Read the explanation
The digital-root page treats its date as a string of digits. For nineteen ninety-six June thirteenth, the digits sum to thirty-five. Summing three and five gives eight, which the source repeats to form code eight eight eight. The digits are calendar notation, not a physical frequency. The displayed arithmetic is reproducible, but it supplies no evidence about personality, destiny or causation. Change the date to nineteen ninety-six June twenty-second. The year and month digits stay the same, while day digits two plus two also sum to four, just as one plus three did. The full digit sum is again thirty-five, so the root and repeated code are identical. This explicit pair is a collision: two distinct valid dates share the label. It shows why matching a code cannot uniquely identify a date, much less establish a shared psychological cause. The source enumerates every valid UTC calendar date in a chosen year window and counts roots one through nine. A one-year window for nineteen hundred contains three hundred sixty-five dates; two thousand contains three hundred sixty-six because of the leap day. Counts depend on calendar structure and the chosen range. A matching share is matches divided by total valid dates, not a prediction score. The exported JSON keeps all counts, digit trace and explicit limitations so the arithmetic can be audited separately from unsupported interpretations.