• Hi Guest Just in case you were not aware I wanted to highlight that you can now get a free 7 day trial of Horseracebase here.
    We have a lot of members who are existing users of Horseracebase so help is always available if needed, as well as dedicated section of the fourm here.
    Best Wishes
    AR

Hong Kong Speed Figures

Got myself in a mess and took a while to unravel all my files after the failed python project, and then to update from start of season, hopefully it's now back to how it was. Thursday Ratings attached.
 

Attachments

  • HKRATINGS.xlsx
    282.1 KB · Views: 3
have a prototype app for hk racing, only version 2 so must be possible improvements, if anyone has a look any feedback you have let me know.
can search horses past performance history with my ratings and get charts for each rating, choose a date and will take to that result with all the ratings also option to look at all winners on that day.
Tremendous work O Outlander (y) (y)
 
Last edited:
Had a bit of fun making a few changes, the ratings charts now have ALL as well as individual ratings sets, where all the ratings for each sectional etc are plotted I like this think will find it useful myself more than anything else can see where the horse has gone above the required ratings and clicking each rating set on the legend isolates each rating. Pretty pleased with that, amazing what you can get AI to do.
 
Sunday ratings attached
Interactive app updated , seems to take a minute to load and splits AW and TURF races, can get TURF races using left side panel, dropdown>ST.
My home town trying to win the Rugby League for the first time since 1955, town was buzzing with anticipation from before 11am this morning. So off down the pub to watch it.
 

Attachments

  • TODAY.xlsx
    280.5 KB · Views: 7
Nearly forgot to post results and card for tomorrow.

O Outlander great work will look closer tomorrow. The hurst still goes on with the Rugby it seems:(
 

Attachments

  • HK Results25.xlsx
    88.5 KB · Views: 0
  • 261004.xlsx
    1.3 MB · Views: 1
O Outlander The Hare The Hare

I wonder if this parallel speed figure chart could be used for the Hong Kong scene.

Other uses for this chart, if horse A is a proven 1:23.0s (78/79 rating) over 7f, what would be his projected time be over 6f? - Answer - 1:10.6s.

Maybe this chart could be used to apply ratings to sectional times.

Please note when reading chart 110.2 is 1m 10.2s (1:10.2).

View attachment 84429

Mike.


View attachment 84430

Sunday ratings attached
Interactive app updated , seems to take a minute to load and splits AW and TURF races, can get TURF races using left side panel, dropdown>ST.
My home town trying to win the Rugby League for the first time since 1955, town was buzzing with anticipation from before 11am this morning. So off down the pub to watch it.
So I am logged in(just joined a day or two ago)), but it tells me I don't have permission to see.

Oops! We ran into some problems.
You do not have permission to view this page or perform this action.
 
I thought that perhaps I could critique the times.
Maybe something to do with being new member, I think your post was also delayed because it wasn’t there before my last post about the rugby but is there now, maybe just a short period before you can post or download attachments, I don’t know for sure, the link to the app should work for anyone though.
 
I just develop my own speed charts, I'm sure your now, getting the hang of AI, put some csv files together and test , I'm using Linear regression and Ratios and a combination of both which is the best way I'd go, Big drops/increases in distance require multiple regression But for instance just the basic ST1400 = 83 normally translates to HV1200 = 70.839 and Ratio = 70.953 and ST1400 =83 at ST1200 = 70.262 LG and ST1200 = 70.380 Ratio thats my universal figures without doing any To and From and From and To Track and looking at the differences from one Cohort group to another, and a fairly wide days between races at either To and From track, You will immediately see class differences especially doing it in quartiles and ignoring Q4 cohorts
 
Yes think we have failed in the final 4 or 5 times now in Super League, they were never in it yesterday, not a match for Wakefield from start to finish. The score flattered Warrington.
Our GF was last night Roosters and Knights, the play was lopsided, but the scores were very close and a field goal the difference, was very entertaining game
 
I just develop my own speed charts, I'm sure your now, getting the hang of AI, put some csv files together and test , I'm using Linear regression and Ratios and a combination of both which is the best way I'd go, Big drops/increases in distance require multiple regression But for instance just the basic ST1400 = 83 normally translates to HV1200 = 70.839 and Ratio = 70.953 and ST1400 =83 at ST1200 = 70.262 LG and ST1200 = 70.380 Ratio thats my universal figures without doing any To and From and From and To Track and looking at the differences from one Cohort group to another, and a fairly wide days between races at either To and From track, You will immediately see class differences especially doing it in quartiles and ignoring Q4 cohorts
The AI project is on the back burner for me, I wasted a lot of time and effort on that, found a way to get roughly what I wanted using my excel sheets, I tidied them into seasons and that speeded them up with less data, they did a couple of access queries to piece all seasons together, in the end if I give AI full horse history, today’s runners and today’s adjusted ratings it built an interactive dashboard in minutes, I guess the only draw back is AI was promising testing and improvements when it was claiming it would bring all my stuff into pure python, the times and ratios just another way of bringing CDs together same as standard times for each CD. I have what I feel is a proper method for creating standard times for U.K. racings in a python script now, just feed it data and it uses my calculations and methods to produce very accurate times, haven’t adapted for HK yet because so easy to do the same thing manually with just 3 courses. The various past performance ratings i don’t think I can get much better now in my methods, how they are adjusted and interpreted and what ratings are still viable and to what extent older ratings should be decayed, that’s the bit where it’s more art than science.
You can find good reason looking at past ratings why each horse has won, just need to find the reason before they run , optimising the fair odds is the goal.
 
Back
Top