Understanding Track Weight JCapper

Understanding Track Weight

Author: Jeff Platt

Date Last Modified: October 14, 2007

Introduction

Back in 2001, when I first started wrapping my head around taking concepts like track weight and speed and shaping models into the JRating algorithm, PolyTrack didn't exist. All dirt tracks (taken as a whole) seemed to share one common trait: Horses with early speed enjoyed a distinct advantage over horses without it.

As a result I was able to create one JRating algorithm for dirt and another for turf. I then applied these two algorithms to all tracks everywhere. For the past several years that approach has let me (and many other JCapper users) enjoy almost universal success simply by playing JRating rank = 1 horses under a wide variety of circumstances.

When I set about developing JPR in early 2005 I used very similar thinking to what I did when I created the JRating. I created one JPR algorithm for dirt and a second algorithm for turf. And it worked pretty much everywhere I wanted to play.

The advent of PolyTrack changes that. It plays differently than plain old dirt. Horses running on or near the lead on PolyTrack no longer have an advantage over their rivals. Instead, horses with early speed on PolyTrack, unless they are in a race with little or no pace pressure, are at a disadvantage. I am convinced that poly is different enough that it deserves treatment as a separate surface.

As I write this (October, 2007) horses are now racing on artificial surfaces at the following Bris track codes: APX-DMR-HOL-OSA-PID-SAX-TPX-WOX. PolyTrack has become prevalent enough throughout racing that universal Data Window samples/results have become noticeably different from what they used to be.

We can argue about reasons both for and against PolyTrack until we are blue in the face. But the simple truth is this: It is now part of the game. There is a belief among horsemen that artificial surfaces are safer for horses to race on. Wait. It gets better. I have seen track management at some tracks have their maintenance crews change dirt surfaces that have historically favored early speed into deeper more tiring surfaces that now favor closers.

In my opinion the game is currently undergoing a landscape change.

Universal data samples of the type used as a basis for creating the My-Edge UDM in the Finding an Edge Help Doc need to be examined more closely. I have come to believe that UDMs and UPR should really be modeled on a track by track basis.

Not all racing surfaces are created equal. If you have ever undertaken the project of trying to measure track weight you know how complicated things can get in a hurry. You can get bogged down analyzing historical samples broken out by a myriad of factors. In JCapper we have CPace, PctE, V1, DecelFactor, ClosingRatio, AvgE1, BestE2, PAL, TPace, Late3, Running Style, Q Speed Points, Rail Position, and many others. All of them have some degree of effectiveness when trying to nail down a track profile. When trying to profile a track it’s easy to lose yourself in complexity.

The reason I decided to do a write up on track weight is that it doesn’t have to be that way. I wanted to present a process for working with track weight – a way that is both simple and workable – so that JCapper users everywhere could understand track weight and get a handle on using it when creating UDMs and UPR.

Racing Surface Categories

Every racing surface can be described as fitting into one of the following simple categories:

Category Description

1. Favors early speed strongly

2. Favors early speed somewhat

3. Plays fair

4. Favors closers somewhat

5. Favors closers strongly

The thing that confuses most players is that surfaces are not always consistent. A surface might play as a category 1 or 2 one meet and then behave as a category 4 the next. Sometimes, because of weather and track maintenance, this even happens from one week to the next.

But the player who is able to accurately recognize the prevailing current tendency of the racing surface(s) he or she is playing on is the player who can get a significant edge over his or her competition.

There are two things most players lack:

A simple workable way to measure and understand (what category) how a given racing surface has been playing.
Understanding how to take advantage of that knowledge.

Measuring Track Weight (Establishing a Benchmark)

The first step in understanding track weight is to establish a few benchmarks for each racing surface category. To this end I am going to present a simple UDM designed to do nothing more than help you measure track weight.

The theory behind my thinking is pretty simple. Run the Track Profiler UDM against a large sample (all tracks everywhere) to establish a benchmark. Then run the UDM against smaller samples (individual racing surfaces for specific time periods) to get snapshots representing a current track profile. Compare the two and categorize the individual racing surface based on the severity of any differences noted.

