Unlocking the Secrets of Longshots in Horse Racing

In the quest to enhance our understanding of horse racing dynamics, my research team, composed of Salem and Kondo-kun, has focused on a fascinating aspect: how running styles influence horse performance and betting outcomes.

Over the past weekend, we conducted tests using a scoring system designed to evaluate horses differently than traditional methods that rely heavily on popularity. Our scoring table assigned points based on various factors including past race times, the subsequent performance of competitors, current race conditions, jockey influences, and rest intervals.

Focus on Longshots

What stood out during our analysis was the performance of longshot horses. Over two days, we noted that 68 horses ranked 6th favorite or lower finished in the top three positions. This revelation prompted us to investigate further.

But why concentrate on these longshots? The answer lies in the potential returns from trifecta bets—a common betting format in horse racing where selecting the top three finishers is key.

The Impact of Longshots on Betting Returns

We examined trifecta payouts across 1,589 races in the 1 to 3-win class (both turf and dirt) in 2025, focusing on how horses placed 6th favorite or lower performed. Out of these races, a third saw payouts solely based on the top five favorites, averaging around 1,520 yen. In contrast, the remaining two-thirds had at least one longshot in the top three, yielding a much higher median payout of 10,250 yen.

Relying solely on popular horses might yield frequent wins, but the payouts tend to be minimal, limiting return rates in the long run. Our findings confirmed that the key to long-term success in betting lies in identifying that one crucial longshot in every race.

Tracking the Performance of 68 Longshots

Among the 68 longshot horses, our scoring table successfully identified 10 that were clearly top contenders. Additionally, 21 horses finished closely behind in 4th to 6th places, indicating that almost half of the successful longshots aligned with our scoring parameters. Notably, one successful trifecta payout over 10,000 yen stemmed from this segment.

The remaining 37 longshots ranked 7th or lower in our assessment, breaking down into two categories:

• Data-poor Horses: About 15 horses lacked sufficient performance history, either being new or having previously underperformed.
• Overlooked Horses: Approximately 22 horses had relevant data that wasn't considered in the scoring system, which raised questions about our current methodology.

Revising Key Factors in Scoring

Analysis of the overlooked horses revealed crucial factors absent in our initial scoring setup:

• Forward-running style: Horses consistently placing in the top 30% at the 4th corner in their recent races.
• Strong performances on the same racetrack: One horse showcased consistent success at the Kyoto dirt track, finishing in the top three six times out of eleven.
• History of strong racing in higher classes: Some longshots had bad finishes in recent races but previously excelled.
• Horses returning from breaks: Many scored poorly due to long intervals between races.
• Adjustments based on race conditions: Changes in surface type (turf to dirt) and race form (flat to steeplechase) also played roles.

Testing with a Larger Dataset

To further validate our findings, we expanded our analysis to include data from about 13,000 horses classified as 6th favorite or lower in 2025 for races in the 1 to 3-win categories. The statistics revealed that longshots capable of leading showed a 12.4% success rate, while those coming from behind exhibited a 7.8% rate. Moreover, horses with a track performance record of 40% or more achieved 12.7%, compared to 9.6% for those without.

Additions to the Scoring Table

After analyzing these factors, we decided to revise our scoring table by adding points for forward-moving horses and those with good racecourse history. However, upon retesting 1,385 races, we discovered an unexpected outcome: while we increased the number of identified longshots, the frequency of those finishing in the top three declined.

The reason behind this perplexing fall-off lay in the nature of the horses boosted by the additional points; many were not backed by strong performance histories or timings. As such, these previously overlooked longshots fared poorly compared to those already recognized due to their accomplishments.

Refining Our Approach

Consequently, I have decided not to integrate this scoring addition due to its adverse effects. Instead, the emphasis will shift towards honing in on time-based performance metrics to promote reliable selections. The challenge remains to disentangle track conditions from horse performance times.

Looking Ahead

Next on our agenda is the analysis of a remarkable horse known for consistently finishing second as a longshot in the Queen Elizabeth II Cup for three consecutive years. This intriguing trend will prompt us to explore how market perceptions can overshadow true potential.

We appreciate your support and welcome any feedback regarding our ongoing research, a collaboration between horse racing analysts and AI researchers. As we strive for continual improvement in our methods, we look forward to sharing more insights in the future.