- This topic has 47 replies, 10 voices, and was last updated 17 years ago by
.
-
Posts
-
or official handicap ratings who outperforms who, my moneys on the bhb anyone got any stats on this
Class Tells
Below are a few stats from my evaluations, July 1 to Nov 4, 2006. Flat only.
They show the ranking of the winners ability rating.top 20 14%
2nd 14 10%
3rd 21 14%
4th 14 10%
5th 15 10%
6th 10 7%
7th 8 5%
8th 8 5%
9th 11 8%
10th 3 2%
11th 7 5%
12th 5 3%
13th 2 1%
14th 1 1%
15th 2 1%
16th 3 2%
17th 1 1%
18th 0 0%
19th 1 1%
races 146 100%So the top rated on ability won 14% of races, second top rated 10% etc.
These races include 2-y-o and 3-y-o races, handicaps and stakes. I will reproduce the above for 4-y-o and upwards races only, as this will give a better comparison. I did not keep a record of O.R’s in these races but it is possible to add them, although this could take some time.
Now! Tell me how do you compare O.R’s to ability ratings in a handicap.looking at these figures there may not be a lot between ability ratings and top rated bhb even in handicaps i would have said bhb ratings would outperform ability ratings to compare just ignore weight ,top ability V top weight in handicaps who,s best ,looking at the % of ability ratings how close does it mimic top weights top 5 good percentages then tail off
Class Tells
Below are a few stats from my evaluations, July 1 to Nov 4, 2006. Flat only.
They show the ranking of the winners ability rating.top 20 14%
2nd 14 10%
3rd 21 14%
4th 14 10%
5th 15 10%
6th 10 7%
7th 8 5%
8th 8 5%
9th 11 8%
10th 3 2%
11th 7 5%
12th 5 3%
13th 2 1%
14th 1 1%
15th 2 1%
16th 3 2%
17th 1 1%
18th 0 0%
19th 1 1%
races 146 100%So the top rated on ability won 14% of races, second top rated 10% etc.
These races include 2-y-o and 3-y-o races, handicaps and stakes. I will reproduce the above for 4-y-o and upwards races only, as this will give a better comparison. I did not keep a record of O.R’s in these races but it is possible to add them, although this could take some time.
Now! Tell me how do you compare O.R’s to ability ratings in a handicap.[/quoteGARSTONF
THAT IS EXCELLENTClass Tells
Below are a few stats from my evaluations, July 1 to Nov 4, 2006. Flat only.
They show the ranking of the winners ability rating.top 20 14%
2nd 14 10%
3rd 21 14%
4th 14 10%
5th 15 10%
6th 10 7%
7th 8 5%
8th 8 5%
9th 11 8%
10th 3 2%
11th 7 5%
12th 5 3%
13th 2 1%
14th 1 1%
15th 2 1%
16th 3 2%
17th 1 1%
18th 0 0%
19th 1 1%
races 146 100%So the top rated on ability won 14% of races, second top rated 10% etc.
These races include 2-y-o and 3-y-o races, handicaps and stakes. I will reproduce the above for 4-y-o and upwards races only, as this will give a better comparison. I did not keep a record of O.R’s in these races but it is possible to add them, although this could take some time.
Now! Tell me how do you compare O.R’s to ability ratings in a handicap.[/quoteGARSTONF
THAT IS EXCELLENTGarstonf on your figures for the races you checked would i be right in saying top 5 ability won 58% of the races
Garstonf,
Whilst your research is admirable surely any statistical survey like that needs to be related to number of runners.
100% winners from the top 5 ability rating/ BHB rating is meaningless if only taken from 5 runner fields
A better comparison (IMO) for such statistics is % winners in top quartile or top 20% of field etc.
An interesting figure, Garstonf.
Just 2.6% better than random. I would think that quite a jolt to those who think weight plays little part in the equation with class.
Crock
Forgive me, but I can’t follow your logic. If I understand Garston’s posts correctly, 27.6% of the winners in the races he has surveyed were won by horses in the top quartile of the AR ranking. I’m not clear how that bears on weight, one way or the other.
Interesting and certainly needs attention drawing to it.
The sample is way too small to form any solid conclusions, though, but it’s a start in the right direction.