Horses that rank 1-3 in CPace vs. a given field are generally those with early speed. The idea behind my Track Profiler UDM is to provide a solid way of measuring the effectiveness of early speed horses. The UDM starts with factor constraints for CPace ranks 1-3. I chose ranks 1-3 rather than rank 1 only because ranks 1-3 provide more data to work with than rank 1 alone. You’ll see the importance of this once we dive into looking at smaller samples. For example, when playing a new track for the first time – I’ll often look at a track profile using just the most recent 3-4 days. In a case like that I want all the data I can get. CPace ranks 1-3 give me more to work with than CPace rank 1 alone.

If you study my Track Profiler UDM, you’ll notice I’ve added factor constraints for PaceIndex, Race Volatility, and Late3 rank. I did this in the interest of establishing a tighter benchmark to refer back to later - because not all CPace rank 1-3 horses are the same. These factors do impact the win rates of CPace horses. I wanted to control them later when running my Track Profiler UDM against small samples (snapshots) for specific time periods at single tracks.

But in your own implementation – when you create a Track Profiler UDM of your own – feel free to roll your own. It is the concept of measuring smaller samples (snapshots) against larger samples (established benchmarks) that is important here. Please do not feel obligated in any way to let yourself get bogged down in the minutia of what I’m trying to present here.

Below are Data Window results for my own Track Profiler UDM run against my calendar year 2006 database on dirt surfaces only:

UDM Definition: TRACKPROFILER

Divisor: # UDM Def Divisor: 999

Surface Req: *ANY Surface*

Distance Req: *ANY Distance*

CPace: MinRank= 1 MaxRank= 3

MinVal= -999 MaxVal= 999

MinGap= -999 MaxGap= 999

Late Pace (last3) MinRank= 2 MaxRank= 5

MinVal= -999 MaxVal= 999

MinGap= -18 MaxGap= 999

Pace Index: MinVal= 80 MaxVal= 999

Race Volatility: MinVal= 75 MaxVal= 119.99

Running Style: ALL

Data Window Settings:

Divisor = 999

Filters Applied:

Surface: (ALL*) Distance: (All*)

(From Index File: C:\2007\pl_CPace_123_2006.txt)

Data Summary Win Place Show

Mutuel Totals 13317.60 13694.80 13391.50

Bet -15696.00 -15696.00 -15696.00

Gain -2378.40 -2001.20 -2304.50

Wins 1574 3032 4121

Plays 7848 7848 7848

PCT .2006 .3863 .5251

ROI 0.8485 0.8725 0.8532

Avg Mut 8.46 4.52 3.25

And there you have it. The above numbers represent the benchmark numbers I actually use when establishing categories for racing surfaces.

Measuring Track Weight (Snapshots)

I’m going to use two very different racing surfaces to illustrate how I go about taking snapshots to measure track weight. The first track is APX where PolyTrack debuted with the start of their 2007 racing season. The second track is BEL which runs their races on plain old dirt. Both tracks produced widely different results. But in both cases, the process for measuring and categorizing track surface was exactly the same.

Arlington Park 2007

Let’s start with Arlington. I started by using the Quick Index File Extract tool to create a Date Ranged Index for the first two weeks of their 2007 meet. I then ran my Track Profiler UDM through the Data Window using the date ranged index file – which gives me a snapshot of what happened during the first two weeks of the meet. Here’s what the data looks like:

UDM Definition: TRACKPROFILER

Divisor: # UDM Def Divisor: 999

Surface Req: *ANY Surface*

Distance Req: *ANY Distance*

CPace: MinRank= 1 MaxRank= 3

MinVal= -999 MaxVal= 999

MinGap= -999 MaxGap= 999

Late Pace (last3) MinRank= 2 MaxRank= 5

MinVal= -999 MaxVal= 999

MinGap= -18 MaxGap= 999

Pace Index: MinVal= 80 MaxVal= 999

Race Volatility: MinVal= 75 MaxVal= 119.99

Running Style: ALL

Data Window Settings:

Divisor = 999

Filters Applied:

Surface: (ALL*) Distance: (All*)

(From Index File: C:\2007\APX\pl_From_05-04-2007_to_05-18-2007.txt)

Data Summary Win Place Show

Mutuel Totals 14.60 25.00 39.20

Bet -46.00 -46.00 -46.00

Gain -31.40 -21.00 -6.80

Wins 2 6 11

Plays 23 23 23

PCT .0870 .2609 .4783

ROI 0.3174 0.5435 0.8522

Avg Mut 7.30 4.17 3.56

Now, compare the above results for Arlington to the benchmark results for all dirt surfaces in 2006. What do you see? Based on what you see, how would YOU categorize the PolyTrack surface at APX? Forget the roi for a second. Pay attention to the win percentages. My own interpretation is that this snapshot makes the APX surface look an awful lot like a surface in the category 5 range: Favors closers strongly.

