In the spirit of anticipation about getting my hands on a QRNG I have publicly released my meta-testing harness. Happy flipping! May you rise above 50% and stay there!
Truth in the Flip: A 3-Trillion-Bit Meta-Guessing Experiment
Can you guess the outcome of a coin flip with better than 50/50 accuracy? Traditional statistics says absolutely not. Independent events have no memory.
But what if you aren’t guessing the coin? What if you are anticipating the nature of randomness itself?
I have always operated on a specific philosophical premise regarding entropy and standard deviation: You can count on random to be random. It tends to maintain standard deviation, meaning long, uninterrupted runs are mathematically much less likely to occur than alternating noise. Because you can rely on random for change, my hypothesis is simple: whenever you see a pattern, bet against it.
To test this, I built a high-performance, multithreaded C# testing harness. It doesn’t guess “Heads” or “Tails.” It evaluates the relationship between consecutive states:
-
If the last two flips were the same, the algorithm anticipates the next flip will be different.
-
If the last two flips were different, the algorithm anticipates the next will be a repeat.
The 3-Trillion-Flip Milestone
To prove an edge against a perfectly random baseline, you cannot rely on small sample sizes. Localized variance (luck) will always create artificial streaks. You need an astronomical amount of data to calculate a definitive Z-Score—the statistical measurement of how many standard deviations your result is from the expected 50/50 mean.
My testing rig utilizes a custom BitFactory memory-pooling architecture to feed 16-Kilobyte(adjustable) chunks of pseudo-random data to parallel worker threads. The simulation has been running in real-time for about three days, but thanks to the multithreaded engine maximizing the CPU, it has already compiled over 11 days of cumulative processing time, churning through over 47.7 million flips per second.
We just crossed 3.15 Trillion total flips.
The overall anticipated edge has settled at a microscopic 50.000013%. While that number is incredibly small, the vital takeaway is the consistency. The Z-score has been fluctuating but has remained strictly positive for the last 1.4 trillion consecutive flips. The math is currently projecting that we will hit a Z-score of 1.96 (a 95% confidence interval) in about 12 days.
The Next Step: True Quantum Hardware
Currently, this harness is running against a standard pseudo-random number generator (PRNG). If the edge holds and breaches a Z-score of 3.00, it proves that the algorithm is successfully exploiting the deterministic, hidden math equations buried inside the PRNG’s code.
The ultimate goal of this project is to swap the PRNG data stream for a true Quantum Random Number Generator (QRNG) operating in full entropy mode. Measuring this algorithm against true, physical quantum anomalies will be the definitive test of the negentropy hypothesis.
The Code is Open
I have officially made the TruthInTheFlip repository public. The architecture is completely decoupled, meaning the random input is injected via a thread-safe delegate. It is trivial to swap my C# pseudo-random source for your own data streams, APIs, or physical hardware.
I invite you to pull the code, spin up the multithreaded engine on your own workstation, and watch the Z-score calculate in real-time.