[Ted Craven]

I spoke just now with both Dave Siegel, TrackMaster President and Jim Vanderbosch VP about the data error reported by Mike W and confirmed by others in this thread.

The short report is: they acknowledge the data error and today have now re-instituted a previously discontinued Quality Control check to scan for such missing values BEFORE all Past Performance data files are published for download either by RDSS or by browser from their website (e.g. for use by Speculator or Validator). This particular QC check was discontinued at some point in the past because it never found any problems, but it's back in effect now and the data problems should not recur.

The longer report is: they maintain a database which includes about 10 million thoroughbred running lines going back about 10 years. Due to some 'deep-in-the-server' issues which occur very occasionally during updates from Equibase, a few of their Speed Ratings and Inter Track Variants (ITVs) which were formerly correct become missing. Siegel told me that yesterday, in response to my enquiry following Mike W's report, they ran a query of their database over the past 2 years, about 1 million running line records, and found a total of 8 records which had missing SR/ITV values (of which we reported 2 of those - this one, and another which RichieP reported re PPs for a race run last night). All 8 (out of 1 million) of these records with missing SR/ITV values were re-run and corrected. Siegel speculated that at any time, even over the past year, there may have been a similar number of records which go bad and are then fixed by periodic review of data-integrity checking scans: 2 out of a million, 10 out of a million, whatever - a small number.

Now, as a result of these reports by us, as stated above, they are running their reinstated QC scan on ALL their PP files for ALL their products BEFORE they are published, looking for any missing values and will rerun any PP files if any missing values are ever found. Siegel told me that he is extremely confident that this re-instated PP file check should result in no further missing SR or ITV values going public. That said, he urged me to communicate to all customers that if EVER they find what looks like it might be 'bad data' to feel free to contact them directly either by phone (800 321-3343) or email (jim@trackmaster.com) with details, and not only to report it here at Pace and Cap or to me. I have seen over the past year how their level of responsiveness to Sartin software users has grown (commensurate with how much more data they're selling and hope to sell). Believe me, they are paying attention now!

From the sounds of it, they took this last report very seriously and responded immediately and in some depth to our concerns. I personally am satisfied that they are presently doing all they can re quality of PP data, given their data volume, but I do think they need to work further on what causes even such a small number of previously good records to occasionally go bad, and told Siegel so.

And remember that the bad data we're referring to is, effectively, the TrackMaster/Equibase Inter Track Variants used to adjust running times - not running times themselves, or positions or beaten lengths or money stats. You should expect a valid TrackMaster Speed Rating (and an underlying ITV) if a thoroughbred race has:

1. a Final Time
2. the horse finished the race
3. there are beaten lengths reported (for other than the winner)
4. it's not a steeplechase

If any of these is missing, the TrackMaster Speed Rating (and ITV) will be missing and this is NOT an example of 'bad data' which needs to be reported to TrackMaster. Otherwise, report it to TrackMaster and report it to me, please. I need to update RDSS (next scheduled major update) to watch for such missing items and simply mark the PP line as unusable (if ITVs are being used). Spec and Val appear to do this already, I think.

Regarding a 'certificate of authenticity' - I think that can never be 'given' by them to us, but rather must be given by us - to them. Siegel told me he definitely wants to know, and directly, if any apparently bad data is found. Each person must decide for themselves how long to wait to judge the data 'authentic'. If anyone wants a refund from TrackMaster for their data account, or from me for their RDSS subscription, just write them, or me - no worries. For myself, I'm going to move on and use the TrackMaster data and RDSS - and encourage other RDSS, Speculator and Validator users likewise - and continue to do my best to make a profit on my wagers ongoing. The data and the software are more than good enough to accomplish that.

All questions and comments are welcome!

Ted