GL
Crock
Forgive me, but I can’t follow your logic. If I understand Garston’s posts correctly, 27.6% of the winners in the races he has surveyed were won by horses in the top quartile of the AR ranking. I’m not clear how that bears on weight, one way or the other.
Hensman,
My logic is that a random selection method, let’s say splitting them in alphabetical order would expect to find 25% winners in any given quartile.
I would expect a ‘merit’ system, by which I’d include a class rating in that, to outperform random by more than 2.6%.
My conclusion therefore is that either the ability rating is little better than random or the effect of weight penalties related to that merit is negating that merit.
My guess would be that if you performed the same exercise on non handicap racing you’d find a significantly higher percentage than 27.6% in the top quartile of ability ratings. Here there are only limited weight penalties to negate any class advantage.
I’d further suggest that any ‘weight based’ rating system say Timeform or Postmark would find more than 27.6% winners in the top quartile.still say bhb handicap ratings will outperform vdw ability ratings but cant prove it
Crock,
A couple of questions if I may? First are you saying the ability rating doesn’t work, or isn’t up to the job because it doesn’t take weight into account? Also looking at Garstonf’s figures he appears to be using the bare rating. As you know when looking at the examples the selection is often well down the ability rankings, so how do G’s figures prove anything?
As you also know I don’t use/look at weight at any time, however I would be more than happy to put my ratings against any of the weight based rating you mention. Unfortunately I can’t give you stats as I inadvertently included a formula into my analysis sheets (also the sheets I used as a record of past races) so the figures have automatically updated, at times making a nonsense of past races.
Be Lucky
Class Tells
After taking out all races that contained 2 and 3-y-o’s there were only 18 left. So instead, it is as before but with all 2-y-o races removed and also any race where the winner did not have an official rating such as Hurricane Run in the King George and Queen Elizabeth, who was French trained.
As stated before, the stats are from my evaluations, July 1 to Nov 4, 2006. Flat only.
They show the ranking of the winners’ official rating and ability rating for comparison.______O.R.________Ability
Top____31___23%___18___13%
2nd____16___12%___16___12%
3rd____16___12%___20___15%
4th____17___13%___12____9%
5th____15___11%___17___13%
6th_____8____6%____6____4%
7th_____6____4%___11____8%
8th_____6____4%____7____5%
9th_____3____2%____5____4%
10th____2____1%____9____7%
11th____4____3%____5____4%
12th____0____0%____3____2%
13th____2____1%____1____1%
14th____3____2%____0____0%
15th____2____1%____1____1%
16th____1____1%____3____2%
17th____1____1%____0____0%
18th____0____0%____0____0%
19th____0____0%____0____0%
25th____1____1%____0____0%
Races_134_100%__134__100%From the 31 winners who were top on official rating, 19 were from stakes races and 12 from handicaps. 4 of these handicap winners were 3-y-o against older horses, so were not top weight due to weight-for-age allowance.
Now for my quartiles. The number of runners in each field was divided by 4 so that the ranking would fit into a specific box. So with an 8 runner field if the winner was top or second on ability it would be in the first or top quartile, if it was fifth or sixth it would be in the third quartile, seventh or eighth in the bottom quartile. For 5 or 6 runners, only the top rated would fit into the top quartile and only the bottom rated would fit into the bottom. I hope all that is clear.
Quartile O.R.________Ability
Top_____50___37%___37___28%
2nd_____34___25%___31___23%
3rd_____33___25%___44___33%
4th_____17___13%___22___16%
Races__134__100%__134__100%As you can see a different picture has now emerged with regards to ability rating. Almost a third of the winners had an ability rating that was ranked in the bottom third quarter in the field.
My thanks to Crock for mentioning quartiles, as the last thing I would have wanted was to give statistics that were misleading.Grand Lodge
Just as a matter of interest. What size sample would you say is sufficient for an acceptable survey?
Crock
I have read your post and split the quartiles.
Stakes races only
Quartile O.R.________Ability
Top____27___45%___20___33%
2nd____17___28%___13___22%
3rd____12___20%___17___28%
4th_____4____7%___10___17%
Races__60__100%___60__100%Handicaps
Quartile O.R.________Ability
Top____23___31%___17___23%
2nd____17___23%___18___24%
3rd_____21___28%___27___36%
4th_____13___18%___12___16%
Races__74__100%___74__100%
- You must be logged in to reply to this topic.