PCT e

[mikesal57]

Lets get back to real handicapping.....

I'm sure most have heard of PctE when dealing with "pace" programs.

In the 1st screen....its obvious where the winner will(should) come from..

But, in the 2nd screen , where do you think the winner is coming from..??

A friend of mine takes ROI over win percentage , I believe its the opposite..

Your thoughts?

Mike

     query start:         7/8/2016 9:07:50 AM
     query end:           7/8/2016 9:07:51 AM
     elapsed time:        1 seconds

     Data Window Settings:
     Connected to: C:\JCapper\exe\JCapper2.mdb
     999 Divisor  Odds Cap: None

     SQL:  SELECT * FROM STARTERHISTORY
           WHERE TRACK = 'BEL' 
           AND [DATE] >= #01-01-2016# 
           AND INTSURFACE <= 3 
           AND DIST < 1760


     Data Summary          Win         Place          Show
     -----------------------------------------------------
     Mutuel Totals     1869.70       1895.70       1761.60
     Bet              -2184.00      -2184.00      -2184.00
     -----------------------------------------------------
     P/L               -314.30       -288.30       -422.40

     Wins                  151           304           436
     Plays                1092          1092          1092
     PCT                 .1383         .2784         .3993

     ROI                0.8561        0.8680        0.8066
     Avg Mut             12.38          6.24          4.04


     By: SQL-F34 Rank

     Rank       P/L        Bet        Roi    Wins   Plays     Pct     Impact
     -----------------------------------------------------------------------
      1     -145.60     332.00     0.5614      10     166   .0602     0.4356  
      2      -84.20     308.00     0.7266      20     154   .1299     0.9392  
      3       36.40     332.00     1.1096      30     166   .1807     1.3069  
      4      -71.10     292.00     0.7565      25     146   .1712     1.2383  
      5       22.20     298.00     1.0745      28     149   .1879     1.3590  
      6      -63.80     236.00     0.7297      15     118   .1271     0.9193  
      7       11.90     172.00     1.0692      11      86   .1279     0.9250  
      8       48.20     108.00     1.4463       9      54   .1667     1.2053  
      9      -34.40      52.00     0.3385       1      26   .0385     0.2781  
     10      -13.90      34.00     0.5912       2      17   .1176     0.8508  
     11      -12.00      12.00     0.0000       0       6   .0000     0.0000  
     12       -6.00       6.00     0.0000       0       3   .0000     0.0000  
     13       -2.00       2.00     0.0000       0       1   .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

     query start:         7/8/2016 9:14:15 AM
     query end:           7/8/2016 9:14:15 AM
     elapsed time:        0 seconds

     Data Window Settings:
     Connected to: C:\JCapper\exe\JCapper2.mdb
     999 Divisor  Odds Cap: None

     SQL:  SELECT * FROM STARTERHISTORY
           WHERE TRACK = 'BEL' 
           AND [DATE] >= #01-01-2016# 
           AND INTSURFACE <= 3 
           AND DIST >= 1760


     Data Summary          Win         Place          Show
     -----------------------------------------------------
     Mutuel Totals      928.50        867.20        864.00
     Bet              -1154.00      -1154.00      -1154.00
     -----------------------------------------------------
     P/L               -225.50       -286.80       -290.00

     Wins                   84           168           253
     Plays                 577           577           577
     PCT                 .1456         .2912         .4385

     ROI                0.8046        0.7515        0.7487
     Avg Mut             11.05          5.16          3.42


     By: SQL-F34 Rank

     Rank       P/L        Bet        Roi    Wins   Plays     Pct     Impact
     -----------------------------------------------------------------------
      1     -116.20     168.00     0.3083       8      84   .0952     0.6542  
      2       30.80     168.00     1.1833      11      84   .1310     0.8995  
      3      -49.10     168.00     0.7077      13      84   .1548     1.0631  
      4      -61.00     168.00     0.6369      13      84   .1548     1.0631  
      5      -17.20     168.00     0.8976      16      84   .1905     1.3084  
      6      -57.50     146.00     0.6062      10      73   .1370     0.9410  
      7       -2.70      94.00     0.9713       7      47   .1489     1.0230  
      8       68.60      34.00     3.0176       4      17   .2353     1.6162  
      9      -11.00      24.00     0.5417       1      12   .0833     0.5724  
     10       -4.20      10.00     0.5800       1       5   .2000     1.3738  
     11       -2.00       2.00     0.0000       0       1   .0000     0.0000  
     12       -2.00       2.00     0.0000       0       1   .0000     0.0000  
     13       -2.00       2.00     0.0000       0       1   .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

[mikesal57]

I guess all the cappers are at work...

