User talk:SandyGeorgia/ArbVotes2011/Guides

Counting

It's hard. Thank you for catching my error :) NW (Talk) 22:42, 26 November 2011 (UTC)

Metrics for rating guides

The concept of rating the guides on how successful they are at predicting the results is an interesting one. There are at least two other 'guides to guides' this year (in case you are not aware of them): User:Monty845/ACE2011 and User:John Vandenberg/ACE/2011 guides. I'm currently trying to work out whether what I wrote last year here counts, as I supported all the candidates that got elected (well, when supporting 17 of the 19, it wasn't that hard). So looking at the second table here, I believe my score would have been 14 (12 supports were correct and 2 opposes were correct, with 5 of my supports going to candidates that didn't get in). Of course, this year, with only 8 seats available for 17 candidates (compared to 12 from 19 last year), it will be a bit harder to get the same success rate. But I thought it was worth pointing out that getting a high number of successful predictions last year wasn't that difficult (it can't have been, if I managed it!). Carcharoth (talk) 09:14, 27 November 2011 (UTC)

I just looked at your guide, and I gave up before I ever figured out who you were voting for and against-- perhaps other less-busy readers will have longer attention spans than mine, but I'm unclear why readers would want to read about how your thought processes evolved and your votes changed. SandyGeorgia (Talk) 16:55, 27 November 2011 (UTC)

You have a point. When I've finished it should be clearer. Not that this has anything to do with what I originally posted above. The point I was making (badly) is that it shouldn't really be at all relevant how closely a guide matches the overall result. I think the approach John Vandenberg has taken is better. Actually critique each guide on how good they are as guides. How well does each guide bring out something about the candidates that other guides fail to do? Though annoyingly I see he too has fallen into the trap of thinking that guides that best match the overall vote of the community are somehow 'leaders' (a better metric would be guides that support arbs that do well in the role, such as producing good case decisions and not flaming out or resigning early). Carcharoth (talk) 19:45, 27 November 2011 (UTC)