Process OR First Step OF A 2 Step Process



Process OR First Step OF A 2 Step Process

My diagram will look like a Chinese crossword puzzle

But I will try and explain it as best as I can.

[pic]

1- The first thing that you have to do is click “List Cards” and “List Results”.

Be sure that the appropriate files reside in the appropriate folders (race cards in “tsn1” and results in “tsnresults1”. Also, make sure those 2 folders reside in the appropriate directories:

“C:\tsn1\” for race cards and “C:\tsnresults1\” for results.

Click on “With Results?”, only, and not “Daily Play?”.

2 – Once your cards and results appear, select the series of race cards you wish to enter in the “Race Card(s) Selection” dialog box, similar to the way I entered them for AQU.

The races you have indicated will formulate your “ZZOUTPUT” file when you click on “Run Data”. This may take some time to complete so be patient.

When it does complete, you will see a dialog box (Timer) come up letting you know this.

Click on “OK”

Then you click on the “Output” button and your “ZZOUTPUT” file, for AQU, will have been created and reside in your “tsnresults1” folder.

3. You can now close the “AllData NI Batch” workbook and open the “AllDataBase” workbook.

[pic]

AllDataBase, Sheet “D”

Now that you’re race cards and results have been processed in AllData NI Batch and copied into “ZZOUTPUT.xls” which now resides in your “tsnresults1” folder, you are ready to import that “ZZOUTPUT” file into AllDataBase.

[pic]

Make sure the 2 “file folder paths” in cells Z1 and Z2 match the directories you had in AllData NI Batch: “C:\tsn1\” for race cards, and, “C:\tsnresults1\” for results.

1. Before you import “ZZOUTPUT.xls” into AllDataBase, you have 2 options: you can add the ZZOUTPUT race cards and results data to your existing database, if you have imported data previously, or, you can create a fresh new database, consisting of only the ZZOUTPUT data you’re about to import. If you want to erase a previous database, first, and create a new database, go to sheet “C” and click the “ERASE” button. Then return to sheet “D” and continue with Step 2. If you want to add the ZZOUTPUT data to data you already have in this database, just continue with Step 2.

2. Click “Import Database”, the “ZZoutput.xls” file will be copied to sheet “C”, starting in cell A17 and extending to column BZ. Just check this area and assure that the correct race cards appear here.

3. Click “Reset”, this will reset all filter settings, except “Date” and “Distance” back to default settings, open the “Distance” filter settings, so that all races can be queried, by setting the “Distance” settings to: -1 and 5000, this setting means that all distances from 0 yards (0 furlongs) to 4999 yards (22+ furlongs) will be queriable.

[pic]

4. Click “Query Database”, when the query finishes running, click “Unique Dist”, this lists all race distances in the database and how many plays there are for each “unique distance”. To see these unique distances, scroll down below the graph, you’ll see something very similar to this:

[pic]

As you can see, in this database, there are 10 unique distances ranging from 4.50 f to 9.00 f. You also see the distances in yards to aid you in setting the distance filter settings (they must be entered in yards, not furlongs). Beside each yardage is the number of plays for that distance. So, if you wanted to look at only 8.00 f races, you would see that there are 443 plays, at 8 f in this database.

5. Ok, now you’re ready to start getting some research done. First you have to decide what races or race types you want to research. Do you have data from more than 1 track in this database? If so, you might want to separate the database by track. Or you might want to research only “dirt” races, or “claiming” races, or races run on “sloppy” surfaces, etc..

[pic]

6. These types of “divisions” of the database can be accomplished in row 21, starting in column A and extending to column G, as shown in the above screenshot. As you can see, I have decided to research only races run at “HOU” (the track code for Sam Houston Raceway) and I only want to look at races run on the dirt (D).

7. With the 2 filters I have set I, if I query the database without setting any other filters I will be dividing my database into only races from Houston and only on the dirt. So, if this is what I want, for now, then I am ready to divide the database.

8. Click “Query Database”. Now look at the area below the graph.

[pic]

To the right of the “Divide 1st Segment” button, you see “NUM/SAMP” and below that you see “285”. This means that there are 285 races, in this sample, that are from Houston and were run on the dirt. Below that you see

START

1

END

285

I have decided to look at all 285 races from Houston that were run on the dirt. Click “Divide 1st Segment”. Look at the graph:

[pic]

As you can see, my starting bankroll was $1000 and declined steadily to $0 if I had bet every horse in every race in this sample, by the 519th play. Now click “Analyze”. Now scroll over to the right of the graph and you’ll see this:

[pic]

This is the “Analyze Utility”, in this area you see 3 groups of numbers under each filterable factor. The “RUNNERS”, in blue highlighting, shows the number of horses involved in all these races.