How can I say this? Well the win rate for horses selected by my Track Profiler UDM (early speed) in the snapshot is dismal compared to my Benchmark. Some group of horses has to be hitting the wire first in every data sample. It’s kind of like that piece of childhood playground equipment called a teeter-totter or see-saw: If early speed horses aren’t winning the races – if their win rates are substantially depressed – then some other group of horses has to be winning the races at an elevated win rate: closers.

Remember that when I created the algorithm for dirt JPR I based it on dirt surfaces behaving as categories 1-3. Here’s that two week APX snapshot again, this time broken out by JPR Rank:

Data Window Settings:

Divisor = 999

Dirt (All*) Distance: (All*)

From Index File: C:\2007\APX\pl_From_05-04-2007_to_05-18-2007.txt

Data Summary Win Place Show

Mutuel Totals 1300.00 1122.90 1134.80

Bet -1314.00 -1314.00 -1314.00

Gain -14.00 -191.10 -179.20

Wins 76 150 225

Plays 657 657 657

PCT .1157 .2283 .3425

ROI 0.9893 0.8546 0.8636

Avg Mut 17.11 7.49 5.04

By: JPR Rank

Rank Gain Bet Roi Wins Plays Pct Impact

1 -47.80 150.00 0.6813 15 75 .2000 1.7289

2 -56.60 150.00 0.6227 11 75 .1467 1.2679

3 2.00 150.00 1.0133 10 75 .1333 1.1526

4 -7.60 150.00 0.9493 7 75 .0933 0.8068

5 13.00 150.00 1.0867 12 75 .1600 1.3832

6 -31.00 150.00 0.7933 6 75 .0800 0.6916

7 83.80 138.00 1.6072 8 69 .1159 1.0023

8 8.00 114.00 1.0702 3 57 .0526 0.4550

9 19.80 76.00 1.2605 2 38 .0526 0.4550

10 -22.80 46.00 0.5043 1 23 .0435 0.3759

11 39.20 26.00 2.5077 1 13 .0769 0.6650

12 -14.00 14.00 0.0000 0 7 .0000 0.0000

13 0.00 0.00 0.0000 0 0 .0000 0.0000

14 0.00 0.00 0.0000 0 0 .0000 0.0000

15 0.00 0.00 0.0000 0 0 .0000 0.0000

16 0.00 0.00 0.0000 0 0 .0000 0.0000

17 0.00 0.00 0.0000 0 0 .0000 0.0000

18 0.00 0.00 0.0000 0 0 .0000 0.0000

19 0.00 0.00 0.0000 0 0 .0000 0.0000

If you’ve read the Foundations of Database Handicapping Help Doc then you know that win rates for the top ranked JPR horses in the above track profile snapshot are severely depressed. If early speed isn’t winning the races then something else must be.

How about late speed?

Here is that first two week Arlington track profile snapshot broken out by PAL (JCapper Pace Adjusted Late) –

Data Window Settings:

Divisor = 999

Dirt (All*) Distance: (All*)

