Your Ask The Experts ID
is separate from your
Order Online Account ID
 Race of the Week:  2024 Kentucky Oaks/Derby Days Final Figures Churchill Downs May 4 & May 5, 2024 
Order Online
Buy TG Data
Complete Menu of
TG Data products
Simulcast Books
Customize a Value
Package of Select
TG Data
Sheet Requests
Order The Last Figure for Any Horse
Free Products
Redboard Room
Download and Review previous days' data.
Race of the Week
With detailed comments
ThoroTrack
Email notification when your horse races
Information
Introduction
For newcomers.
Samples and Tutorials
For Horsemen
Consulting services and Graph Racing
Sales Sites
Where to buy TG around the country
Archives
Historical races and handicapping articles
Handicapping
Hall of Fame
Major handicapping contest winners
Home Page
Re: Weirdness (617 Views)
Posted by: TGAB (IP Logged)
Date: March 24, 2006 03:52PM

Here's the problem. Equibase, our data source provider, switched over to a new file format in the first week of the new year. One of the disadvantages we find is the procedure for amending, updating European race results. The procedure calls for deleting the entire record on file and resending the complete updated new record. Prior to the new year, only the amended variables were resent; the entire record was not deleted, only the changed value. That's the way US results are updated and have been for a while, no change in the new year.

How does the procedure change manifest itself? Well, your seeing some of manifestations---the multiple gelding notations, duplicate races.

To be fair it should be said that Equibase is responsible for the accuracy of US race results, not European results. But as the major US race database provider, they do gather European race results and provide them to customers. That involves getting results from many different foreign organizations over which they have no control. Their primary purpose is make sure that foreign result data are as complete as possible for foreign runners that show up in the US.

Remember you are looking at a Dubai card, not a US race card. Also it should be noted that race and horse information records are voluminous. There's a lot of information you don't see, we don't use. For instance, internal call data, earnings, complete races, records, ownership, riding colors, etc. Any time one of these change for a foreign race we already have stored, the entire record is zapped and replaced.

The problems come because once upon a time we tried gather and code gelding information. Obviously this is a major change to a colt or horse's physical being which often has imminent repercussions. The Jockey Club gathers and reports this info but there's no timely notification process which makes the reporting of it haphazard at best. For this reason we abandoned the effort and purposely chose not to denote it on the sheets, except in the breeding line atop the sheet. We stored the record and chose not to pick subsequent changes except as noted above. However with European records now, whatever we have in the database gets zapped and resent everytime the horse appears. So I spent a good deal of time yesterday going over each horse and looking for erroneous or multiple gelding codes and zapping them horse by horse. In addition we were also appending omitted past performances and TG figures for these performances as well. It was time consuming and a lot of work but necessary to get out the card. I obviously missed a horse or two but believe me there were a lot more castration notices.

Now the duplicate races. You've probably noticed a lot more duplicates races, boxes, since the new year started. A long time ago we realized the problems inherent with foreign data. We had our programmer develop software to allow us to append race information omitted. So we took whatever Equibase had, added omissions, appended figures, and fixed the multiple codings. Once stored, prior to this year, we only updated relevant variables. When horses showed up again, we only had to check subsequent information. Not so this year with data format change. In addition to the procedure change, Equibase decided to code foreign race numbers differently with three, four and five number designations. These designations never match with what we have, like 1,2, 3 etc., so they always get added. Hence we have duplicates. I tried to get rid of all of those as well but obvious was unsuccessful and frankly on some days when I'm pressed for time, I don't bother deleting the duplicates, the figure takes priority to the cosmetic changes although I do realize the duplicate boxes are annoying and can be confusing. Bear with us we've put this on the list for the programmer to fix, but it hasn't been done as yet.

It's bad for us inhouse as well. Frankly, if the procedure hadn't changed, we'd
could get rid of the boxes, gelding listings one time and that would be it. Also now if we enter a figure for new foreign race record, it will get zapped once it shows up again, unless we manually enter and join the result to the figure which would result in a duplicate listing since we have no idea of the race number designation scheme Equibase uses. Believe me this is a problem and we will fix it in the near future.

For some races we are not able to make a figure. The figures we do show do not include ground loss information but we think, they're pretty accurate, regardless.

Now you know some of the back room doings and problems we face dealing with foreign horses and presenting data for foreign races.

Go Super Frolic.

Just saw the question concerning the race summary discrepancies. A few points here. One, the DRF has more resources at hand to gather, report, and ensure the accuracy of foreign race data and the guy they got in charge there is good at what he does. Two, and this is speculative, but they either didn't face or have overcome some of the data problems we confront now. Three, and this is most important, focus. Our focus is making the figure and if we show a figure for a race that figure reflects accurate inputs of race result data. I realize that point is subject to question when the race summary is incorrect but I assure you, a Thoro-Graph figure reflects accurate inputs. Often we don't see the race summaries until they're printed out and it takes too much time to fix them and if we do we run into the problems detailed above. The DRF shows running lines with a comment and that's their main focus. Not so with us. The figure is the main focus and we use accurate source materials to ensure the accuracy of the figure we make to the fullest extent possible. You can get running lines from various sources which is why race summary accuracy is a lower priority. And by the way you won't see these inaccuracies with domestic data.




TGAB



Edited 4 time(s). Last edit at 03/24/2006 05:00PM by TGAB.



Subject Written By Posted
Weirdness (983 Views) SoCalMan2 03/24/2006 10:45AM
Re: Weirdness (599 Views) sighthound 03/24/2006 11:24AM
Re: Weirdness (585 Views) elkurzhal 03/24/2006 11:47AM
Re: Weirdness Continues (573 Views) Upper Nile 03/24/2006 01:37PM
Re: Weirdness Continues (570 Views) Upper Nile 03/24/2006 01:46PM
Re: Weirdness (617 Views) TGAB 03/24/2006 03:52PM
Re: super frolic (612 Views) RICH 03/24/2006 04:13PM
Re: super frolic (561 Views) TGAB 03/24/2006 04:37PM
Re: super frolic (598 Views) Chuckles_the_Clown2 03/24/2006 09:14PM
Re: super frolic (557 Views) bobphilo 03/24/2006 09:44PM
Re: super frolic (518 Views) ROBERT49 03/24/2006 09:50PM
Re: super frolic (582 Views) bobphilo 03/24/2006 09:57PM
Re: super frolic (824 Views) TGAB 03/24/2006 10:42PM
Re: super frolic (563 Views) tmon 03/24/2006 11:07PM
Re: super frolic (579 Views) TGAB 03/24/2006 11:40PM
Re: Weirdness (571 Views) tmon 03/24/2006 05:39PM
Re: Weirdness (532 Views) davidrex 03/24/2006 05:43PM


Sorry, only registered users may post in this forum.
This forum powered by Phorum.

Thoro-Graph 180 Varick Street New York, NY 10014 ---- Click here for the Ask The Experts Archives.