The top group shows the winners’ average rankings for each factor, and below that the place horses’ average rankings, below that the show horses’ average rankings, etc..

The middle group shows the top ranked horses’ win % for each factor, below that the 2nd ranked horses’ win %, below that the 3rd ranked horses’ win%, etc..

The bottom group shows the top ranked horses’ ROI, below that the 2nd ranked horses’ ROI, below that the 3rd ranked horses’ ROI, etc..

What we are looking for is a factor that shows both a good win % and a good ROI.

We’re only looking at the first 7 factors, if you were to scroll further to the right you would see all the other factors listed just like these are. As you can see, in this screenshot, there are 2 factors here that have decent win % but none with a good ROI, I could show the other factors, too, but believe me when I say that none of our factors, by themselves, showed both a good win% and a good ROI. This is common, as we are looking at a very broad range of races and distances. Almost never will you find a single factor that produces good win % and good ROI, when looking at such a broad sample of races.

9. Scroll back over to the main filter area as shown in this screenshot:

[pic]

As you see, in cells B2 and B3, I have changed the “Distance” filter settings to a minimum of 1300 yards and a maximum of 1350 yards. This setting means that when I query again only races that were run at distances from 1301 yards to 1349 yards will be returned. Loo at cells K2 and K3, you see 6.14 and 5.91. These are the equivalent furlong distances for the 2 yardage distances I entered as distance filter settings. What falls between 5.91 furlongs and 6.14 furlongs? There’s only one race distance that falls between that, 6f. So, I have set my distance filter to return only races run at Houston, on the dirt, and only at 6f.

10. Now, click “Divide 1st Segment”. The graph looks a little different, not good but a little better than what we had before.

[pic]

Now click “Analyze” and scroll over to the “Analyze Utility” area:

[pic]

Ah, things look a little more promising now. We have decent Win % and ROI in several factor categories: “Average Win%”, “Best Speed @ Distance”, good old TSN/Bris Prime Power, and, “Best L2 Comp2”.

Of these 4 positive factor, “Best L2 Comp2” has the best ROI, by far, 45.9%. That’s pretty darn good, and a good starting place for our first “model” or as we call it in AllDataBase, “Setup”. Look at the top group of numbers for that factor. The winning horses ranked an average of 7.5 for this factor. This gives us an idea of how we need to set our filter settings for this factor. I usually drop down a point, or 2, below that for my minimum setting and leave the maximum at 11, so our setting for “Best L2 Comp2” would be a minimum of 7 and a maximum of 11, this will return all horses that ranked 8th, 9th, or 10th in this factor for all the races we are researching, Houston, dirt, 6f.

[pic]

As you can see in the screenshot I have changed the filter settings for “Best L2 Comp2” to 7 and 11, which will return 8, 9, and 10 ranked horses, 10 being the best rank possible. Now let’s query this factor setting and see what happens to our graph. Click “Divide 1st Segment”:

[pic]

Wow! Our graph looks much better, not good yet, but better, we’re on the right track. Let’s change out filter settings again and see if the graph improves. Let’s drop the 8th , and 9th ranked horses out of the setting and see how the top ranked horses do. Let’s change the settings to 9 and 11, which will return 10, the top ranked horses in this factor.

[pic]

You see that I have changed the settings for “Best L2 Comp2” to 9 and 11, which will return 10, the top ranked horses only. Now click “Divide 1st Segment”, again, and look at the graph:

[pic]

Does that look a little better? Our bankroll increased from $1000 to slightly less than $1600, and we made 67 plays. Not bad, a little herky-jerky, but we can probably smooth that out some with the addition of another factor or 2. First, when you find a setup like this, that shows a lot of promise, it’s a good idea to save it, so you can continue to experiment with factors but won’t forget the one you started with. So----

11. Click “Save 1st Analyze”. Now look below the “Analyze Utility” area and you’ll see an exact copy of it, so you don’t forget where you started. Each time you develop a better setup, just click “Save 1st Analyze” and it will be copied below the “Analyze Utility” area.

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download