[traynor]

The results of ANY study of pace are strongly influenced by the criteria of the racelines selected for analysis. Tweak the process of selecting input (even slightly) and the results are tweaked.

Most who are (or want to be) pace handicappers realize (or should realize) this fundamental aspect of race analysis, and explicitly declare the process used to arrive at the results generated.

PctE can be usefully predictive. It can also be totally misleading, contingent on the specifics of how the analysis is set up. Specifically (as just about every pace handicapper has discovered to his or her dismay), the lines used to generate the results (on a specific group of races) may generate quite different results when applied to a different group of races, or when slight tweaks are made in input content.

[raybo]

%E is a measurement of how a horse "might" expend its energy (early, late or neutral). Which leads many to the assumption that, in shorter sprint races, horses that tend to expend more energy early are advantaged. And, in the long term, that is probably true, when looking at hit rates. However, much depends on the individual field dynamics (running styles, pace pressure, and early speed capabilities of the entire field).

The theory of early energy expenditure being important, in shorter sprint races, is well known, and as a result, well bet. Therefore, the better early expenditure horses will often be bet down, to the point that a normal hit rate (an expected hit rate under the specific circumstances) will probably not overcome the low averageodds associated with these types of horses.

IMO, for win wagering and probably horizontal wagering, if %E is going to be your selection priority, regarding the shorter sprint races, and your priority is going to be hit rate, then that hit rate had better be unusually high, to the point of being unattainable, or at least not sustainable over time. Therefore, ROI becomes more important, in my mind anyway.

[mikesal57]

Ray...thx for answering

Tray...I need someone from MIT to translate for me

Mike

[traynor]

Originally Posted by mikesal57

Ray...thx for answering

Tray...I need someone from MIT to translate for me

Mike

How is the value %E derived? What--exactly--goes into the calculation? However it is defined, it is virtually guaranteed that another "pace analyst" will use a different set of criteria to "calculate" a rating he or she then defines as "%E." Are only recent races used? Only one race? Two races? Best two of last howevermany? Only this distance? Only this track? Only races in which the entry in question was a serious contender? The list is endless.

Profiling pace (and profiling "energy distribution") seems easy if one simply crunches numbers and declares whatever comes out as whatever one chooses. The notion of %E is intimately connected the notion of "track profiles" and both are subject to (highly subjective, individual, and wildly differing) "interpretations."

One person's %E output will only be in the same ballpark as another person's %E output if and only if the exact same inputs and calculations are made. That is rarely the case.

Consider--would you calculate the (expected, and presumed "predictive") %E of an entry that had not been in contention at any point in any of its last three, four, or five races? Or of an entry that "normally" is fast early, then quits in the stretch, but only wins when it trundles around midpack and closes?

Simplistic (and seriously misleading) "pace calculations" are far more impressive in the viewing stage than in the betting stage. Pace analysis can be highly predictive, but is way more complex than parsing a few values from recent races, adding, subtracting, averaging, and otherwise massaging those values, then declaring them to be a "pace analysis."

Fortunately, few are willing to analyze pace (and probable pace) on anything more than a superficial, simplistic level. That leaves a LOT of opportunity for those willing to do more sophisticated pace analysis.

[BCOURTNEY]

Originally Posted by mikesal57

Ray...thx for answering

Tray...I need someone from MIT to translate for me

Mike

Tray means that method by which you identify and select input variables for whatever pace approach you choose should be consistent because slight variations in data and data selection will aggravate the results and make analysis difficult or impossible, especially when trying to peel a few pennies off every dollar. Your approach should be well defined at the outset, so you can truly test if it is achieving the desired result.

I just noticed Tray was responding as I submitted, wasn't trying to provide a response different than his.

[mikesal57]

Lets take it back a step....

I have a program ,as you can see, is JCAPPER...

I don't have a clue on how Jeff programmed his factor PctE
I'm sure after 10+ years he has a hold on how he did it & how it works...

So , back to my post , with all things being as it is

look at screen 2...

My contenders are 2 & 5

Who would you bet and why?

