Strategy A
Only high-confidence edges that pass strict quality filters. Noise-resistant.
- PnL
- +$223.34
- ROI
- +5.83%
- Win Rate
- 48.3%
- Bets
- 770 / 383 settled$3,830 staked
- W / L
- 185 - 198
- Open
- 387
Live · Operational•Last update: —
Tukuyomi
Proprietary arbitrage strategies that capture prediction market inefficiencies. Three independent strategies paper-trading 24/7. All figures are auto-aggregated from live operations.
Cumulative PnL
+$855.37
$9,760 staked · 976 settled
Total Bets
1,812
719 open
Uptime
7d
Running continuously
System Status
Operational
107 edges detected (latest)
Strategy Performance
Three strategies running in parallel. PnL / ROI finalize after settlement.
Strategy B
Bets on every detected edge equally. Baseline benchmark.
- PnL
- +$190.70
- ROI
- +4.98%
- Win Rate
- 49.09%
- Bets
- 666 / 383 settled$3,830 staked
- W / L
- 188 - 195
- Open
- 283
Strategy C
Directional strategy mimicking proven betting patterns.
- PnL
- +$441.33
- ROI
- +21.02%
- Win Rate
- 52.38%
- Bets
- 376 / 210 settled$2,100 staked
- W / L
- 110 - 100
- Open
- 166
Daily Activity
Last 30 days — placed bets and cumulative PnL
Strategy Design Philosophy
Why three strategies run in parallel
Structural inefficiencies exist in prediction markets, but no single strategy can capture them all. Tukuyomi runs three independent strategies simultaneously to compare their real-world performance in parallel. The differences in their results are the core data for market understanding and strategy selection.
Strategy A
SelectiveFilter out noise and measure baseline win rate
Only edges that pass multiple quality indicators are selected for this strategy. A conservative benchmark designed to verify whether market inefficiencies truly exist where they appear.
Validation
Testing: win rate of curated edges, long-term stability, opportunity cost from over-filtering
Strategy B
BaselineObserve every detected edge as-is
No quality filters applied — every detected edge receives an equal bet. This baseline captures the raw signal distribution of market inefficiencies. Comparing with A and C reveals where filtering adds value.
Validation
Testing: population distribution of detected edges, false positive rate, marginal utility of filtering
Strategy C
MimicReplicate advantages observed in public data
This strategy mimics the directional characteristics of proven betting patterns observed in publicly available data. Instead of independent judgment, it reflects the structure of existing success cases to reproduce their edge.
Validation
Testing: reproducibility of mimicked patterns, correlation with proprietary strategies, persistence of historical advantages
Specific filter thresholds, reference data sources, and implementation methods are not disclosed. This dashboard exists to publish the integrity of results — not to make the strategy logic reproducible.
Open Positions
719
active bets
+18.2%
avg edge
$7,190
total staked
5
sports covered
By Strategy
A328
B253
C138
By Sport
MLB257
NBA198
Tennis149
NHL89
Soccer26
Recent Scan Cycles
Live execution log — published as proof of operational integrity
| Time (UTC) | Edges | Duration | Status |
|---|---|---|---|
| 08:25 | 107 | 127.5s | OK |
| 08:20 | 105 | 126.7s | OK |
| 08:15 | 105 | 120.7s | OK |
| 08:10 | 105 | 128.4s | OK |
| 08:05 | 105 | 127.1s | OK |