ALL DATA Free

[headhawg]

Originally Posted by headhawg

I have a feeling that he will return but that's just speculation on my part.

If only I could pick horses like this. Welcome back raybo.

[raybo]

Originally Posted by headhawg

If only I could pick horses like this. Welcome back raybo.

Thanks! I missed this place.

[maddog42]

Originally Posted by headhawg

If only I could pick horses like this. Welcome back raybo.

Welcome back Raybo !!!!!
I'm a little slow on the realization that you are back. I tend to Zoom-in on the threads that I have been posting on, to the exclusion of others. This Forum is
better with you back, and is truly the BEST on the Web.

[raybo]

Maddog,

Thanks!!

[BIG HIT]

Glad your back pace advantage and you.Are a couple guy's never would have meant if not for computers which i terrible with.Two men that both man up and solve thier difference and move on.Maybe the country leaders should check out pace advantage on how to be leaders LOL.

[raybo]

Thanks to one of our users we have discovered a problem with AllData NI Batch and AllData_J1. It shows up on race cards that have a large number of horses. The card we found it in was the Saratoga card for 8/27/2011. In the Travers, race 12, only 7 horses were processed, and only one horse was processed in race 13. One of the macros in the workbook was set to look at 2000 records in the raw data file, that's about 133.33 horses. On the Saratoga card, the first 11 races had 127 horses entered, which meant that in race 12 the macro wold only look at the first 6+ horses.

Harry found the problem and passed it on to me. I have made the change to AllData_J1 on the AllData website. For those of you who already have AllData_J1, or any of the earlier versions of AllData NI Batch on your computer, you can make the change very easily, and don't need to re-download the workbook. Here's what you need to do to fix the problem.

Open AllData and go to the "Summaries" worksheet. Scroll up to the top of the worksheet, then scroll to the right until you can see column "DS". Click in cell "DS6", then look up in your formula bar, just above the column labels and you'll see the following formula there:

=COUNTIF(DL1:DL2000, ">0")+1

In that formula, change the "DL2000" portion to "DL3500". Then click the green check mark to the left of the formula. Then save your workbook.

The corrected formula should look like this:

=COUNTIF(DL1:DL3500, ">0")+1

That's all it takes to fix the problem.

If you have problems making the correction, let us know here or by email.

By the way, the "Pace Picture" and the "Black Box" had the winner in the Travers, using the default paceline settings. Of course, it was the post time favorite, but what the hey, at least you could have said you had it.

As a matter of fact, the "Black Box" had 3 of the top 4 finishers in the Travers, in it's top 4 picks. Only missed the place horse.

[raybo]

The latest update to "AllData_J1.xls" is uploaded on the AllData website. The download fixes a couple of formulas on the "Summary Distance Equalizations and Pace Ratings" view.

Here's the link to the download page, on the site (scroll down to the bottom of the page for the download):

AllData NI Batch - Download page

[jeebus1083]

Raybo,

We exchanged e-mails awhile back regarding this, and you disagreed at the time, but I have to maintain my stance on this: The cells for the positional points of call are not consistent with the cells for the fractional points of call, and will result in the inaccuracy of the Sartin data. I compared the data to the Points of Call index in the Brisnet library, and found this discrepency.

Past Performances are generally presented as 5 positional calls: Start, 1st Call, 2nd Call, Stretch Call, and Finish. Depending on the distance, the positional calls change. Since Sartin information is extracted from the 1st, 2nd, and Finish positional calls, you have to use the appropriate fractional calls that match the positional calls, otherwise, the positions/beaten lengths will not be accurate for the point of call.

I took the liberty of making the appropriate changes to my version of AllData, and they now come out right. Races that are less than 6F will be blank, as the 1st positional call is for the 3/16 call, and the 1st fractional call is the 1/4 mile - there is no 3/16 fraction recorded. Also, races greater than 1 3/4 miles (same distance as the BC Marathon) also come up blank, since there is no way to fit the appropriate positional/fractional data correctly beyond that. Besides, there are not many everyday races that run marathon distances.

I made changes to the 1st call/2nd call/Finish call distance by feet in the "Distance Equalization" tab to reflect the changes made in AllData.

If you wish, I can e-mail you my version of AllData, complete with changes, and you can be the judge. However, I stand by my claim that the Sartin calculations in my modified version are correct, and the calculations in the version available for download off of your site are inaccurate.

[raybo]

Originally Posted by jeebus1083

Raybo,

We exchanged e-mails awhile back regarding this, and you disagreed at the time, but I have to maintain my stance on this: The cells for the positional points of call are not consistent with the cells for the fractional points of call, and will result in the inaccuracy of the Sartin data. I compared the data to the Points of Call index in the Brisnet library, and found this discrepency.

Past Performances are generally presented as 5 positional calls: Start, 1st Call, 2nd Call, Stretch Call, and Finish. Depending on the distance, the positional calls change. Since Sartin information is extracted from the 1st, 2nd, and Finish positional calls, you have to use the appropriate fractional calls that match the positional calls, otherwise, the positions/beaten lengths will not be accurate for the point of call.

