Why did Kaspa replace its earlier block-sampling method?

Kaspa's previous block-sampling method — which used the hash of each window block to decide which past blocks to include in a difficulty calculation — turned out to be gameable by miners. Because a miner could influence hash values, they could bias which blocks got sampled and potentially skew the difficulty estimate in their favor. The replacement approach defines the sample window by two simple, miner-agnostic parameters: `length` (the real-time period in seconds) and `size` (the number of blocks to draw). A miner who wants to affect the sample now has to pay a cost comparable to discarding a valid block — making the attack economically unattractive. For a beginner, the takeaway is that Kaspa's difficulty algorithm was deliberately hardened so no single miner can tilt the playing field.