The home of intelligent horse racing discussion
The home of intelligent horse racing discussion

Speedratings

Home Forums Archive Topics Speedratings

Viewing 13 posts - 16 through 28 (of 28 total)
  • Author
    Posts
  • #99165
    Daylight
    Member
    • Total Posts 369

    My speed ratings are a little bit more basic and the backbone of the program although there are a hundred other things that have been added and disregarded over the years the main platform has never really changed much. It is my own design and one that has served me well over the last 3 seasons over distances 5 furlongs to a mile in competative handicaps:<br>

    • <br>
    • Expected Time in Seconds = (Standard Time + Going Allowance)

    • Lbs in One second  = ((30 / race distance) x (15 / race distance))

    • Expected time in Lbs = (Expected Time in Seconds x Lbs in One second )

    • Actual Winners Time in Lbs = (Winners time x Lbs in One second)

    • Beaten by winner in Lbs = (distance beaten x (15 / race distance))

    • Actual time in Lbs = (Actual Winners Time in Lbs + Beaten by winner in Lbs)

    • Raw rating = ((Expected time in Lbs – Actual time in Lbs) + weight carried in Lbs)<br>

    <br>That’s the calculation from what I can remember – it may look daunting and complicated but it isn’t really. This was all done from memory so don’t quote me on this as I may have put a wrong figure in somewhere along the lines, sorry can’t be more exact without searching my program for a few hours (which I haven’t got right now).<br>The going allowances I can’t remember but I think if it’s heavy going I add 11% of the standard time on.

    #99167
    guskennedy
    Member
    • Total Posts 759

    To paul101…the following is an extract from something I wrote a few years ago explaining how I calculate a going allowance:

    "How is the going allowance calculated for any particular day’s racing at any particular track?  The first step is to sub-divide the returns into individual ‘courses’ at the track e.g. the straight course and the round course at Doncaster, the sprint course and the round course at Sandown etc.  Don’t assume that the going allowance on each of these ‘courses’ will be the same on any given day.  Next, carry out the calculations previously set out.  This leaves you with a ‘raw’ timerating based on a theoretical ‘nil’ going allowance.  

    As an example on 16th July 1994 there were five races on the straight course at Newmarket ( July course).  A ‘nil’ allowance would produce the following timeratings: Jawaal (official rating 73) 129, Lion’s Mane (no official rating) 100, Bintalshaati (official rating 88) 129, Princess Oberon (official rating 70) 96 and Loyalize (no official rating) 120.  Remember that there is a 14lb difference between the two scales so normally Jawaal could be expected to produce in a truly-run race a timerating in the region of 87.  A timerating of 129 would represent improvement of 42 lbs!  Although Jawaal was certainly a lightly-raced and improving colt at that time that degree of improvement is just not feasible and so we can clearly see that conditions on that day were fast and a ‘minus’ going allowance is indicated  but to what degree?

    Esentially what is involved at this point is exercising a degree of judgement based on experience, looking at the timeratings which would be produced by various going allowances and reaching a ‘best fit’ decision i.e. which going allowance best fits what we know about the horses which raced on this particular day?  Anyone who remembers slide-rules will understand what I mean when I say that I am in effect sliding a cursor along the range of possible options to come up with as accurate an allowance as possible."

    [to be continued]

    #99168
    guskennedy
    Member
    • Total Posts 759

    Paul101…it’s going to take me too long to retype everything I typed all those years ago so I’ll change tack by analysing Saturday’s Doncaster card from a time viewpoint.  

    It’s actually quite a good example as there was no rain during the meeting and six of the seven races were over the straight course and we’ll concentrate on those.

    In race order, the winners, their ages, weights carried, official ratings and times relative to standard were as follows:-

    1.50 (5f) Intellibet One 2 7-13 (no OR) 6.49 secs slow<br>2.20 (8f) Ellen Mooney 3 8-6 (OR 78) 10.59 secs slow<br>3.35 (8f) Zucchero 6 8-13 (OR 91) 8.94 secs slow<br>4.10 (6f) Falcon Hill 3 8-6 (OR 108) 5.26 secs slow<br>4.40 (6f) Lady of Gdansk 3 8-9 (OR 55) 7.24 secs slow<br>5.15 (7f) Lingo 3 7-11 (OR 60) 8.02 secs slow

    Do the calculations set out previously (i.e. with no allowance for the going) and the basic timeratings are IO 21, EM 14, Z 20, FH 47, LoG 23, L 18.

