FYI Re. TSN PCS file update

[BillW]

TSN has published a file format document for their PCS files. Included are several additions that have yet to appear in the files, but I assume that posting this document implies that the new data should begin showing up soon. The format document is linked to the top of the PCS file d'load page.

I don't work for TSN, just watch their file formats closely for changes and undocumented data. :)

Bill

[Handle]

Thanks for the heads up. I'm a bit skeptical though.

For one, the structure says that PCS files have Jockey/Trainer previous year and current year records -- they don't. Just a whole lot of empy fields down around field. 1147:
,,,,,,,,,,,,,,,,,,,,,


Both the PCS and the 50 cent Data Cards specify that they have these gems (trainer stats):

1337 Key Trnr Stat Category #1 X(16) 16
1338 # of starts #1 9999 4
1339 Win% #1 999.99 6
1340 in-the-money (itm)% #1 999.99 6
1341 $2ReturnOnInvestment #1 999.99 6
1342 Key Trnr Stat Category #2 X(16) 16
1343 # of starts #2 9999 4
1344 Win% #2 999.99 6
1345 in-the-money (itm)% #2 999.99 6
1346 $2ReturnOnInvestment #2 999.99 6
1347 Key Trnr Stat Category #3 X(16) 16
1348 # of starts #3 9999 4

They don't. After BRIS began including them I asked TSN if they planned on putting them in. At the time they said "no way, Jose". I don't know why they said that, since my name is not Jose.

So, file structures are good -- accurate file structures a bit more desirable.

-Nathan

[BillW]

Nathan,

A few weeks ago they did "pre-announce" improvements coming soon, including the trainer stats (I think it was in their monthly e-mail newsletter). I seem to remember it being coupled to the introduction of their new PP formatter SW.

My experience with Bris/TSN is that they are guilty of putting undocumented data into their files more often than leaving documented data out. Hopefully they will continue that practice. In any case it's nice to see them attempt to document the format.

Later,

Bill

[delayjf]

BillW,
What undocumented data are you refering to. I ask because I'm haveing a program created that will use the TSN files.

[BillW]

Originally posted by delayjf
BillW,
What undocumented data are you refering to. I ask because I'm haveing a program created that will use the TSN files.

What I was refering to was for instance the jockey data (fields 1367 thru 1373) have been in the PCS files for a while now with no mention in the Multicaps file spec or either (Bris/TSN) single file spec.

Field 11 has been present in the Bris single data files for well over a year now and just documented recently. (I've been using TSN PCS files for the last 3 or 4 months, so I don't know how recently)

I don't have an exhaustive list handy, but these fields are easy enough to spot by just going through the file and comparing to the published specs.

Good luck,

Bill

[BillW]

Some of the documented changes have appeared in the PCS files as of 11/01.

Fields 1332 thru 1336 (Horse's record FAST DIRT) and fields 1337 thru 1365 (Trainer angle data) is now appearing in the PCS files. I am told by TSN that only a maximum of 3 trainer angles will be present in these files (There is room reserved for 6). I haven't independently verified this.

Still missing from the files are fields 1147 thru 1166 (Trainer/Jockey records for Current/Previous year) and fields 1328 thru 1331 (best TSN speeds: Life/Most Recent Year/2nd Most Recent Year/At Track). I'm talking to TSN about this and will post an update when available.

Later,

Bill

[azmike]

BIll W:

I don't use Bris or TSN yet. I am looking at a program that can use Bris or PCS files. I would like to go with TSN PCS files because of the $59.99 per month unlimited plan.

Are the files the same? If not, what would I be missing by going with TSN PCS files.

Thanks,

Mike

[BillW]

azmike,

The answer is a bit more complicated than the question.

The $0.50 TSN files, PCS files, Bris Single/Multi files and Bris Multicap files all use the same format, hence are interchangeable in some commercial software. Two types of differences differentiate the files. The first type is where data in the same field is of a different format (jockey/trainer names, Speed figs, race classification) and the second type where a field is "reserved" or empty in one file and contains data in the other (DRF SR/TV in Bris files ... empty in PCS and TSN single files).

