- This topic has 235 replies, 18 voices, and was last updated 17 years, 11 months ago by
.
- Posts
going to have a go at the scatter gun approach the opposite tomorrow but in the long run its not the way to go
Dave,
The RP stuff is both well known and used by a lot of people, but it’s hard to get away from the logic of it.
The only fault I can find is that horses moving up in class can get ‘run off their feet’ in better races. This doesn’t happen all the time, but we have to be aware of it.
Finishing positions have some merit provided there is sufficient data. Even the best horses can run a few bad races or not be ridden out when all chance of winning has gone. The ratings approach ignores these bad races, concentrating on a horse’s best performances.
I don’t make any specific allowance for class when compiling ratings at present, but it’s always worth considering in the future.
Formath,
Thanks for keeping this interesting topic going. I thought your recent losing run might scupper you, but you are still afloat and sailing on.
i said i,d post up my scatter gun selections today god i wish i had not started picked out the races and stuck with them 5 seemed to be selected easy enough gungadu,labelthou,nozic,turn on the style,the outlier the next 3 were a bit harder to dig out afsoun,preacher boy,marcel, the 2.20 at ascot what a bad grade 1 race,and went for monets garden and the 4.55 at hay madison du berlais? how many winners out of ten will this method give me i say 6 what do you say formath what a great day in front of the tv today
Class Tells,<br>This is the advice I believe VDW would have given you today, quoted from a sports forum letter May 24, 1979:
"If no choice emerges from methodical study, one of two things should become plain, the competition is too keen to justify a bet or the winner must be looked for among the outsiders. May I suggest the first alternative is the one to adopt (temperament)."
1.45 Ascot<br>Gungadu – (4/6 won) up 2lbs, won 1st of 5 +25 lengths, led, blundered, clear, easily. Plenty in hand and clearly the one to beat. Odds-on no bet<br>Jaunty Times (16/1) – same weight, won 1st of 4 +shd, slighly outpaced, rallied, ran on. On a hat-trick but tougher here.<br>Over The Creek (7/2) – up 2lbs, 2nd of 18 -2.5 lengths (penultimate race), chased, stayed on, snatched 2nd. Could be a threat at the weights.
3.45 Haydock<br>The Sawyer – (11/4) up 7lbs, won 1st of 11 +7 lengths, not fuent last, challenged, led, stayed on well, drew clear. No easy task at the weights. Stake 1 point lost, due 2, bank 175.32
Chief Dan George (9/1 won) – up 2lbs,won 1st of 7 +hd, keen, led, all out. Improving and can’t be ruled out.<br>Joacci (6/4) – down 4lbs, won 1st of 13 +6 lengths, made virtually all, went clear, easily. Should have no trouble holding his own.<br>Hills Of Arran (15/2) – down 1lb, 5th of 11 -11.5 lengths, stayed on , hit last, never got competetive. has yet to get head in front over jumps?<br>
(Edited by Formath at 4:27 pm on Feb. 17, 2007)
Dave Jay,
I agree with your opinion that VDW probably bet to coupled odds justifying his claim for 80-90% winning bets in a season.
Have Fun,
With regard to your query re VDW advising course class to be considered at some time in his letters. I can’t recall that and I have no way of checking as I have given all my Moss publication books away and all my copies of the SCHB. The only thing that I have found is a dog-eared photocopy of a VDW letter to sports forum around 1982:
"There is a good maxim in racing which says ‘never back a horse that has not won’.<br>This does not obviate the need to have a proper evaluation of any race and, unlike National Hunt, Flat racing is full of events catering for young two and three-year-old horses.<br>Many three-year-olds have never run as two-year-olds and until it is possible to get a full measure, as is the case with older horses, some means other than that previously shown is required for use in conjunction. I suggest the same rating method is used, but with such youngsters have as a further guide a means based on time."
He then goes on to detail at length a system fo using Split Second ratings to achieve this (if anyone is interested I can add it on).
(Edited by Formath at 3:03 pm on Feb. 17, 2007)
it all went as i thought but on reflection i would not have incuded the outlier did things to quickly this morning,as an after thought when checking my selections against form spot of the mail 9 out of ten were first or secound rated may use this next time 6 out of 10 is not to bad for a first effort
2.30 Fontwell<br>Turko (10/11 won) – up 2lbs, 2nd of 5 -17 lengths, not fluent, effort, blundered, faded. Has the class but going may not suit. Odds-on no bet<br>Garde Champetre (Non-runner) – same weight, 2nd of 5 -6 lengths, chased, effort, kept on same pace. Disappointing and not rock solid.<br>Phar Bleu (8/1 2nd) – same weight, 2nd of 4 -23 lengths, hit 5th, led, headed, no chance with winner. Vulnerable to anything of quality.<br>Just for interest the straight forecast paid £7.99
(Edited by Formath at 4:55 pm on Feb. 18, 2007)
3.25 Market Rasen<br>Ice Melted (7/4 2nd) – up 5lbs, 2nd of 8 -14 lengths, chased winner, no impression, no chance. Progressive over fences, this brings him into contention. Stake 2 points lost, due 5, bank 173.32<br>Character Building (7/4 won) – up 26lbs? 4th of 10 -12 lengths, ridden, weakened, held, hampered last. Improver over fences and could go well.<br>Birdwatch (10/1) – down 24 lbs, rallied, every chance, one-paced. Potentially well treated.