From Index File: C:\2007\APX\pl_From_05-04-2007_to_05-18-2007.txt

Data Summary Win Place Show

Mutuel Totals 1300.00 1122.90 1134.80

Bet -1314.00 -1314.00 -1314.00

Gain -14.00 -191.10 -179.20

Wins 76 150 225

Plays 657 657 657

PCT .1157 .2283 .3425

ROI 0.9893 0.8546 0.8636

Avg Mut 17.11 7.49 5.04

By: PAL Rank

Rank Gain Bet Roi Wins Plays Pct Impact

1 41.80 156.00 1.2679 18 78 .2308 1.9949

2 -6.80 166.00 0.9590 9 83 .1084 0.9374

3 -36.40 152.00 0.7605 8 76 .1053 0.9100

4 -91.00 152.00 0.4013 8 76 .1053 0.9100

5 -23.40 140.00 0.8329 5 70 .0714 0.6175

6 -71.40 140.00 0.4900 5 70 .0714 0.6175

7 -31.40 134.00 0.7657 6 67 .0896 0.7742

8 17.20 116.00 1.1483 6 58 .1034 0.8943

9 48.60 80.00 1.6075 4 40 .1000 0.8645

10 68.80 50.00 2.3760 5 25 .2000 1.7289

11 -18.00 18.00 0.0000 0 9 .0000 0.0000

12 88.00 10.00 9.8000 2 5 .4000 3.4579

13 0.00 0.00 0.0000 0 0 .0000 0.0000

14 0.00 0.00 0.0000 0 0 .0000 0.0000

15 0.00 0.00 0.0000 0 0 .0000 0.0000

16 0.00 0.00 0.0000 0 0 .0000 0.0000

17 0.00 0.00 0.0000 0 0 .0000 0.0000

18 0.00 0.00 0.0000 0 0 .0000 0.0000

19 0.00 0.00 0.0000 0 0 .0000 0.0000

I don’t know about you but I find the above snapshots to be very interesting. If I were making UDMs or UPR for Arlington Park – because I see evidence of the APX PolyTrack surface playing like a category 4-5 surface - I would start looking for something other than early speed to base my UDMs and UPR upon. PAL seems like a very logical place to start.

How did the remainder of the Arlington meet play out? Here is how PAL, CPace and JPR performed in the time period after the first snapshot through the end of the meet:

Data Window Settings:

Divisor = 999

Dirt (All*) Distance: (All*)

