Timeform ratings and DRF

[beertapper]

i wonder why they decided to replace them with these ones.. will brisnet or someone else get rights for them now?

http://www.drf.com/news/article/85701.html

"Racing Post Ratings - the single most important handicapping tool provided by the British racing daily the Racing Post - replaces Timeform Ratings and will appear in the foreign past performance lines of horses who have raced in Britain, Ireland, France, Germany, Italy, Dubai, Hong Kong, Japan, South Africa, and Australia. The ratings will also be included in the foreign results section in the DRF Simulcast Weekly."

[Pgh. Gere]

On another thread, someone had stated that the contract was up. Timeform, now owned by betfair, was asking for alot more money to supply their ratings. I've used the RP Ratings before and have found them to be very good.

[cj]

I personally think the ratings are better. You can also get them for free at the Racing Post site for most horses.

Here is Curlin:

http://www.pacefigures.com/images/curlinRP.jpg

[robert99]

It is a strange business model when two companies providing the same product, lose business and international kudos by putting the price up against a free service competitor.

If you believe any horse performance can be rated by a single number, then Timeform were slightly better as they take into account pace and track variant, also their raters watch and retime each video recording. Racing Posts are very much computer generated to a simple formula, so Curlin's 8 length loss to Street Sense is a 111 - the literal, arithmetic performance, rather than what the race quality analysis might have told you.

[Greyfox]

They may well be a good "handicapping" figure, if every horse in the field has one tagged to it. But...
At least with Time Form Ratings you could get a rough idea as to how they compare to Beyers. You won't be able to do that with these numbers.
As I bet North American races only, this change away from Time Form is not welcomed by me.

[cj]

They may well be a good "handicapping" figure, if every horse in the field has one tagged to it. But...
At least with Time Form Ratings you could get a rough idea as to how they compare to Beyers. You won't be able to do that with these numbers.
As I bet North American races only, this change away from Time Form is not welcomed by me.

They are very similar. If you can convert Timeforms to Beyers scale, you can do the same for RPR ratings. I've already done it.

[cj]

TF Beyer RPR
80 75 90
82 77 91
84 79 92
85 79 93
87 81 94
89 83 95
91 85 96
92 85 97
94 87 98
96 89 99
98 90 100
99 91 101
101 93 102
103 94 103
104 95 104
106 97 105
107 98 106
109 99 107
110 100 108
112 102 109
113 102 110
115 104 111
116 105 112
118 106 113
119 107 114
121 109 115
122 110 116
124 112 117
125 113 118
127 114 119
128 115 120
130 116 121
131 117 122
132 118 123
134 119 124
135 120 125
137 121 126
137 122 127
137 121 128
138 122 129
139 123 130
141 125 131
142 126 132
144 127 133

[robert99]

There is no reliable conversion between a time rating, Beyers, and the "class" ratings of Timeform (which does not directly consider time, despite the name). You can readily see this when you compare Timeform's "class" rating figures with their own but separate time rating figures. However, if you look at the Racing Post example above there will also be a TS (Topspeed) figure supplied for UK, France and Ireland races, which is time based, and will be better collated with Beyer time figures when considering imports. So for USA use the new source is far better.

[cj]

There is no reliable conversion between a time rating, Beyers, and the "class" ratings of Timeform (which does not directly consider time, despite the name).

That is what everyone keeps saying. I'll just keep cashing on those foreign imports with the highest converted rating, or the ones with 1st Lasix that are close.

[cj]

Anyone using Formulator will still see:

Timeform Rating:

They are, however, the RPR ratings from Racing Post. I checked a few horses myself.

[Tom]

Let me see now, there was this one horse who loked really good on that conversion from TF to Beyer. Hmmmm, could it have been........

....INVASOR

Those conversions, they are thee best! :lol:

[Zman179]

Personally, I do not like Racing Post Ratings because the actual figure will change depending on the horse's conditions for today's race (in the Racing Post.) Plus, weight will adversely affect the figure, and for someone like myself who doesn't like to take small weight shifts into account that is not a good thing.

For example, horse A ran a race on the all-weather (dirt) track at Southwell on June 1. The horse's rating will look something like this:

In the race chart for June 1: RPR of 84
In the HTML form for today's race on June 12, the RPR for Horse A's 6/1 race is now: 88
In the PDF form for today's race on June 12, the RPR for Horse A's 6/1 race is now: 94

And for his following race, the rating will be adjusted even further. As you can probably tell, when you compare a horse's HTML rating on the RP website with the PDF file that the RP offers, the ratings will be totally different...for the same horse on the same day! There's simply too many discrepancies in the numbers for me to take them as seriously as Timeform's numbers.