    These are obviously far too slow and we know the ground was soft anyway and so some adjustment by way of a going allowance is clearly needed.

    Assume that the horses were slowed down by the conditions to the tune of 0.70 seconds per furlong and re-do the calculations.  In other words, treat Intellibet One as having run 2.99 seconds slower than standard (i.e. 6.49 minus [5 x 0.70] seconds), Ellen Mooney as having run 4.99 seconds slower (i.e. 10.59 minus [8 x 0.70] seconds) etc.  The adjusted timeratings are IO 80, EM 72, Z 78, FH 106, LoG 81 and L 76.  Is 0.70 the correct going allowance?

    My view is that it isn’t and that the correct going allowance is a bit higher than that.  Again, I’m a bit short of time and I’ll have to go into my reasons a bit later on in the week.  I’ll also explain the use that can be made of the ratings once they’re finalised.

    #99170
    paul101
    Member
    • Total Posts 27

    Hi Guskennedy, thanks for all the input, its really interesting, how I would love to be able to calculate time ratings myself, my hobby is reading form, I can spend 20hours or more a week just looking at form,time and pace, and digging deep, allthough I only show a small profit on my punting, I enjoy the time I put in  reading, but Iam afraid to make my own time figures would be way way above me, I will be  looking forward to your next post.

    cheers Paul

    #99171
    thebairn
    Member
    • Total Posts 72

    Gus / Paul

    Very interesting thread – I’ve just checked my going allowance for Doncaster on Saturday, and it came out at 0.69 seconds/furlong.

    #99172
    buggy
    Member
    • Total Posts 2

    A good bit of reading GUS wanted to do my own speed ratings for a long time and will now have a go<br>hope everybody is fine been away for a while good to be back

    #99173
    guskennedy
    Member
    • Total Posts 759

    What you end up with is a sort of grid which looks like this:-

    <br>                                     0.60 0.65 0.70 0.75 0.80 0.85

    Intellibet One (no OR)    72    76     80    84    88    92<br>Ellen Mooney (OR 78)     64    68     72    76    80    84<br>Zucchero (OR 91)           70    74     78    82    86    90<br>Falcon Hill (OR 108)        98   102   106  110  114  118<br>Lady of Gdansk (OR 55) 73    77     81    85    89    93<br>Lingo (OR 60)                 68    72     76    80    84    88

    <br>So, if the going allowance is 0.60 seconds per furlong it would produce the timeratings in the first column.  If it was 0.65 it would give those in the second column etc.

    Bear in mind that there’s a difference of 14lbs between the two scales (see earlier on this thread) so if, for example, the Lincoln was truly-run Zucchero could be expected to record a timerating of about 105.  The going allowance would need to be about 1.04 seconds per furlong to produce that timerating for Zucchero.  The problem is that that sort of going allowance would produce timeratings for Falcon Hill of about 133 – champion sprinter class – and for Lingo of about 103 and that’s not feasible.  It would mean that Lingo had recorded a timerating 29lbs in excess of his handicap mark.  Lingo is very probably improving but if you accept as I do that at seven furlongs there are two pounds to a length this would mean that all horses that finished within fourteen and a half lengths of Lingo had shown improved form relative to their handicap marks.  Exactly half the field finished within that distance of Lingo – eleven out of twenty two.  It’s too many.  Conclusion: the Lincoln wasn’t truly-run nor was Ellen Mooney’s race.

    Lingo’s race might have been truly-run and so might Lady of Gdansk’s.  The latter was a 33/1 shot and was probably showing improved form.  Lingo won by two lengths with the third horse a further three and a half lengths away in a big field, distances which often indicate a truly-run race.