From Index File: C:\2007\APX\pl_From_05-19-2007_to_09-30-2007.txt

Data Summary Win Place Show

Mutuel Totals 8147.00 7536.80 7308.50

Bet -9476.00 -9476.00 -9476.00

Gain -1329.00 -1939.20 -2167.50

Wins 589 1175 1750

Plays 4738 4738 4738

PCT .1243 .2480 .3694

ROI 0.8598 0.7954 0.7713

Avg Mut 13.83 6.41 4.18

By: PAL Rank

Rank Gain Bet Roi Wins Plays Pct Impact

1 180.60 1254.00 1.1440 134 627 .2137 1.7192

2 -329.60 1268.00 0.7401 100 634 .1577 1.2688

3 -403.60 1230.00 0.6719 74 615 .1203 0.9679

4 -178.20 1264.00 0.8590 85 632 .1345 1.0819

5 -247.20 1154.00 0.7858 57 577 .0988 0.7947

6 -85.40 1052.00 0.9188 59 526 .1122 0.9023

7 -203.00 814.00 0.7506 31 407 .0762 0.6127

8 -21.00 622.00 0.9662 23 311 .0740 0.5949

9 -128.20 406.00 0.6842 11 203 .0542 0.4359

10 216.60 216.00 2.0028 11 108 .1019 0.8193

11 -104.40 148.00 0.2946 3 74 .0405 0.3261

12 -25.60 48.00 0.4667 1 24 .0417 0.3352

13 0.00 0.00 0.0000 0 0 .0000 0.0000

14 0.00 0.00 0.0000 0 0 .0000 0.0000

15 0.00 0.00 0.0000 0 0 .0000 0.0000

16 0.00 0.00 0.0000 0 0 .0000 0.0000

17 0.00 0.00 0.0000 0 0 .0000 0.0000

18 0.00 0.00 0.0000 0 0 .0000 0.0000

19 0.00 0.00 0.0000 0 0 .0000 0.0000

By: CPace Rank

Rank Gain Bet Roi Wins Plays Pct Impact

1 24.80 1172.00 1.0212 109 586 .1860 1.4963

2 -183.60 1174.00 0.8436 95 587 .1618 1.3019

3 -122.40 1170.00 0.8954 83 585 .1419 1.1413

4 149.00 1176.00 1.1267 84 588 .1429 1.1492

5 -354.20 1166.00 0.6962 68 583 .1166 0.9383

6 -310.40 1108.00 0.7199 60 554 .1083 0.8712

7 -368.00 916.00 0.5983 38 458 .0830 0.6674

8 -241.80 652.00 0.6291 19 326 .0583 0.4688

9 104.60 432.00 1.2421 14 216 .0648 0.5214

10 -8.80 272.00 0.9676 12 136 .0882 0.7098

11 5.60 162.00 1.0346 5 81 .0617 0.4966

12+ -23.80 76.00 0.6868 2 38 .0526 0.4234

By: JPR Rank

Rank Gain Bet Roi Wins Plays Pct Impact

1 -301.40 1172.00 0.7428 136 586 .2321 1.8669

2 -47.80 1172.00 0.9592 119 586 .2031 1.6335

3 -61.40 1172.00 0.9476 97 586 .1655 1.3315

4 -144.20 1172.00 0.8770 73 586 .1246 1.0021

5 -145.40 1166.00 0.8753 56 583 .0961 0.7727

6 114.00 1104.00 1.1033 50 552 .0906 0.7286

7 -401.00 910.00 0.5593 22 455 .0484 0.3889

8 -63.20 660.00 0.9042 22 330 .0667 0.5363

9 -366.00 436.00 0.1606 3 218 .0138 0.1107

10 140.00 272.00 1.5147 6 136 .0441 0.3549

11 -103.60 164.00 0.3683 2 82 .0244 0.1962

12 51.00 76.00 1.6711 3 38 .0789 0.6351

13 0.00 0.00 0.0000 0 0 .0000 0.0000

14 0.00 0.00 0.0000 0 0 .0000 0.0000

15 0.00 0.00 0.0000 0 0 .0000 0.0000

16 0.00 0.00 0.0000 0 0 .0000 0.0000

17 0.00 0.00 0.0000 0 0 .0000 0.0000

18 0.00 0.00 0.0000 0 0 .0000 0.0000

19 0.00 0.00 0.0000 0 0 .0000 0.0000

While I never advocate blindly basing your play on a single factor – layered tight model UDMs are what I use in my own live play - It should be clear at this point that paying attention to track surface – and the evidence was clearly there as early as two weeks into the meet – you would have had a much easier time finding success at the windows by focusing on late speed in your UDMs and UPR at Arlington Park in the summer of 2007 instead of JPR and early speed.

Interestingly enough, the data suggests that it would have been possible to create a track specific UDM for the 2007 Arlington meet using CPace rank one as a primary factor. Sometimes the crowd realizes and overreacts to the influence track surface has on the outcome of races.

Belmont Park 2007

Let’s do the summer at Belmont 2007. I started by using the Quick Index File Extract tool to create a Date Ranged Index for the first two weeks of their 2007 meet. I then ran my Track Profiler UDM through the Data Window using the date ranged index file – which gives me a snapshot of what happened during the first two weeks of the meet. Here’s what the data looks like:

UDM Definition: TRACKPROFILER

Divisor: # UDM Def Divisor: 999

Surface Req: *ANY Surface*

Distance Req: *ANY Distance*

CPace: MinRank= 1 MaxRank= 3

MinVal= -999 MaxVal= 999

MinGap= -999 MaxGap= 999

Late Pace (last3) MinRank= 2 MaxRank= 5

MinVal= -999 MaxVal= 999

MinGap= -18 MaxGap= 999

Pace Index: MinVal= 80 MaxVal= 999

Race Volatility: MinVal= 75 MaxVal= 119.99

Running Style: ALL

Data Window Settings:

Divisor = 999

Filters Applied:

Surface: (ALL*) Distance: (All*)

From Index File: C:\2007\BEL\pl_From_05-01-2007_to_05-14-2007.txt

Data Summary Win Place Show

Mutuel Totals 40.00 24.70 26.40

Bet -40.00 -40.00 -40.00

Gain 0.00 -15.30 -13.60

Wins 4 5 7

Plays 20 20 20

PCT .2000 .2500 .3500

Now, compare the above results for Belmont to the benchmark results for all dirt surfaces in 2006. What do you see? Based on what you see, how would YOU categorize the dirt surface at BEL? Forget the roi for a second. Pay attention to the win percentages. My own interpretation is that this snapshot makes the BEL surface look an awful lot like all of those category 1-2-3 surfaces I saw when I created the original dirt algorithms for JRating and JPR: Favors early speed strongly and Favors early speed somewhat and Plays fair.

How can I say this? Well the win rate for horses selected by my Track Profiler UDM (early speed) in the snapshot is right in line with my Benchmark. Some group of horses has to be hitting the wire first in every data sample. Again, it’s kind of like a see-saw: If early speed horses are winning the races – if their win rates are in line with the benchmark or even elevated slightly – then some other group of horses has to be suffering: In this case closers are the ones suffering – or rather – bettors backing closers are probably suffering most of all.