[robert99]

Let me see now, there was this one horse who loked really good on that conversion from TF to Beyer. Hmmmm, could it have been........

....INVASOR

Those conversions, they are thee best! :lol:



Tom,

Invasor won 3 group ones prior to BC, class AND in very FAST TIMES.
His RPR/TF ratings were from 83 - 123. It was the top class form and consistent FAST TIMES that won him the BC, not a conversion figure from somewhere betwen 83 and 123. Even then some chimps still did not back Invasor to favourite.

Reliable means being consistent over a series of horses, not just one, so who backed George Washington also at CD with an even better RPR/Timeform rating of 131 from its previous Ascot race - it might still be running?

[cj]

Reliable means being consistent over a series of horses, not just one, so who backed George Washington also at CD with an even better RPR/Timeform rating of 131 from its previous Ascot race - it might still be running?

GWs numbers were on turf. No sane capper would use them on dirt.

[kenwoodallpromos]

"http://www.chef-de-race.com/pfs/comparative_speed_figs.htm"

[robert99]

GWs numbers were on turf. No sane capper would use them on dirt.

Exactly, no UK, France and Ireland horses rated by Timeform/ RP would have been rated on dirt or even on tracks as tight as US turf tracks but still people believe there is a simple conversion.

[robert99]

Personally, I do not like Racing Post Ratings because the actual figure will change depending on the horse's conditions for today's race (in the Racing Post.) Plus, weight will adversely affect the figure, and for someone like myself who doesn't like to take small weight shifts into account that is not a good thing.

For example, horse A ran a race on the all-weather (dirt) track at Southwell on June 1. The horse's rating will look something like this:

In the race chart for June 1: RPR of 84
In the HTML form for today's race on June 12, the RPR for Horse A's 6/1 race is now: 88
In the PDF form for today's race on June 12, the RPR for Horse A's 6/1 race is now: 94

And for his following race, the rating will be adjusted even further. As you can probably tell, when you compare a horse's HTML rating on the RP website with the PDF file that the RP offers, the ratings will be totally different...for the same horse on the same day! There's simply too many discrepancies in the numbers for me to take them as seriously as Timeform's numbers.

The Timeform number is to a 140 base and that is to 10 stone. So the base rating assumes the horse is to carry 10 stone. There is no difference between RP and TF on that score. Yes the ratings (for both) will be adjusted depending on the actual weight carried in an actual race and they will be revised after every race depending on the result. Beyers are not so weight sensitive but will also change every race, so will Timeforms. What you have observed further makes the case of why it is so difficult to relate correctly either rating to Beyers.

[Greyfox]

Thanks Cj for the table. That's exactly what I need. Well done.

[mhrussell]

...for posting the table of RP ratings vs. Timeform and Beyer. And I agree, the conversion formulas work well enough to cash on some nice prices!

best,

[Tom]

Tom,

Invasor won 3 group ones prior to BC, class AND in very FAST TIMES.
His RPR/TF ratings were from 83 - 123. It was the top class form and consistent FAST TIMES that won him the BC, not a conversion figure from somewhere betwen 83 and 123. Even then some chimps still did not back Invasor to favourite.

Reliable means being consistent over a series of horses, not just one, so who backed George Washington also at CD with an even better RPR/Timeform rating of 131 from its previous Ascot race - it might still be running?

I'm talking about his first US start - at Pimlico. His CJ made from a converted TF made him no surprise in that race. Which is what you want the conversion to do. CJ has incorporated the conversion to make shippers compatible wtih US ratings. :ThmbUp::ThmbUp::ThmbUp:

[Zman179]

The Timeform number is to a 140 base and that is to 10 stone. So the base rating assumes the horse is to carry 10 stone. There is no difference between RP and TF on that score. Yes the ratings (for both) will be adjusted depending on the actual weight carried in an actual race and they will be revised after every race depending on the result. Beyers are not so weight sensitive but will also change every race, so will Timeforms. What you have observed further makes the case of why it is so difficult to relate correctly either rating to Beyers.

I appreciate your explanation. However, in regards to the Beyer numbers, once the number has been made for a race (i.e. a horse wins and receives a 102 for his effort), then that number is set in stone. Regardless of the conditions for his next 1, 5, or 20 races, that 102 for that race will always remain a 102. I just don't believe that figures should be fudged based on today's conditions unless the handicapper wishes to fudge them himself. But I do understand that that's simply the way it is.

