok but in the end the time i could spend on them wasnt enough. With the database you can put a Rating in for the winner and will work out the rest of the race for you after putting in certain preferences. Which is why asked the question about pounds per length, you can also allow for WFA scale and Jockeys allowance, thats another question do you guys take into account the jockeys allowance, when looking at Ratings?

Trilobite, .... How successful are your Speed Ratings, are you happy with the filtering, how do they compare with the 'trade' speed ratings like Topspeed etc...do you get many that result in the same selection?.

My ratings/software is still in the very early stages of development.

I download the results of all the flat racing in the UK (not Irish) everyday, and give every horse a speed rating, and update each horses entry.

How good are they ? Dunno really, time will tell


Here are my top ratings for the race Carl mentions at AtTheRaces website above, 7f only filtered

Pretty similar results i would say :

95 7f Satyricon Lingfield (TURF) Good to Firm

93 7f Coconut Queen (IRE) Catterick (TURF) Good, Good to Firm in places

87 7f 2y Coconut Queen (IRE) Chester (TURF) Good

85 7f Satyricon Lingfield (AW) Standard

82 7f 26y Teasing Warwick (TURF) Good to Firm

82 7f Rainbow Fox Southwell (TURF) Good

82 7f Celtic Change (IRE) Newcastle (TURF) Good to Firm, Good in places

And here are the ratings i gave them for the results that night :

94 Satyricon
94 Celtic Change
88 Coconut Queen
87 Cassie's Choice
87 Sea Rover
86 Rainbow Fox
85 Onatopp
81 Teasing
79 Mandy's Maestro
78 Naughty Thoughts
78 Here's Blue Chip

[BillyHills]

Trilobite, looks pretty good mate.
Do you use the 'best' figure as your yardstick or an average figure over say 3 runs on that going. I'm always sceptical about using 'best' figure as it can have been some time ago, when do you let previous form fall off from the calculations.
I know some 'form' experts use top ratings from over 2 years ago, for me thats too long since the horse has reproduced his best and pretty unlikely it will again. Similar question to how long you give a horse after a lay off i suppose.

I'll have a look at the speed figures i got for that race later.

Thanks, BH

[Trilobite]

I use a mixture of LTO, an average (from 2,3 last races) and top speed.

But more importantly to me is when a horse appears at the top (ish) of the filtered list several times.

I can also bring up any horses records and see all the races i have recorded for it, with race details + comments. So i can see if any top speed was years ago or last week and if its likely to make the break and set the pace, try to make all etc etc.

[Remmer7]

I wish i knew something about anything, it all sounds very interesting, but I have no idea where to start in really studying horses, I have a lot of time to do it, and will so for the next few months, so if anyone has tips of how to get started or where to get more info it would be much appreciated, and I will be more than happy to do detailed studies of several races each day to post here where I would be greatful for feedback in my first few attempts

thanking in advance


...and great thread billyhills, its interesting to see peoples different points of view

[BillyHills]

Remmer, as you can see from this thread, i asked a few polite questions and the guys around here will always be willing to help and suggest things without taking the mickey out of 'novices'.

Ive been studying betting for 30 years and worked in the business for at least 10 of them, its hard to know where to start sometimes but reading the Racing Post columns online, looking at stats to familarise yourself with the best trainers and jockeys, reading the form and take notes of the comments in running.
Try to watch as much as possible, i record every race, every day and watch them later and rewind them to see if anything catches my eye for next time, i'm always very interested in why a paticular horse didnt win, its not all about winning today, gathering info for the future is just as important, if it helps write notes, draw up a list of horses that caught your eye for next time etc..
When i started i just concentrated on certain races, whether it be Maidens or Handicaps or whatever, its easier that way to keep tabs on things, there is far more racing now then when i used to watch them but on the plus side there is much more info available, take advantage of it.
Dont be afraid to ask questions, better to ask if your not sure as it can be a costly experience learning this game.
We dont confess to be 'experts' but we will try and help.

Good luck

[Remmer7]

Wow, thats great, exactly the kind of response i was looking for

Thanks very much mate for taking the time to r