Photo Finish LIVE — The Charts
The highest levels of competition will look for that critical .01% edge and the best stables will know what data to analyze in order to find it.
The Commish has spent the last 15 years of his professional life synthesizing complex data structures and locating the needles of value in the haystacks of information. Prior to that, he spent 15 years of his personal life playing video games (primarily Zelda, Goldeneye, FIFA, Madden and Socom franchises) and relished in the competitiveness of gaming, which is really just the digital manifestation of an intellectual duel. A chess match, variation.
None of that means anything… other than DYOR.
Third Time has been gracious enough to provide us with some data represented in colorful visual illustrations. The data referenced in this article only pertains to the individual horses and can be located on the horse profile page:
These charts are all lagging indicators, meaning they are the results of performance, not the causes of performance.
Going to assume all races are run on preference. For those who run off preference often, you can just add ‘No Preference Considerations’ to the Cons for each chart.
Very important to understand the information available, because if you don’t, that means you should assume your competition knows more than you.
Let’s dive in…
Race Results by Finish Position: Bar Chart
Pros:
-Provides visual highlight of finish position results
Cons:
-No field size context or considerations
— Race in 4 horse fields consistently and ‘over’ podium, skewing results
Alpha:
-None… Just a basic highlight of results
How to improve this chart:
-Change the colors of the finishing positions to match the other charts
— Yellow for 2nd, Orange for 3rd, Clay-ish for Off Podium
-Have only 1 scale or have 2 scales for all horses, not relative scales
— The scale should be static for every horse, regardless of race quantity
— Influenced by race count, a full bar could represent different amounts
— — This is true for both the Y and X axis dynamics
-Net Profit / ROI by place would be a better indicator of efficiency in results
— 3rd place in small fields loses or wins little
— 3rd place in large fields wins much more than breakeven
Opinion:
-Don’t love this chart, missing context, no alpha, minimal value
Race Results by Distance: Stacked Bar Chart
Pros:
-Provides visual highlight of finish position results by furlong
-Color scales provide aid to decipher clues where the best results are
-Stacked counts provide condensed logic on finish positions
Cons:
-No field size context or considerations
— Race in 4 horse fields consistently and ‘over’ podium, skewing results
Alpha:
-Like the children’s game of “Hot and Cold”, this chart can easily give you clues to truncate the time to find your horse’s distance preference (DP)
— This is the discovery phase for racing in PFL, finding your DP
-The faster you uncover the DP, the less you spend on discovery
— ROI positive skill in your toolbox
-Potentially visualizing variance (presumably Temper)
— High race counts with ample Wins and Missed Podiums highlight this
How to improve this chart:
-Have only 1 scale or have 2 scales for all horses, not relative scales
— The scale should be static for every horse, regardless of race quantity
— Influenced by race count, stacked bar could represent different amounts
— — This is only true for the Y axis
Opinion:
-Big fan, especially for wins, green is a clear indicator of potential success
-Clear feedback of where to keep smashing those and adjacent furlongs
A Chart Has No Name: Scatter Plot
Pros:
-Provides visual highlight of finish position by field size by furlong
-Color scales provide aid to decipher clues where the best results are
-Competitiveness visualized (presumably Heart)
— More larger field races filling would enhance the data’s story telling
— More races and larger fields should highlight where Heart takes over
— — If Heart is truly showcased in big fields
Cons:
-Hard to decipher how frequent repeat results are for: position x field size
Alpha:
-If the skill of Heart is truly showcased in big fields, this chart would be able to broadcast those patterns fairly easily
— If this is true for Heart, is the threshold the same field size for all horses?
— — Do some have big hearts from 8+ horses? 10+ horses? 12+ horses?
-Assists with race selection either attacking bigger fields or avoiding them
How to improve this chart:
-Give it a name, sheesh
-Have only 1 scale or have 2 scales for all horses, not relative scales
— The scale should be static for every horse, regardless of race dynamics
— Influenced by horse field and finish position
— — This is true for the X and Y axis dynamics
-Swap the data from X to Y axis and Y to X axis
— Would look like this:
-Left to Right reads more intuitively then Bottom to Top
-Create background bands to highlight which aggregate results are profitable, like this:
Opinion:
-Valuable chart, provides clues on field size, structures to look for:
-Purple Box: High Field Count + Successful Result
— This is the high +ROI area, big fields and podiums, likely heart
-Red Box: General regressive results as field size grows
— Highlights potentially standard results, ROI neutral-ish
-Yellow Box: Poor Performance in Small Fields
— Do not want to be here, ROI negative area
-20 Horse races are outliers, improbable to quantify accurately
Racing ROI: Bar Chart layered with Line Chart
Pros:
-Provides visual highlight of financial results
-Best trend analysis we have
Cons:
-Too many assumptions on $Crown payments converting to a $Derby value
Alpha:
-Horses have proven to be streaky, look at the consecutive green candles
— When you recognize you hit the hot streak, run it until it’s legs fall off
-Aging impacts should also be highlighted here and you watch the trendline
— The white line is a great indicator of the horse’s ROI trajectory
How to improve this chart:
-The X-Axis on both sides should have 0 on the same plane, currently do not
— Looks like this horse turns profitable on 6/6/23
— — When the white line crosses from brown to green
— Horse was actually profitable from the first race, check right X-Axis
-Have only 1 scale or have 2 scales for all horses, not relative scales
— The scale should be static for every horse, regardless of race quantity
— — This is true for both the Y and X axis dynamics
Opinion:
-Love this chart, gives me trend, ROI, streaks, aging, love it
Race Time Comparisons: Box Plot
Pros:
-Provides visual highlight of race results for horses with same overall grade
-Comparative data that provides context — you vs them
-Another aid in determining distance preference
-Filters for specific archetype comparisons
Cons:
-Not enough data early on, but will improve over time
— S+ horses had almost no data to compare against
-Unclear if the data will continue to build for every generation or if the results are reset at some point…or if it only contains rolling timeframes
Alpha:
-Prefer to look at this graph horizontally ACROSS furlongs, not performance WITHIN a furlong
— End of the day, your horse may or may not be fast, so ‘waiting’ or looking for its dot in the upper quartiles may never come
— Therefore, just look across the furlongs for the highest dots and race there, that is where you know the horse has its most potential, whichever furlong has the highest dot… regardless of how high it is
-Aging impacts should also be noted here as you watch potentially once high dots start to drift lower and lower on the box and whiskers
-Temper impacts will also be visible here… do you see tight groupings or wide groupings? Pretty clear answers from the results will determine temperament strength
How to improve this chart:
-Make the latest race result a different color dot, like red
— When there are alot of races, impossible to know where last race landed
-For the filters, should only show the horses races on those filters, not all races on every filter
— If you want to see how your horse performed ‘off pref’ then if you click the off pref filter, it currently still shows all race results, not just the races ran on that archetype
Opinion:
-Love this chart and anything speeding up distance preference discovery
As these charts receive enhancements or when new information emerges, this blog will undergo continuous updates, offering our perspective on how to extract the most alpha.
