3 ms·
Are you sure this isn't just due to load balancing? You might be hitting a different node everytime, they probably don't sync too quickly, and with the amount o
by sgk284 17y ago
Are you sure this isn't just due to load balancing? You might be hitting a different node everytime, they probably don't sync too quickly, and with the amount of traffic digg gets, the numbers could easily vary by that much in a fairly short period of time.
Assuming malice seems premature.
- viraptor 17y agoIt should be relatively easy to prove... Poll the results for ~1 hour for all (or front page only) stories. Generate proper statistics for the speed of change for each entry (stddev per article per each phase - raise, sustain, fall). If all ads jump as much as this sample, it would be quite easy to show the difference. Even a side-by-side graph of votes for articles and for ads should show a if anything strange is going on. But showing a sample of size == 1 is more or less pointless.