4.10 Market Rasen<br>Loulou Nivernais (11/4 won) – down 3lbs, won 1st of 9 +nk, headway, challenged, stayed on to lead. On a hat-trick but facing better quality opposition. <br>Stake 3 points won = +8.25, due 0, bank 181.57<br>Spring Lover (9/2) – same weight, 3rd of 13 -3.25 lengths, ridden, every chance, no extra. Going may not suit and others may be too good.<br>Nonantais (6/1)- same weight, won 1st of 8 +.5 of a length, chased, led, headed, led, ridden, rallied. Likely to make presence felt.<br>It’s Bertie (9/4) – down 5lbs, 4th of 8 -13.5 lengths, ridden, one-paced. Of interest.
(Edited by Formath at 4:57 pm on Feb. 19, 2007)
Struggling to find anything worth assessing on the cards today and these two races are just to keep the pot boiling as we are at the start of a staking progression.<br>3.10 Taunton<br>Brave Villa (15/8 won) – down 2lbs, 4th of 12 -15.5 lengths, headed, stayed challenging, weakened, blundered. More suitable distance today. Stake 1 point won = +1.87, due 0, bank 183.44<br>Kayceecee (9/2 3rd) – down 2lbs, 2nd of 5 -4 lengths, held up, ent 2nd, found little. Needs to find more.
3.30 Southwell<br>Beldon Hill (7/4 lost) – up 15lbs? won 1st of 8 +5 lengths, chased, led, hung, driven clear. C&D hard to beat.<br>Stake 1 pt lost, due 2, bank 182.44<br>Jazrawy (3/1 won) – up 1lb, won 1st of 10 +shd, quickened, ridden, kept on gamely. Tougher but can’t be ruled out
(Edited by Formath at 5:52 pm on Feb. 20, 2007)
…and I thought I was struggling yesterday<br>4.00 Ludlow<br>Ouninpohja (5/6 2nd) – up 8lbs, 3rd of 14 -11 lengths, effort, not quicken. Suited by the going.<br>Odds-on no bet<br>
(Edited by Formath at 5:11 pm on Feb. 21, 2007)
Hi Formath .. my own bet in this race is Iffy. I use the VDW method after a sort but factor in actual prices .. although Ouninpohja is the most likely winner, I think mine is a better bet (Win and Place) .. any thoughts?<br>:cool:
Hi Dave,<br>I did consider Iffy in my VDW platform and it was an each-way chance along with Kickahead and Nikola:
Iffy – up 16lbs, 3rd of 9 -10.5 lengths, chased, no impression, weakened. Some chance here.
Nowdays I have price as the most critical factor taking RPR as my source of expert opinon. I support runners at RPR odds to return 100 where shorter than the market – meaning I make a killing if the market odds are greater than the RPR opinion. <br>I only consider non-hcaps with smallish fields owing to the fact that ratings tend to be tight in hcaps. However, the 4 o’clock race today would have been great:
Headings – Cloth No, RPR, RPR odds, stake, SP.
1 136 5/1 16pts 5/6 no bet<br>2 132 15/2 11pts 33/1 lost Lord Baskerville<br>3 132 15/2 11pts 15/2 lost Iffy<br>4 133 7/1  12pts 33/1 lost Nikola<br>5 136 5/1  16pts 7/2 no bet<br>6 136 5/1  16pts 11/1 WON +176pts Kickahead<br>7 134 6/1  14pts 14/1 lost Rustler<br>8 122 100/1 (nothing over RPR 10/1 considered)
Hindsight is a wonderful thing and I could have made 128pts. On the downside if Ouninpohja had gone in I would have been down 64pts.
(Edited by Formath at 5:57 pm on Feb. 21, 2007)<br>
(Edited by Formath at 1:56 pm on Feb. 22, 2007)
Hindsight is a wonderful thing, similar to time travel .. :biggrin:
I like what you have done there with the RPR .. very clever. I really didnt like Ouninpohja, because short priced favs in races like this are just cronic bets, or so I have found .. roll on tomorrow.
Formath,
I’ve tried something similar, converting RPR ratings(with some added points for other factors) into odds. Do you use a base figure such as Top-rated minus X as a starting point for your calculations?
Artemis, deduct the lowest rating +1 from the highest rating, then divide each individual score by the sum of the total scores.
Bet value = percentage converted to points.
If I may be so bold .. :)
3.00 Huntingdon<br>Chief Yeoman (6/5 won) – up 3lbs, won 1st of 4 +15 lengths, chased, led, clear, canter. Promising and a bold bid likely.<br>Stake 2 points = +2.4, due 0, bank 184.84<br>Echo Point (15/8 3rd) – down 2lbs, won 1st of 13 +22 lengths (penultimate race), made all, drew clear, easily. Useful hurdler and could need catching.<br>Garde Champetre (5/1 2nd) – same weight, 2nd of 5 -6 lengths, chased, effort, same pace. Hesitant chaser and trip may not suit.<br>Just for interest the straight forecast paid £7.18<br>
(Edited by Formath at 4:30 pm on Feb. 22, 2007)
Dave,
I tried that for a while, but a lot of the prices didn’t look right. We know from years of statistics that the SP odds are fairly representative of a horse’s chance, bearing in mind that these stats are based on averages and that value can be found if you look for it.
However, the RPR(or any other ratings into odds method) seems to extend the prices of favourites, sometimes quite alarmingly so that 1/1 becomes 4/1.<br>Clearly, this model cannot be right or realiable as a predictor, otherwise you could be laying horses at inflated odds with disastrous results.
I’m still working on it. I was allowing about 4 percentage probability points for 1 ratings point, when last I tried it. I didn’t manage a formula – I used iteration or adjusting the odds until they fitted the model and looked about right.
- You must be logged in to reply to this topic.