Wow I can't believe that no one's mentioned it: gmail shots FTW. See here: viewtopic.php?f=6&t=90380
The score comes in. Search your mail for "180." If you don't get that score, take a shot. Now search "179." Still nothing? Shot. And so on...
I had planned to do this since day 1, literally. However, the more I think about it, and the more I think my score is going to be about 5 points below my average, the more I think I'm going to do it forwards...
Let's say my range of scores that I thought I could reasonably get is 140-160. Both the 140 and 160 are outliers, i.e. more than 3 standard deviations away from what I should get; I reasonably think that I should get somewhere around 152. Start at 140....there should be no way I did this poorly. Check for this score. Nope? AWESOME! Shot. 141...shot, 142...shot.
With the original method, if you get a high score, it will be a surprise, but you'll be sober, and unless it's damn near perfect, you're getting only bad news on the way. With the "forward shots" method, if you end up with a low score, it's a surprise all of the sudden when you come across it, but unless it's at or near the lowest possible score, you continually get good news at least until you find it out. If you end up with a high score, you are continually getting pumped as you keep checking. And getting drunk.
Obviously, there's the concern that you hit your high score and all of the sudden realize that you must have scored even below what you thought was your lowest possible score. So if you think your low score is 165 and high is 180, and you get to 180 and haven't found the email yet, you're very quickly going to be pissed. Thus, you have to be generous on the low end, which is why someone who reasonably said they got a 152 should go to 140 on the low end.
Also, I think 1/2, 1/3, or beer shots are the way to go...if your range is 20, and you score on the higher end of it, you won't be able to see your score when you finally come across it. I think about 10-12 points below your predicted score is a good number to be safe; 1/2 or 1/3 shots should play at this range.
Now that I think about it, another way to do this would be to start with a score that's 1 or two points above your average. Work down until you're confident you didn't do that poorly. That way, as you're working down, there's still a chance that you did better than your average, so there's still hope as you move down. Alternatively, start at your predicted score and check alternately above and below that score. You alternate between really good news and really bad news.
Finally, it also might be helpful to know the curve before this exercise. If you know the curve was a dastardly -8, you should adjust your low range accordingly. If it was maybe more generous a la December 2009's -14, then you might not need to go as low.