Top Horse Analytics: 09/19/2016 Weekly Statistics

[Augenj]

Welcome to a public subset of the weekly statistics produced by Top Horse Analytics.
Column descriptions follow:

Note: "place" bet types have been created where possible

BET TYPE
Favorite Win - Bet the favorite to win
Favorite Place - Bet the favorite to place
MLF Win - Bet the morning line favorite to win
MLF Place - Bet the morning line favorite to place
Fav/MLF Win - Bet the favorite to win when it's also the morning line favorite
Fav/MLF Place - Bet the favorite to place when it's also the morning line favorite
TH/Fav Win - Bet the THA top horse to win when it's also the favorite
TH/Fav Place - Bet the THA top horse to place when it's also the favorite
TH/MLF Win - Bet the THA top horse to win when it's also the morning line favorite
TH/MLF Place - Bet the THA top horse to place when it's also the morning line favorite
TH/Fav/MLF Win - Bet the THA top horse to win when it's also the favorite and morning line favorite
TH/Fav/MLF Plc - Bet the THA top horse to place when it's also the favorite and morning line favorite
TH Win - Bet the THA top horse to win
TH Place - Bet the THA top horse to place
TH Show - Bet the THA top horse to show
TH Exacta - Bet a straight exacta with THA top horse and 2nd horse
TH Trifecta - Bet a straight trifecta with THA top horse, 2nd horse, 3rd horse
TH Super - Bet a straight superfecta with THA top horse, 2nd horse, 3rd horse, 4th horse
TH Double - Bet a straight daily double with THA top horse in two consecutive races
TH Pick 3 - Bet a straight pick 3 with THA top horse in three consecutive races
TH Pick 4 - Bet a straight pick 4 with THA top horse in four consecutive races
TH Pick 5, 4/5 - Bet a straight pick 5 with THA horse in five or four of five consecutive races
TH Pick 6, 5/6 - Bet a straight pick 6 with THA horse in six or five of six consecutive races
TH 2nd Win - Bet the THA 2nd horse in a race to win
TH 2nd Place - Bet the THA 2nd horse in a race to place
TH 3rd Win - Bet the THA 3rd horse in a race to win
TH 3rd Place - Bet the THA 3rd horse in a race to place
TH 4th Win - Bet the THA 4th horse in a race to win
TH 4th Place - Bet the THA 4th horse in a race to place
TH 5th Win - Bet the THA 5th horse in a race to win
TH 5th Place - Bet the THA 5th horse in a race to place
TH 6th Win - Bet the THA 6th horse in a race to win
TH 6th Place - Bet the THA 6th horse in a race to place
Any J Win - Bet any horse to win that has the Best Jockey flag (J)
Any J Place - Bet any horse to place that has the Best Jockey flag (J)
Any T Win - Bet any horse to win that has the Best Trainer flag (T)
Any T Place - Bet any horse to place that has the Best Trainer flag (T)
Any S Win - Bet any horse to win that has the Best Speed flag (S)
Any S Place - Bet any horse to place that has the Best Speed flag (S)
Any F Win - Bet any horse to win that has the Best Finish flag (F)
Any F Place - Bet any horse to place that has the Best Finish flag (F)
Any E Win - Bet any horse to win that has the Best Earnings flag (E)
Any E Place - Bet any horse to place that has the Best Earnings flag (E)
Any W Win - Bet any horse to win that has the Best Workouts flag (W)
Any W Place - Bet any horse to place that has the Best Workouts flag (W)
Any L Win - Bet any horse to win that has the Best Lead Pack flag (L)
Any L Place - Bet any horse to place that has the Best Lead Pack flag (L)
Any C Win - Bet any horse to win that has the Best Closer flag (C)
Any C Place - Bet any horse to place that has the Best Closer flag (C)
Worst Odds Win - Bet the horse with the highest odds in a race to win
Worst Odds Plc - Bet the horse with the highest odds in a race to place

BET SIZE
Lowest bet size for a bet type.

OCC PC
Occurs percent is calculated by dividing the NBR RACES column by the NUMBER OF RACES at the top. (and multiplying by 100)

NBR RACES
The number of races for this bet type.

NBR HITS
The number of hits for this bet type.

HIT PC
Hit percent is calculated by dividing the NBR HITS column by the NBR RACES column. (and multiplying by 100)

DOL$ IN
Dollars In is calculated by multiplying the BET SIZE column times the NBR RACES column.

$DOL OUT
Dollars Out is calculated by summing the individual payoffs for this bet type during the year.

HI PAYOFF
High payoff is the highest payoff for this bet type during the year.

PROFIT
Profit is the $DOL OUT column minus the $DOL IN column for this bet type.

ROI
Return On Investment is calculated by dividing $DOL OUT by $DOL IN.

