- This topic has 94 replies, 33 voices, and was last updated 15 years, 11 months ago by
.
- Posts
This’ll end up boring the pants off everyone. <!– s:lol: –>
<!– s:lol: –>There is already a thread 4 pages long on this subject on the betfair forum, but I didn’t want to contribute to something that disappears once it reaches the bottom of the page.
Greg Wood has reported that Phil Smith is going to assess the Timeform ratings given to Arkle, Flyingbolt and Mill House, and will give his verdict at the same time as the Anglo-Irish ratings are published at the end of the season.
<!– m –>http://www.guardian.co.uk/sport/blog/20 … phil-smith<!– m –>
There seems to be a little bit of knocking of the ratings achieved in h’caps, which I find a bit weird, because surely these are more solid than the ones achieved in conditions races. Also, surely one of the main reasons for producing ratings is to assess how much weight should be allocated to horses in h’caps.
I think there is an inextricable problem in producing ratings for staying h’cap chases in Soft ground, because once the elastic has broke, and the other horses have reached the end of their tether, the gaps between the runners become huge. How is one to decide how many lbs there should be per length in those circumstances.
It seems to me that what is required is what speed handicappers do; par times need to be established for races over several years, and then these par times are used to calculate the going variant for the meeting. You then use the lbs per second at that distance for the time difference from the par, and then take the weight carried into consideration as well (which quite a lot of speed h’cappers don’t).
It seems to me that what is required is what speed handicappers do; par times need to be established for races over several years, and then these par times are used to calculate the going variant for the meeting. You then use the lbs per second at that distance for the time difference from the par, and then take the weight carried into consideration as well (which quite a lot of speed h’cappers don’t).
All sounds good in theory Gerald but the reason people don’t take that approach is that the margins for error are significant. Over ten years, say, you could get a wide variety of ground conditions, making it very difficult to develop ‘par’ times. Fences are sometimes missed, race distances (placing of running rails, etc) vary and, possibly most significantly, the pace races are run at can vary considerably, all of which makes makes assessments of ability based on ‘time’ studies fraught with danger.
I did hear on BBC yesterday that Phil Smith was planning to ‘re-evaluate’ Arkle’s rating and it sounds an interesting idea.
I’ll have a go during January at working out the ratings for the Cheltenham Festival for the past 20 years, and see what happens.
All 4 of Kauto’s winning times were faster than Arkles in 1965, even though 3 of the 4 renewals were run on officially slower ground than 1965. On comparative official going descriptions Kauto’s winning time in 2008 is over 12 seconds faster.
It is not just very unlikely that chasing talent could be distributed in such a way across a half-century of horseflesh. It is many millions-to-one against – so implausible in statistical terms, in fact, that it is effectively impossible.
It’s only outlandish odds if you assume statistical independence, which would be a bit silly. Even then how can it be millions to one against when far less than a million horses have run under rules in that time?
The case against is crushingly simple. The next horse in Timeform’s all-time list is Flyingbolt, Arkle’s stablemate, on 210, followed by a 19lb gap to Mill House and – as of Saturday – Kauto Star on 191. So we are being asked to believe that of the hundreds of thousands of steeplechasers to have raced since the mid-1960s, two were nearly a stone and a half better than all the others, and they just happened to occupy adjacent boxes.
I don’t see this as being inconcievable. Human height is often held up as the classic example of a well-behaved natural distribution. The tallest people table looks very similar:
http://en.wikipedia.org/wiki/List_of_tallest_people
I’m not saying Sleepy Hollow haven’t dropped a clanger here but doubt I’d put much faith in Phil Smith and his methods doing a better job. He can’t even callibrate modern horses shipping in from abroad with any degree of accuracy, so how’s he going to come up with accurate marks for horses racing four decades apart?
- Total Posts 17716
I’ll have a go during January at working out the ratings for the Cheltenham Festival for the past 20 years, and see what happens.
The fences have moved too many times to constructed a solid analysis imo, thats for sectionals anyway.
- Total Posts 17716
Might sound a daft question, but did Cheltenham actually have a ‘New’ course in 1965? If not, time comparisons are meaningless.
