This pattern suggests a direct correlation between increased latency and the frequency of grandparent references over the expected three-parent scenario.

In response to this finding, a modification to the existing difficulty algorithm was proposed and tested. The original difficulty equation was augmented to incorporate an adjustment factor based on the presence of excess grandparents relative to a benchmark of three parents per block. This adjustment was designed to scale the difficulty in accordance with the discrepancy between observed grandparents and the ideal number of parent references, thereby compensating for latency-induced imbalances. The effectiveness of this revised formula was evaluated under various conditions, demonstrating its capacity to maintain stability during hashrate attacks and under normal operating scenarios alike. Notably, the adjustment proved particularly effective when addressing the challenges posed by a significant minority of miners experiencing up to four times slower latency.

Further analysis revealed that optimizing the adjustment for blocks with two parents—rather than focusing solely on the grandparent metric—yielded more reliable outcomes. This insight stems from the observation that two-parent scenarios are significantly more common than excessive grandparent references under conditions of homogeneous latency across the mining network. The term "homogeneous" here implies that the system functions optimally when latency disparities among miners are minimized, regardless of the absolute values of those latencies. Subsequent adjustments to the difficulty calculation were proposed to better reflect the prevalence of two-parent blocks, further refining the approach to mitigate latency inequality's adverse effects.

However, limitations were noted in the scope of protection this algorithm adjustment could offer, particularly when a smaller fraction of the mining population is affected by high latency. While the revised algorithm demonstrated enhanced performance in scenarios where a larger proportion of the hash rate was subject to higher latencies, achieving similar improvements for cases involving only a minor percentage of disadvantaged miners proved more challenging. This highlights the need for ongoing research and development to extend the benefits of such adjustments to a broader segment of the mining community, ensuring equitable participation and resilience against latency-related disadvantages in blockchain networks.