[Tom]

Actually, Beyer verifies his numbers and makes adjustments to them as needed. Ususlaly not a lot, but I 've seen 3-4 month periods where all numbers at FL were raised by 2 points when he revised his track-to-track.

I modify ( well, actually, CJ does it for me) numbers by age - alllowing for improvement over time in uyoiunger horses.

[Zman179]

Revisions to make the numbers more reliable are one thing, revisions because the horse is carrying 114 lbs. today is something entirely different.

Besides, I think that Beyer had to revise his Finger Lakes numbers so that he wouldn't have a winner with a -0 figure. :lol: What was it, not too long ago a FL winner had a figure of 2?

[Pgh. Gere]

Zman- Timeforms racecards do the same thing, but what we will see in the DRF is the final rating figure for that race. Both TF & RPR and Beyer will go back and change a number after the fact only if subsequent results show that they need to.

[Tom]

Zman, when Beyers first came out, there were lots of horses with "Beyer Bagels" - 0 speed figs! at the thumb.:eek:

[Cratos]

There is no reliable conversion between a time rating, Beyers, and the "class" ratings of Timeform (which does not directly consider time, despite the name). You can readily see this when you compare Timeform's "class" rating figures with their own but separate time rating figures. However, if you look at the Racing Post example above there will also be a TS (Topspeed) figure supplied for UK, France and Ireland races, which is time based, and will be better collated with Beyer time figures when considering imports. So for USA use the new source is far better.

You are absolutely correct and Phil Bull would roll in his grave if he heard this conversion nonsense.

[cj]

You are absolutely correct and Phil Bull would roll in his grave if he heard this conversion nonsense.

Does this mean I have to give the money back? Here is one of countless examples:

http://www.pacefigures.com/archives/bc03/distaff.html

Check out Buy the Sport at Lingfield, and then in the Gazelle. I bet this horse playing in the War Room here at PA. Her Timeform adjusted rating earned at Lingfield, her only non-turf race, put her on top. The only horse close, favored Lady Tak, was very questionable at the distance. The horse won and paid $98.

Others from Breeder's Cup races in the last few years that have won with high rated conversions that don't work:

Six Perfections
Domedriver
High Chapparel
Ouija Board
Shirocco
Red Rocks

[robert99]

I'm talking about his first US start - at Pimlico. His CJ made from a converted TF made him no surprise in that race. Which is what you want the conversion to do. CJ has incorporated the conversion to make shippers compatible wtih US ratings. :ThmbUp::ThmbUp::ThmbUp:

Tom,

His Timeform rating, before shipping, at Nad Al Sheba was only 106. This on CJs chart is a Beyer of only 97. Racing Post only rated Invasor at 83 which is below a 75 Beyer. CJ's chart says a Timeform 106 is a RP 105 - but it was not in this case, being only 83. You must have second sight to use a conversion that gave Invasor a good chance of a USA GR1 on first start.
If you talked about his Maronas performances that would be more convincing.

[cj]

Invasor's adjusted rating was a 96. It is in my archives section for last year's BC.

[Cratos]

Does this mean I have to give the money back? Here is one of countless examples:

http://www.pacefigures.com/archives/bc03/distaff.html

Check out Buy the Sport at Lingfield, and then in the Gazelle. I bet this horse playing in the War Room here at PA. Her Timeform adjusted rating earned at Lingfield, her only non-turf race, put her on top. The only horse close, favored Lady Tak, was very questionable at the distance. The horse won and paid $98.

Others from Breeder's Cup races in the last few years that have won with high rated conversions that don't work:

Six Perfections
Domedriver
High Chapparel
Ouija Board
Shirocco
Red Rocks
No you do not have too and I am not disputing your winning, but what I am saying is that you are using “conversion” for “comparison.” Non-parametric statistical theory talks very well to this issue.

[Robert Fischer]

I think Curlin recieved a 5 pt higher score than RTR for the Belmont because of the weight.

They are interesting to see a ballpark idea of how a horse was running.
My favorite use of that site is for the pedigree notes at the top, and also comments in running where they provide an alternative call of the race to the charts.

[mudnturf]

I always felt that the Timeform product was much more than just ratings.
I found their description of each horse extremely helpful by indicating its preference for such things as turf condition and track layout.... some horses preferred "galloping" tracks, while others were more useful on tracks with tighter turns.
And while the Ratings are useful, the symbols that often accompany those ratings make them even more useful.
For example, a plus sign means "might be better than we have rated it", while a question mark means "the rating is suspect".