I took the liberty of making the appropriate changes to my version of AllData, and they now come out right. Races that are less than 6F will be blank, as the 1st positional call is for the 3/16 call, and the 1st fractional call is the 1/4 mile - there is no 3/16 fraction recorded. Also, races greater than 1 3/4 miles (same distance as the BC Marathon) also come up blank, since there is no way to fit the appropriate positional/fractional data correctly beyond that. Besides, there are not many everyday races that run marathon distances.

I made changes to the 1st call/2nd call/Finish call distance by feet in the "Distance Equalization" tab to reflect the changes made in AllData.

If you wish, I can e-mail you my version of AllData, complete with changes, and you can be the judge. However, I stand by my claim that the Sartin calculations in my modified version are correct, and the calculations in the version available for download off of your site are inaccurate.

What I did in AllData, when the calls' distances differed from the fractional times distances, I defaulted to the fractional times distances. So, in a 5f race, the 1st call is at 3/16 and the 2nd call is at 3/8, but when you look at the fractional times, they are at the 1/4 and the 1/2, so, I used 1/4 and 1/2 as the call distances, otherwise you have a discrepancy between the calls and the fractional times. In my opinion, it is better to have a velocity for these fractional times than to have nothing at all for them.

Is this what you are referring to?

[jeebus1083]

Originally Posted by raybo

What I did in AllData, when the calls' distances differed from the fractional times distances, I defaulted to the fractional times distances. So, in a 5f race, the 1st call is at 3/16 and the 2nd call is at 3/8, but when you look at the fractional times, they are at the 1/4 and the 1/2, so, I used 1/4 and 1/2 as the call distances, otherwise you have a discrepancy between the calls and the fractional times. In my opinion, it is better to have a velocity for these fractional times than to have nothing at all for them.

Is this what you are referring to?

I should have specified in my initial response that the erroneous data was coming from route races, but failed to do so. I'm fine with what you did there (and will have to go back and fix), but when you cross over to route races, you get some problems. AllData processes the raw times using the 1st Fraction/2nd Fraction/3rd Fraction cells, which would be fine, but when you get to races 1 1/8M or greater, those fractional time cells do not align with the positional points of call.

Ex: In a 1 1/8M race, the 1st positional call is the 1/2, and the 2nd positional call is the 3/4. For some reason, AllData processes the 2nd fractional call time (3/4) as the 1st call, and the stretch fractional call (mile) as the 2nd call time. Because of this, the feet per traveled in distance is also incorrect.

To solve the problem, I ditched using the 1st/2nd/3rd fraction cells in the formulas. For the appropriate calls, I used the 2f/4f/6f/8f cells where appropriate and wrote the formulas as such. I also changed the feet traveled to reflect the positional points of call. I may have had to modify the position formulas, but I don't remember at this point. They now align just right.

I will send you the sheet tonight.

[jeebus1083]

Raybo, check your e-mail. It explains everything, and my version of the sheet (with said modifications) is also attached.

[raybo]

Originally Posted by jeebus1083

Raybo, check your e-mail. It explains everything, and my version of the sheet (with said modifications) is also attached.

Thanks, Tim, I'll get to it as soon as time allows!

[jeebus1083]

Sorry to beat the horse (figuratively speaking )... I did find (and fix) and error on the pace pressure gauge in your "Pace Picture"... The 2nd component of the pressure gauge are just the E horses with 5+ QSPs, not all E horses regardless of QSP points.

Needless to say, that is now fixed!

[raybo]

Originally Posted by jeebus1083

Sorry to beat the horse (figuratively speaking )... I did find (and fix) and error on the pace pressure gauge in your "Pace Picture"... The 2nd component of the pressure gauge are just the E horses with 5+ QSPs, not all E horses regardless of QSP points.

Needless to say, that is now fixed!

Tim,

Is the "E horses with 5+ QSPs" a recent change? I have not bought Randy's books but tried to reproduce information that was "public knowledge" on his website, for free, and not copywrite protected (to my knowledge). I could swear that, originally, the second number was the number of "E" horses, with no Quirin minimum applied.

If that is now true, I will make the changes to AllData.

Thanks again, for the "heads-up"!

[jeebus1083]

Originally Posted by raybo

Tim,

Is the "E horses with 5+ QSPs" a recent change? I have not bought Randy's books but tried to reproduce information that was "public knowledge" on his website, for free, and not copywrite protected (to my knowledge). I could swear that, originally, the second number was the number of "E" horses, with no Quirin minimum applied.

If that is now true, I will make the changes to AllData.

Thanks again, for the "heads-up"!

From PaceAppraiser.com:

Please note that the second number in the gauge represents velocity, which is the number of one dimensional early pace horses with 5 or more speed points. The higher the numbers in each category (pressure and velocity) the more likely the winner will be a P/C or C.