Remember that when I created the dirt surface algorithm for JPR I based it on dirt surfaces behaving as categories 1-2-3. Here’s that first two week BEL snapshot again broken out by JPR Rank:

Data Window Settings:

Divisor = 999

Dirt (All*) Distance: (All*)

From Index File: C:\2007\BEL\pl_From_05-01-2007_to_05-14-2007.txt

Data Summary Win Place Show

Mutuel Totals 616.50 529.90 565.30

Bet -668.00 -668.00 -668.00

Gain -51.50 -138.10 -102.70

Wins 47 94 135

Plays 334 334 334

PCT .1407 .2814 .4042

ROI 0.9229 0.7933 0.8463

Avg Mut 13.12 5.64 4.19

By: JPR Rank

Rank Gain Bet Roi Wins Plays Pct Impact

1 -5.10 94.00 0.9457 18 47 .3830 2.7216

2 -1.40 94.00 0.9851 10 47 .2128 1.5120

3 -40.10 94.00 0.5734 4 47 .0851 0.6048

4 -44.80 94.00 0.5234 5 47 .1064 0.7560

5 29.80 90.00 1.3311 5 45 .1111 0.7896

6 -22.40 72.00 0.6889 3 36 .0833 0.5922

7 -29.00 56.00 0.4821 1 28 .0357 0.2538

8 99.50 36.00 3.7639 1 18 .0556 0.3948

9 -22.00 22.00 0.0000 0 11 .0000 0.0000

10 -10.00 10.00 0.0000 0 5 .0000 0.0000

11 -4.00 4.00 0.0000 0 2 .0000 0.0000

12 -2.00 2.00 0.0000 0 1 .0000 0.0000

13 0.00 0.00 0.0000 0 0 .0000 0.0000

14 0.00 0.00 0.0000 0 0 .0000 0.0000

15 0.00 0.00 0.0000 0 0 .0000 0.0000

16 0.00 0.00 0.0000 0 0 .0000 0.0000

17 0.00 0.00 0.0000 0 0 .0000 0.0000

18 0.00 0.00 0.0000 0 0 .0000 0.0000

19 0.00 0.00 0.0000 0 0 .0000 0.0000

If you’ve read the Foundations of Database Handicapping Help Doc then you know that win rates for the top ranked JPR horses in the above track profile snapshot are a little above what you’d normally expect. Early speed seems to be doing quite well in the above snapshot.

How about late speed?

Here is that first two week Belmont track profile snapshot broken out by PAL (JCapper Pace Adjusted Late) –

Data Window Settings:

Divisor = 999

Dirt (All*) Distance: (All*)

From Index File: C:\2007\BEL\pl_From_05-01-2007_to_05-14-2007.txt

Data Summary Win Place Show

Mutuel Totals 616.50 529.90 565.30

Bet -668.00 -668.00 -668.00

Gain -51.50 -138.10 -102.70

Wins 47 94 135

Plays 334 334 334

PCT .1407 .2814 .4042

ROI 0.9229 0.7933 0.8463

Avg Mut 13.12 5.64 4.19

By: PAL Rank

Rank Gain Bet Roi Wins Plays Pct Impact

1 -41.20 94.00 0.5617 9 47 .1915 1.3608

2 -60.30 94.00 0.3585 7 47 .1489 1.0584

3 -59.30 100.00 0.4070 5 50 .1000 0.7106

4 23.50 90.00 1.2611 10 45 .2222 1.5792

5 -7.60 94.00 0.9191 7 47 .1489 1.0584

6 22.20 72.00 1.3083 4 36 .1111 0.7896

7 -58.00 58.00 0.0000 0 29 .0000 0.0000

8 6.70 34.00 1.1971 3 17 .1765 1.2541

9 130.50 24.00 6.4375 2 12 .1667 1.1844

10 -6.00 6.00 0.0000 0 3 .0000 0.0000

11 -2.00 2.00 0.0000 0 1 .0000 0.0000

12 0.00 0.00 0.0000 0 0 .0000 0.0000

13 0.00 0.00 0.0000 0 0 .0000 0.0000

14 0.00 0.00 0.0000 0 0 .0000 0.0000

15 0.00 0.00 0.0000 0 0 .0000 0.0000

16 0.00 0.00 0.0000 0 0 .0000 0.0000

17 0.00 0.00 0.0000 0 0 .0000 0.0000

18 0.00 0.00 0.0000 0 0 .0000 0.0000

19 0.00 0.00 0.0000 0 0 .0000 0.0000

I don’t know about you but I find the above snapshots to be very interesting. If I were making UDMs or UPR for Belmont – because I see evidence of the BEL dirt surface playing like a category 1-2 surface - I would be looking to early speed to base my UDMs and UPR upon instead of late speed. JPR and CPace both seem like very logical places to start.