    It’s impossible to seize on any one race or any one horse through which to "timerate" this Doncaster card.  The "best fit" I can come up with is a going allowance of 0.75 seconds per furlong.  This allows for both Lingo and Northern Nymph (the first two in the last) having put up performances on time in advance of their handicap marks and it credits Lady of Gdansk with 16lbs improvement on her handicap mark.  The timeratings for these two races could in theory have been a little higher but that would have resulted in a corresponding increase in the timeratings of the juveniles in the first race which was, after all, an early season maiden auction event in which many horses finished quite close up.

    The problem is that the handicap marks of Lady of Gdansk, Lingo and Northern Nymph are likely to shoot up when the new marks are published tomorrow.  I’ll report back then.

    #99174
    guskennedy
    Member
    • Total Posts 759

    God knows what happened to the grid.  It looked all right when I posted it.

    I must give up drinking…

    #99175
    buggy
    Member
    • Total Posts 2

    Gus then how do you work out the rating for the second / third horses etc

    #99176
    guskennedy
    Member
    • Total Posts 759

    buggy…in answer to your question, it’s simply a case of adjusting the rating according to how far the other horses were beaten and what weight they carried relative to the winner.  I allow 3lbs per length at sprint distances, 2lbs between 7f and 10f, 1.5 lbs from 11f-14f and 1lb a length 15f and above.  In Lingo’s race, for example, Northern Nymph carried 19lbs more than the winner and was beaten two lengths which equals 4lbs at 7f.  Eagles High carried 10lbs more than Lingo and was beaten a total of five and a half lengths (equals 11lbs).  Lingo’s timerating was 80, so NN’s was 95 (80+19-4) and EH’s 79 (80+10-11).  You have to remember to allow for weight-for-age where the horses are different ages.

    The new handicap ratings were published today and Lingo goes up to 69, Northern Nymph to 83 and Lady of Gdansk to 70.  From a pure time perspective, Lady of Gdansk is the most interesting as this new rating means that she will be attractively weighted in future handicaps and, as she comes from an unfashionable yard, she is likely to be a decent price. She would also be of interest in any 0-70 rated stakes as she would receive the fillies’ allowance.

    #99177
    guskennedy
    Member
    • Total Posts 759

    In case anybody is interested, the fastest timerating of the Flat season so far is the 122 recorded by Kyllachy at Newbury on 19th April.  I see he’s entered for the Palace House at Newmarket on Saturday.

    The official handicapper has underestimated the six-furlong maiden for three-year-olds run at the Craven Meeting.  The speedratings for the first four were Indian Country 108, Greenslades 103, Green Line 100 and Faiza 92.  Indian Country has been given an official rating of 83.  He’s entered in a 7f handicap at HQ on Friday although I’m not sure he’ll be suited by that trip.  The second and third haven’t been given a handicap mark yet but Faiza has been left on 69 and she’ll be interesting in ratings-related races as well as handicaps.

    The handicapper often gets this race wrong and in the past couple of years San Salvador and Polar Kingdom have got into handicaps on winning marks after running in this race.

    #99178
    guskennedy
    Member
    • Total Posts 759

    Blowing your own trumpet isn’t the done thing but nobody else is going to do it so here goes…

    Of the five horses mentioned above, four have now run.  Green Line won a maiden at odds-on last Friday.  Kyllachy was backed from 4/1 and won the Palace House and today Faiza won a Brighton handicap at 5/1 (from 7/1), having been beaten in a similar race at Bath last week.  She did so from a bad draw and will not go up too much in the handicap.  Indian Country is the only non-winner of the four.  He was very heavily-backed in a Haydock handicap twelve days ago but flopped.  The ground was soft and it had been soft also when he was beaten first time out at odds-on.  By contrast, his Newmarket win was on fast ground and I think that’s the key to this horse.

    #99179
    robgomm
    Member
    • Total Posts 224

    <br>Well done Gus!! You should be TRF’s speed ratings expert, they are obviously working well…i hope you’ve made some money this season :) Rob.

Viewing 13 posts - 16 through 28 (of 28 total)
  • You must be logged in to reply to this topic.