I finally broke my first hydrometer a few days ago, so I picked up two new hydrometers at the LHBS. I was calibrating them today in tap water, at about 78F, and after adjusting for the temperature, one of the hydrometers is off a full 0.010! That's a lot of erm, error. The actual reading was 1.080, so I used the paper included and added ~ 0.002 points to adjust. The other wasn't off by much, after adjusting it for temperature, it has an error of only 0.003. both readings are high.
I'm using the most accurate one for now, and probably will not fuss over the 0.003 correction too much. But man, 0.010? If I hadn't checked, I would have been pretty screwed.
In light of that, I was wondering if we could make a histogram of the error of our hydrometers. I mean, surely we'll be able to get a good sample size of at least a few hundred on a site that requires each member to have one! If the poll option looks funny, I understand. I mean it to say "my hydrometer reads low by 0.003".
edit: it seems that we can only vote once. Oh well. Choose whichever one you wish, if you have more than one. We should still be able to get reasonable statistics.