How did the remainder of the Belmont meet play out? Here is how JPR, CPace, and PAL performed in the time period after the first two week snapshot through the end of the September, 2007:

Data Window Settings:

Divisor = 999

Dirt (All*) Distance: (All*)

From Index File: C:\2007\BEL\pl_From_05-15-2007_to_09-30-2007.txt)

Data Summary Win Place Show

Mutuel Totals 3392.30 3378.50 3465.60

Bet -4546.00 -4546.00 -4546.00

Gain -1153.70 -1167.50 -1080.40

Wins 321 638 919

Plays 2273 2273 2273

PCT .1412 .2807 .4043

ROI 0.7462 0.7432 0.7623

Avg Mut 10.57 5.30 3.77

By: JPR Rank

Rank Gain Bet Roi Wins Plays Pct Impact

1 41.30 630.00 1.0656 113 315 .3587 2.5402

2 -150.20 632.00 0.7623 60 316 .1899 1.3445

3 -6.70 632.00 0.9894 54 316 .1709 1.2100

4 -161.40 632.00 0.7446 37 316 .1171 0.8291

5 -425.00 612.00 0.3056 15 306 .0490 0.3471

6 -272.50 540.00 0.4954 17 270 .0630 0.4458

7 -47.40 394.00 0.8797 13 197 .0660 0.4673

8 -51.40 258.00 0.8008 7 129 .0543 0.3842

9 -46.20 144.00 0.6792 4 72 .0556 0.3934

10 -54.00 54.00 0.0000 0 27 .0000 0.0000

11 23.80 14.00 2.7000 1 7 .1429 1.0116

12+ -4.00 4.00 0.0000 0 2 .0000 0.0000

By: CPace Rank

Rank Gain Bet Roi Wins Plays Pct Impact

1 -93.50 636.00 0.8530 80 318 .2516 1.7814

2 -84.80 630.00 0.8654 55 315 .1746 1.2364

3 -139.70 634.00 0.7797 53 317 .1672 1.1839

4 -77.40 630.00 0.8771 50 315 .1587 1.1240

5 -366.30 610.00 0.3995 25 305 .0820 0.5804

6 -189.70 540.00 0.6487 25 270 .0926 0.6556

7 -108.40 404.00 0.7317 17 202 .0842 0.5959

8 -25.90 252.00 0.8972 10 126 .0794 0.5620

9 -35.80 140.00 0.7443 5 70 .0714 0.5058

10 -16.20 54.00 0.7000 1 27 .0370 0.2623

11 -12.00 12.00 0.0000 0 6 .0000 0.0000

12+ -4.00 4.00 0.0000 0 2 .0000 0.0000

By: PAL Rank

Rank Gain Bet Roi Wins Plays Pct Impact

1 -265.30 690.00 0.6155 61 345 .1768 1.2520

2 -159.50 720.00 0.7785 71 360 .1972 1.3965

3 -212.70 692.00 0.6926 52 346 .1503 1.0642

4 -161.40 636.00 0.7462 43 318 .1352 0.9575

5 -26.70 598.00 0.9554 42 299 .1405 0.9947

6 -217.40 496.00 0.5617 20 248 .0806 0.5710

7 51.30 362.00 1.1417 23 181 .1271 0.8998

8 -148.00 200.00 0.2600 3 100 .0300 0.2124

9 13.60 112.00 1.1214 5 56 .0893 0.6322

10 -28.00 28.00 0.0000 0 14 .0000 0.0000

11 2.40 10.00 1.2400 1 5 .2000 1.4162

12+ -2.00 2.00 0.0000 0 1 .0000 0.0000

Again, I almost never advocate basing your play blindly on any one factor. But it should be clear at this point that paying attention to track surface – and the evidence was there as early as two weeks into the meet – you would have done well to focus on JPR and early speed in your UDMs and UPR at Belmont Park in the summer of 2007 while downgrading the importance of late speed.

Summary

All racing surfaces have their own individual characteristics and tendencies. It pays to understand them. Using a Track Profiler UDM to establish a benchmark and comparing specific time period track profile snapshots to a benchmark can be an excellent way to get an understanding of what those individual track surface tendencies actually are.

-jp