Photo Finish Live

Project Details

Studio: Third Time Games

Team Size: 15

Engine: Unity

Languages: C#, React

Hardware: PC

Roles: Lead Systems Design, Gameplay Scripting

Photo Finish Live is a real-money virtual horse racing ecosystem. The game combines a non-deterministic —but largely skill-driven—  race algorithm with a robust and realistic breeding system. The result is a market full of unique horses whose value is based on real earning potential. Players race against other owners for fiscal prizes and sell their horses through a player-driven marketplace.

My Duties:

1. Invent a true-to-life genetic system that determines all racing, breeding, and life characteristics of a gamified racehorse.

2. Write a race simulation calculation that is simultaneously deterministic enough for players to express game skill through success but unpredictable enough for the outcome to be wagerable with real money.

3. Generate all races and tracks in accordance with real-world standards and accommodate all spend levels, skill points, race conditions, and game partnerships.

4. Propose a revenue model which satisfies the needs and expectations of the investors and the development team.

Major Accomplishments:

1. Invented a genetic system that is (as far as I know) unique to the world of videogames. It uses real-world genetic rules to determine every feature of a game horse including racing, aging, injury, breeding, and coat coloring. It is hard to express the extent of the work here, but I view it as my current greatest accomplishment in the industry. Here are some features of of the system: 

1. It controls the expected rate of system progression along a multi-year timeline while accommodating thousands of individual fluctuations due to play rate and outlier values.

2. It simultaneously balances player expectations against desired system targets such that foals (newly bred horses):

   1. Clearly follow the strengths of their parents.

   2. Benefit significantly from skilled breeding strategy.

   3. Do not constantly creep up in power outside our intended rate.

   4. Are not predictable enough to “crack the genetic code” (make future breeds predictable).

3. It ensures that even poorly performing horses can produce winners and even great winners do not continuously produce great children. Players do not need a top tier horse to find success and cannot buy their way into long-term unskilled profits.

4. It uses real features of genetics —like regions, recessive traits, and allele expressions— to balance outcomes instead of relying on brute force.

2. Developed a set of tools in Google Sheets that generate game-ready json data. In addition to a significant improvement in development time, the system also:

   1. More than halved the number of schedule mistakes as compared to direct json editing.

   2. Generates all components of a race separately for fine tuning before combining them for json conversion.

   3. Comfortably accommodates both randomly generated and hand-written content without interrupting the tool flow.