Assuming you are talking about the Bris $1 single/multiple pp files (and not the $2 multicap files that are the Bris equivalent to the PCS files).

The speed figs are different and can't be mixed. The Bris files contain figs calculated by Bris with their magic formula. I'm not sure of the origin of the figs. in the TSN files ... possibly Equibase?

The TSN files do not have the DRF SR/TV numbers and contain the Equibase nomenclature for race classifications in the PP's rather than DRF.

Bris files do not contain and race ratings /class ratings for each pp, the PCS files do. (same for Recency class rating)

PCS files do not contain Current/Previous year jockey/trainer data.

PCS files do not contain "Best speed" #'s .

Trainer/jockey names are inconsistent. (as they are from entry to pp in each file :))

The above list is off the top of my head ... the file formats are documented on the respective sites.
The Bris document is in the Library section of the site and TSN document is on the product page for the PCS files.

All the above being said, if you are using a commercial package only, you will probably notice little difference. You should contact the author of the software you are interested in and discuss it with them.

Hope this helps,

Bill

[azmike]

Bill:

I appreciate your time and effort in answering my question. I will contact the author of the software per your suggestion.

Mike

[takeout]

Originally posted by BillW
The TSN files do not have the DRF SR/TV numbers and contain the Equibase nomenclature for race classifications in the PP's rather than DRF.

The nomenclature for the race classifications as well as the trainer and jockey names are something that I would've thought that EQ and DRF would've gotten synchronized on when they kissed and made up. Was I wrong or what! Switching files for a program based on trainer names is a headache of major proportions and minor track players are still getting screwed with DRF's nonexistent "nonwinners of" (in so many months) category.

[BillW]

Originally posted by takeout
The nomenclature for the race classifications as well as the trainer and jockey names are something that I would've thought that EQ and DRF would've gotten synchronized on when they kissed and made up. Was I wrong or what! Switching files for a program based on trainer names is a headache of major proportions and minor track players are still getting screwed with DRF's nonexistent "nonwinners of" (in so many months) category.

takeout,

The Bris files (DRF) have the "non-winners of" in the race condition, TSN files don't (C1000n2l vs. Clm10000). In the TSN files, you have to parse it out of the race description text. The number of months is never designated in DRF data and only designated in past performance data in TSN files (C10000n1y vs Clm10000NW16M). "Todays" race condition is better defined in the Bris files and past performance race conditions better in the TSN files.

The trainer/jockey names are inconsistant from past-performance data to entry data in the same file as well as from Bris -> TSN and vice versa. Usually the differences are consistant, and a translator can be written.

If you are using someone elses software, you are at their mercy. :mad:

Bill

[takeout]

The number of months not being designated in the DRF past performance lines is the one that I find unconscionable. This is not some subjective number but actual factual information that they choose to leave out. I think it's ironic that some of the well known writers that are now associated with the publication seem to be blind to this missing data as they are the ones that taught many of us, via their books, what race conditions were in the first place.

[rrbauer]

BillW wrote:
My experience with Bris/TSN is that they are guilty of putting undocumented data into their files more often than leaving documented data out. Hopefully they will continue that practice. In any case it's nice to see them attempt to document the format.


Perfecto Mundo, Bill. Every year or so, all of a sudden my parsing of a BRIS comma-delimited data file blows up due to an illegal data type. I start checking and sure enough they have added a field(s) right in the middle of a record and voila: Everything gets shifted and all bets are off. Of course they never give you a heads up or any clue that they've made a change. I send a scathing email to Happy Broadbent or Corby Corbett and the reply is usually something like, "Gee! Sorry about that!"

[Tom]

Originally posted by takeout
The number of months not being designated in the DRF past performance lines is the one that I find unconscionable. This is not some subjective number but actual factual information that they choose to leave out. I think it's ironic that some of the well known writers that are now associated with the publication seem to be blind to this missing data as they are the ones that taught many of us, via their books, what race conditions were in the first place.

The operative word here is "writers," then, they were trying to sell us books. Today, they are trying to sell us newspapers.