* * *

The following graphs simulate wagers from the start of the current logical year to the present.

Yellow cells indicate 0.9001 ROI to 1.0000 ROI. (close but no cigar)
Green cells indicate 1.0001 ROI and above. (profitable)

Questions and suggestions are welcome. :)

[Augenj]

Belmont Park 1/2

[Augenj]

Belmont Park 2/2

[Augenj]

Churchill Downs 1/2

[Augenj]

Churchill Downs 2/2

[Augenj]

Gulfstream Park 1/2

[Augenj]

Gulfstream Park 2/2

[VigorsTheGrey]

Augenj,

The data on the sheets 2/2 don't seem correct...just so i'm clear about the place percentages....the percentage indicates any aggregate condition meaning the horse could have finished 1st or 2nd and be counted as finishing 2nd, right...one would just subtract the winning percentage from the place percentage to arrive at the percentage that the horse finished 2nd only...that is why the data in the 2/2 sheets doesn't seen right...

[Augenj]

Augenj,

The data on the sheets 2/2 don't seem correct...just so i'm clear about the place percentages....the percentage indicates any aggregate condition meaning the horse could have finished 1st or 2nd and be counted as finishing 2nd, right...one would just subtract the winning percentage from the place percentage to arrive at the percentage that the horse finished 2nd only...that is why the data in the 2/2 sheets doesn't seen right...
You're right. If the horse has a place payoff, whether a winner or not, it's summed up and averaged. Maybe it's just me but I don't see a need for statistics on horses that finish 2nd only. If you bet a horse to place, you get paid whether it finishes 1st or 2nd as you know.

[Augenj]

Those numbers on 2/2 do look odd in the sense that the wins and places are equal or nearly equal so I ran the program interactively with a debug trap looking for when there was a place payoff but no win payoff. This condition happened rarely as the stats showed earlier. It's just an anomaly with the "Any X" horses I guess although I'm not comfortable with that. I'll keep looking for logic errors although it's the same logic as all of the other "normal" place bets.

[VigorsTheGrey]

It must be that there is a double counting effect going on....it is significant to note that the "worst odds" win and place figures do not show the anomaly and spread hit percentage-wise in a reasonable fashion....I think it must double, triple, even quadruple count and average because more than (1)"Type X" bold factor routinely appear on one horse......?

...If the data were correct, would it mean that if the "Type X" horse did not win, that it did not finish 2nd either...that it almost never finished second? This is not the case, intuitively...

[Augenj]

It must be that there is a double counting effect going on....it is significant to note that the "worst odds" win and place figures do not show the anomaly and spread hit percentage-wise in a reasonable fashion....I think it must double, triple, even quadruple count and average because more than (1)"Type X" bold factor routinely appear on one horse......?

...If the data were correct, would it mean that if the "Type X" horse did not win, that it did not finish 2nd either...that it almost never finished second? This is not the case, intuitively...
Agreed. It might be upstream where the Any X conditions are selected. The place logic is the same everywhere so it can't be that since I tested it also.

[VigorsTheGrey]

The smoking gun is the "Any C" win and place, which in the 2/2 sheet for Gulfstream shows an identical 22.13 for both hit percentage rate...could it be that the program is selecting the same line item for both win and place?

I know nothing about code writing so don't feel like you have to explain anything to me...thanks for your effort here.... :)

[Augenj]

The smoking gun is the "Any C" win and place, which in the 2/2 sheet for Gulfstream shows an identical 22.13 for both hit percentage rate...could it be that the program is selecting the same line item for both win and place?

I know nothing about code writing so don't feel like you have to explain anything to me...thanks for your effort here.... :)I'll look at this tonight when everybody is sleeping and I can hear myself think. :D

[Augenj]

Dummy here forgot that there is more than one place payoff in the results arrays. I had cloned and modified the win logic to create the place logic without taking this into account. I'll also have to modify the win logic to accommodate dead heats when there are multiple win payoffs along with reviewing other place logic besides the "Any X" stats.

It will take a day or so to correct all of this and re-post the corrected weekly statistics.

Thanks to VigorsTheGrey and ReplayRandall for their suggestions and error reporting. ;)

[Augenj]

The incorrect results affected only the "Any X" lines. All other lines checked out OK. The "Any X" conditions were difficult to code to begin with since the whole system was designed around using the Top Horse and conditions that are coupled with it.

The "Any X" wins were affected since sometimes there are multiple payoffs from dead heats and entries. As we all know (now), there are multiple place payoffs in the results race arrays and this change fixes that also.

Thanks for pointing out the errors, VigorsTheGrey.

Following are the corrected 2/2 graphs for each track.

[Augenj]

Belmont Park 2/2

[Augenj]

Churchill Downs 2/2

[Augenj]

Gulfstream Park 2/2