Mike

[JustRalph]

From the jcapper factors page


"PctE - or Percent Early. This number represents the percent of total energy expended by a horse during the first two thirds of a race. This number is an average taken from each of the running lines found in a horse's past performance record. The PctE number is calculated using the Bris E1, E2, Late Pace, and Final Time based Speed Figure. Each number is weighted according to the impact of historical occurances and each of the separate factors is then combined into a single number. The higher the PctE, the more likely it is that a horse will contend for the lead in today's race. The lower the PctE, the more likely it is that the horse will not contend for the lead in today's race. Hint~ PctE can often be a very useful identifier of track specific speed and closing biases when they are present. (Accessible via the Profile Table Interface)"

[cj]

The original %E is from Sartin stuff. I think it worked like this:

2nd Call Feet Per Second / (2nd Call Feet Per Second + Last Fraction Feet Per Second)

So an example from a 6f race:

22 45 1:10

2640 / 45 = 58.67

1320 / (70 - 45) = 52.80

(These two added together were called Total Energy)

%E = 58.67 / (58.67 + 52.80) = 52.63%

Where it has evolved over the years I don't know, but that was the original as best I remember.

[traynor]

Originally Posted by JustRalph

From the jcapper factors page


"PctE - or Percent Early. This number represents the percent of total energy expended by a horse during the first two thirds of a race. This number is an average taken from each of the running lines found in a horse's past performance record. The PctE number is calculated using the Bris E1, E2, Late Pace, and Final Time based Speed Figure. Each number is weighted according to the impact of historical occurances and each of the separate factors is then combined into a single number. The higher the PctE, the more likely it is that a horse will contend for the lead in today's race. The lower the PctE, the more likely it is that the horse will not contend for the lead in today's race. Hint~ PctE can often be a very useful identifier of track specific speed and closing biases when they are present. (Accessible via the Profile Table Interface)"

I think most (if not all) professional-level pace analysis software goes way beyond the averaging of all races concept. As with nearly every other speed/pace rating generated with computer analysis, such averaging suffers from the conceptual flaw of regarding every horse in every race as having exerted its optimal effort to win that race. In any race in the PPs in which this is NOT the case (and in numerous other instances), inclusion of that clump of data in the overall analysis is flawed. The result is more semantic noise than prediction (or even accurate description).

This is in no way a criticism of jcapper (or anyone else's) definition of, or use of, %E as a factor in pace analysis. I only mean to point out that the term is used for a wide spectrum of output that is comprised of an equally wide spectrum of input.

[cj]

Originally Posted by traynor

I think most (if not all) professional-level pace analysis software goes way beyond the averaging of all races concept. As with nearly every other speed/pace rating generated with computer analysis, such averaging suffers from the conceptual flaw of regarding every horse in every race as having exerted its optimal effort to win that race. In any race in the PPs in which this is NOT the case (and in numerous other instances), inclusion of that clump of data in the overall analysis is flawed. The result is more semantic noise than prediction (or even accurate description).

This is in no way a criticism of jcapper (or anyone else's) definition of, or use of, %E as a factor in pace analysis. I only mean to point out that the term is used for a wide spectrum of output that is comprised of an equally wide spectrum of input.

Awesome. Now lets try to respect the premise of the thread and get back on topic.

[mikesal57]

I guess I have to settle on the college course I got from all here..

But no one answered my question

I give up!!

[cj]

Originally Posted by mikesal57

I guess I have to settle on the college course I got from all here..

But no one answered my question

I give up!!

Hey, I just tried, give it time! (and mine was only algebra )

That said, since most of us don't use Jcapper, explain. Why is the first screen obvious where the winner should come from as a starter?

[whodoyoulike]

Originally Posted by traynor

I think most (if not all) professional-level pace analysis software goes way beyond the averaging of all races concept. As with nearly every other speed/pace rating generated with computer analysis, such averaging suffers from the conceptual flaw of regarding every horse in every race as having exerted its optimal effort to win that race. In any race in the PPs in which this is NOT the case (and in numerous other instances), inclusion of that clump of data in the overall analysis is flawed. The result is more semantic noise than prediction (or even accurate description).

This is in no way a criticism of jcapper (or anyone else's) definition of, or use of, %E as a factor in pace analysis. I only mean to point out that the term is used for a wide spectrum of output that is comprised of an equally wide spectrum of input.

This is the first I've heard of PctE but, this isn't describing pace analysis but rather energy expended which is an indication of stamina or energy remaining. It seems the value is noting the value over several records for each horse. If you're able to estimate the energy expended for each horse at each of the four fractional calls for the upcoming race then I can see it as a part of pace analysis.