Genuinely interesting pattern and one I hadn't considered before stumbling onto this thread. I mostly follow MIB encounter data but there's actually a decent overlap with "corridor" theory, the idea that certain linear geographic features - old rail lines, ley lines, river valleys - seem to concentrate anomalous activity rather than just being incidental to it.
Whether that applies to Dogman specifically I couldn't say with any confidence. But railway cuttings are worth thinking about. They're often decades or centuries old, they cut through terrain that would otherwise be largely undisturbed, and teh embankments create cover that most large animals would find useful. That's before you even get into anything stranger.
What I'd want to know is whether the clustering holds up once you control for reporting bias. Old railway lines often pass near towns and roads, meaning more potential witnesses. Anyone here actually mapped this with coordinates rather than just eyeballing a few reports and seeing what they want to see? Would take this a lot more seriously with actual data behind it.