I just want to do the past 20 years so that I can get the methodology sorted out.
I’m aware of many of the difficulties. For example, nowadays they cut the corner when they go pass the stands.
I just want to develop an accurate speed handicapping method, based in lbs, for use generally, and I’m just using the Cheltenham Festival as a starting point.
The way I see it, form handicapping is fairly accurate, but it is subjective. I am basically a scientist. Speed handicapping is essentially objective, but it is inaccurate in the sense it isn’t too close to the truth at times. My main problem will be after I’ve worked out the going variants etc and then want to convert things to lbs. Is there a scientific or convincing rationale anywhere for how many lbs per length or per second there should be at varying distances run???
Last time at lwast two of Kautos King Georges were run on a track that was partly an alweather surface, this trying too compare him with Arkles performance is nonsense as they didnt run over the same course and distance.
Just a few observations.
Ground conditions can differ within 24 hours so I would have thought trying to compare over 40 years would be pointless.
Why has there been no reassessment before? I suppose at least it gives Phil Smith another chance to get his name in the papers and make him feel important for a few days.
Arkle was before my time but I am aware of his record. After an initial defeat my Mill House he was only ever beaten by weight or injury. He raced in competitive handicaps because there were few ‘easy’ condition races to target. In those condition races such was his air of invincibility very few horses took him on. Although Kauto Star has been tremendously impressive at Kempton he doesn’t have a record of conceding lumps of weight in handicaps or frightening away the opposition.
My own starting point would be to check Arkle’s rating against his contemporaries. Was he actually 21lbs superior to Mill House? If so, and if Mill House’s rating relative to other of their contemporaries was correct then perhaps the rating can be justified.
In all of this, is it the case that Flyingbolt is the forgotten horse in National Hunt history? Only 2lbs off the best ever and well clear of anything else that’s ever run but unknown outside racing and I’d suspect his career achievements could only be recited by the odd anorak or two.
I can’t claim to recite Flyingbolt’s complete record but he had probably a third of Arkle’s runs at top level. That is probably why he is considered by some to be the ‘forgotten horse’. Add to that there appears to be little or no film of his races.
First season; unbeaten over hurdles, including the Gloucester Hurdle. Following season chasing; again unbeaten including a runaway success in what was to become the Royal and Sun Alliance. As a 6 year old won the Massey Ferguson Gold Cup by 15 lengths carrying 12st 6lbs . Won the 2 mile Champion Chase by 15 lengths and ran in the Champion Hurdle the following day [came 3rd?]. Then won the Irish Grand National carying 12st 7lbs conceding 40lbs to the runner up. Hethen sadly caught a rare virus and was never the same again.
Pat Taaffe rode both and was utterly convinced that had they met, Arkle would always have won at distances 3 miles plus.
Gambling Only Pays When You're Winning
I have only just spotted this on Betfair and now on here and would have posted something similar to Glenn had he not posted already.
The idea that the ratings of horses in neighbouring boxes should be regarded as independent is ridiculous, and it is disappointing to see the usually first-rate Greg Wood implying as much.
I think the BHA handicappers do a good job on the whole, but that is hardly a surprise given how many of them there are. If you judged their handicapping methodology on what comes out of Phil Smith’s mouth – a mouth he seems very fond of using – you would not take any of their ratings seriously.
I am actually relatively interested in what Phil Smith comes up with here as, although I know little about Arkle’s form, I have always thought the ratings awarded to him and his contemporaries sounded totally implausible.
It can’t be that difficult (if a little tedious) to put the relevant seasons’ form through the handicapping "mincer" and come up with some more sensible numbers?
I am actually relatively interested in what Phil Smith comes up with here as, although I know little about Arkle’s form, I have always thought the ratings awarded to him and his contemporaries sounded totally implausible.
It can’t be that difficult (if a little tedious) to put the relevant seasons’ form through the handicapping "mincer" and come up with some more sensible numbers?
Your post doesn’t make a great deal of sense as you state that you know little of Arkle’s form and then imply that the rating is incorrect.
<!– s:lol: –>
- You must be logged in to reply to this topic.