Indeed, it could be interesting to make a heatmap based on explicit responses, then another based on HN's internal IP logs (w/ the clunky geo IP deductions that

Indeed, it could be interesting to make a heatmap based on explicit responses, then another based on HN's internal IP logs (w/ the clunky geo IP deductions that can be made from that), to see if there's any areas of large divergence – then hypothesize if those are more due to: - selective propensity-to-respond-to-poll - misleading geo IP data - other software-agents/traffic-drivers/usage-differences (eg: some areas/cultures skim headlines once a day, others get deep in the comments)