Home › Forums › Archive Topics › Trends, Research And Notebooks › onTarget Trends
- This topic has 93 replies, 14 voices, and was last updated 19 years, 4 months ago by
carlisle.
- AuthorPosts
- December 1, 2006 at 19:20 #32136
Carlisle – can you write script for extracting data?
December 1, 2006 at 20:40 #32137Hi cormack
No,  but I don’t think that is the way to go.  Data can be accessed from a commerical database. It will not be a problem, trust me.
Exactly what data fields do you need?
byefrom<br>carlisle
December 1, 2006 at 21:14 #32138Ok C –
for last ten years King George lets see if we can get for each runner –
Year<br>Horses name<br>Horses finishing position<br>Trainer<br>Age<br>Weight carried<br>SP <br>Whether fav or not<br>Dyas since last run<br>Last Time Out form fig<br>Class of race ran in LTO<br>Runs that jumps season (since July 1st)<br>Best previosu RPR<br>last time out RPR<br>Whether won over distance
Let me know if that can be readily accessed.
December 1, 2006 at 23:20 #32139Hi cormack
lots of the data is easily be scalped from the race card, we <br>obviously don’t have 10 year of King George race cards.<br>I take it that you still want to produce, all, your own trends.
The following in red will have to be entered manually
Year <br>Horses name <br>Horses finishing position <br>Trainer <br>Age <br>Weight carried <br>SP <br>Whether fav or not<br>Days since last run<br>Last Time Out form fig<br>Class of race ran in LTO<br>Runs that jumps season (since July 1st) <br>Best previousRPR <br>last time out RPR <br>Whether won over distance
I will help with some manual data entry.
byefrom<br>carlisle
ps researching 10 years of race data is trickier than<br>providing data for a fully declared yet to be run race.
pps I will try to find another way, if possible. Has anyone else got any good ideas.
amended…………
I can do RaceForm Ratings but not RPR’s<br>
(Edited by carlisle at 12:29 am on Dec. 2, 2006)
December 1, 2006 at 23:28 #32140"I take it that you still want to produce, all, your own trends."
What’s the alternative to that Carlisle – isn’t that the whole point?
Reading back thru the thread were you perhaps meaning we should be aiming at producing a rating from teh data put up by the three trendors? If so I can’rt see how that could be done.
But I’m willing to listen.<br>
December 2, 2006 at 00:03 #32141Hi cormack
they have essentially, already, highlighted the trends.<br>I feel that there is no need to perform a wholesale trend production project.
For example if the stated trend was…………<br>At least 8 runs over fences  9/10<br>If true give the horse 9 marks<br>else give the horse 1 mark.
Another more involved way of giving marks can be devised.
I have amended my previous posting.
byefrom<br>carlisle<br>
December 2, 2006 at 00:24 #32142So, let me get myslef straight on this –
You’re suggesting we look at the articles produced by those 3 each week and, using a scoring system, allocate marks/scores and calculate a trend rating from that?
I’m WAY down another path! lol
What I am looking at is doing our own trend analysis on historical data to ascertain the key trends and their relative importance in each race (it varies race to race of course) and then using that data to produce a weighted trend rating for each runner.
I’m really struggling to see how we could get consistent data from those articles to produce anything meaningful, although I’d be delighted to be proved wrong!
As an exercise – could you perhaps look at those three articles this weekend and come up with some numbers. If I could see a way to get a set of numbers from the articles arriving at a trend rating would be a straightforward number crunch.
Meanwhile – back to the King George data extraction (manual)for me!! (of which more later)<br>
December 2, 2006 at 00:41 #32143Hi cormack
view the amended posting, I think manual inputs can now be minimised. I will let you have all the data you requested on Sunday.
As I have previously mentioned, lets pursue different methods. Then compare our findings, their is no conflict of interests. Lets see what both methods come up with.
I am sure I will learn lots along the WAY.
byefrom<br>carlisle
ps goodnight.<br>
December 2, 2006 at 00:45 #32144Ok Carlisle – agreed – I’ll keep you informed here on progress and you do likewise.
Good stuff – good to have a thread on here with an actual outcome in terms of action/activity. I am absolutely determined that, by the time we’re finished, one way or another, we’ll have a trend rating mechanism devised (even if I have to make them up!!).
December 2, 2006 at 00:50 #32145mutiny :)
December 2, 2006 at 01:00 #32146lol
December 2, 2006 at 12:15 #32147and today’s selection is … ?
December 2, 2006 at 12:50 #32148Hi dave jay
As far as I am concerned, there will never be a trends selection.  My idea is to produce a list of Race Trend Ratings, that can be viewed alongside other forms of analysis.
"men at work…….."
byefrom<br>carlisle
(Edited by carlisle at 12:52 pm on Dec. 2, 2006)
December 2, 2006 at 19:48 #32149Oh right.
December 2, 2006 at 20:06 #32150Come on Dave – give us a chance!
December 2, 2006 at 20:32 #32151You’ve had a week and reading the last few posts you still don’t know what you are actually trying to achieve .. if anything !<br>:biggrin:
December 2, 2006 at 21:35 #32152evolution.
Besides I have been cast adrift with all these bread fruits………
ps cormack I am working on your data.
<br>
- AuthorPosts
- You must be logged in to reply to this topic.