Boat Race AI System
Backtested on 12,308 real races over 4 years using a proprietary algorithm
A proprietary indicator calculated for each racer, combined with environmental filters, selects only the most profitable races.
① Race Filtering
Limited to 11 venues where Lane 1 wins 55%+ of the time, with an A1-ranked racer in Lane 1.
② Proprietary Indicator Selects the Opponent
A confidential indicator is calculated for each racer. The outer-lane racer (Lanes 4–6) with the highest score is designated as “Lane 1’s opponent.” Without this indicator, the same betting method yields only 63% ROI (a heavy loss). With it: 186%. This is the entire edge.
③ Exclusion Filters
Days where Lane 1’s win rate drops below 70% are excluded based on multiple conditions. Dangerous wind conditions on race day are also filtered out.
④ 2-Ticket Bet (Exacta + Trifecta)
An Exacta bet for steady returns, plus a single Trifecta for high-payout upside. ¥100 × 2 tickets = ¥200 per race.
Verified on 12,308 real races from 2022–2025 (payouts adjusted to 80%).
Annual Profit
Profitable All 4 YearsROI
185.9%For every ¥100 wagered, ¥185.9 is returned on average — far exceeding the typical 75% return in boat racing.
Hit Rate
Exacta 8.0%Wins roughly once every 12–13 races. Looks low, but average payout per hit is ¥1,252.
The Indicator Effect
Without It = Loss| With Proprietary Indicator | Without It | |
|---|---|---|
| ROI | 185.9% | 63.7% |
| Annual P&L | +¥170,000 | −¥100,000 |
Same venues, same bet type — without the proprietary indicator to select races, the result is a significant loss. Publicly available data (win rate, motor stats) alone cannot break 100% ROI.
Targeting high-payout races where outer-lane boats (4–6) finish 2nd. Exacta as insurance, Trifecta as bonus.
2-Ticket System
| Ticket | Bet (Example) | Role | Avg. Payout |
|---|---|---|---|
| Exacta | 1→6 | Insurance (higher hit rate) | ¥1,252 |
| Trifecta | 1→6→3 | Bonus (high payout) | ¥4,673 |
Cost: ¥100 × 2 = ¥200/race. Exacta hit returns ¥600–3,000. Trifecta hit adds ¥5,000–40,000.
Progressive Staking
Strategy C| Status | Per Ticket | Per Race |
|---|---|---|
| Normal | ¥100 | ¥200 |
| 10 consecutive misses | ¥200 | ¥400 |
| 20 consecutive misses | ¥300 | ¥600 |
| 30 consecutive misses (cap) | ¥400 | ¥800 |
Reset to ¥100 once cumulative P&L turns positive. This roughly doubles monthly profit.
11 venues × 34 rules. Average 2.0 qualifying races per day.
Top 10 Patterns by ROI
| Venue | Lane | Races/Yr | Bet | ROI |
|---|---|---|---|---|
| Wakamatsu | 6 | 31 | Exacta 1-6 + Trifecta 1-6-3 | 324% |
| Marugame | 6 | 14 | Exacta 1-6 + Trifecta 1-6-2 | 258% |
| Kojima | 6 | 32 | Exacta 1-6 + Trifecta 1-6-5 | 200% |
| Kojima | 2 | 14 | Exacta 1-2 + Trifecta 1-2-6 ★ | 175% |
| Tokoname | 6 | 30 | Exacta 1-6 + Trifecta 1-6-2 | 167% |
| Gamagori | 2 | 16 | Exacta 1-2 + Trifecta 1-2-5 ★ | 162% |
| Ashiya | 6 | 65 | Exacta 1-6 + Trifecta 1-6-2 | 156% |
| Kojima | 5 | 18 | Exacta 1-5 + Trifecta 1-5-3 | 147% |
| Miyajima | 4 | 22 | Exacta 1-4 + Trifecta 1-4-2 | 141% |
| Omura | 2 | 56 | Exacta 1-2 + Trifecta 1-2-6 ★ | 107% |
All 4 years were profitable in backtesting, but this does not guarantee future results.
Maximum Drawdown
−¥125,780The worst peak-to-trough decline over 4 years. Temporary losses exceeding ¥120,000 can occur, but every year ended in profit.
Max Losing Streak
62
Based on Exacta results. Dry spells of over a month are possible.
Mindset Required
• Commit to at least 3 months
• Bet mechanically through losing streaks
• Don’t react emotionally to single results
Fully automated selection. Just run the script daily.