Weekly Overview
Yesterday's Canada 28 data shows that over the past seven days, 2,760 draws were conducted, spanning May 25, 2026, to May 31, 2026. Overall, the big/small ratio was 50.94% big numbers to 49.06% small numbers, while the odd/even ratio was 49.53% odd numbers to 50.47% even numbers. These figures closely align with the theoretical probability of 50%, reflecting a reasonable performance under the law of large numbers.
7-Day Trend Analysis
Daily data revealed some fluctuations in big/small and odd/even ratios. On May 29, the proportion of big numbers hit a weekly low at 45.52%, marking the only significant deviation from the average. Conversely, on May 31, the big number ratio rebounded to 52.99%, indicating a short-term reversal trend. Such variations can be attributed to the effects of small sample sizes.
AI Algorithm Performance
Among the 30 AI algorithms analyzed this week, the top three performers were "Monte Carlo," "Mean Reversion," and "Anti-Martingale." "Monte Carlo" led with an overall accuracy rate of 45.4%, based on a sample size of 7,932 predictions. "Mean Reversion" matched this accuracy rate, showcasing equally strong performance. "Anti-Martingale" followed closely with a 45.3% accuracy rate, securing its place in the top three.
Notably, these algorithms demonstrated balanced performance across multiple dimensions. For instance, the "Monte Carlo" algorithm achieved accuracy rates of 50.98% in odd-number predictions and 24.56% in combination predictions. This balance indicates the algorithms' adaptability to fluctuating data patterns.
Extreme Event Highlights
This week saw frequent occurrences of extreme events, with rare triple numbers and span extremes drawing attention. On May 31, the triple number "9+9+9" reappeared, an event with an approximate theoretical probability of 1%, making it exceptionally rare. Additionally, span extremes of 9 were observed on multiple dates, such as May 27 through May 29, where several draws reached a span of 9. These occurrences significantly exceeded theoretical expectations, likely reflecting the randomness and short-term deviations in the data.
Summary and Outlook
The algorithm performance and extreme events this week showcased the diversity and Volatility of the data. The top three algorithms demonstrated stability, while the extreme events provided valuable insights into anomalies. Next week, we look forward to further validating the patterns and probability theories underlying these observations in Canada 28 draws.