You wouldn't catch all of them but I believe you could get a pretty good sample by running a single node. And you won't be able to get historical data unless someone else has been tracking this and shares their data with you.
There's some discussion about reorg frequency/depth on BitcoinTalk[0][1]. Also there is this academic paper[2].
There's some discussion about reorg frequency/depth on BitcoinTalk[0][1]. Also there is this academic paper[2].
[0] https://bitcointalk.org/index.php?topic=1403436.0 (2016)
[1] https://bitcointalk.org/index.php?topic=5351869.0 (2021)
[2] https://dspace.mit.edu/bitstream/handle/1721.1/127476/119301...