Hacker News Reader

Comment by bitL | Hacker News Reader

One of the universities I was attending had a policy that when you get all answers on a multiple-choice test wrong, you get an A, because the probability of that happening by chance is impossibly low. Seems like 1951 Hoffmann is an example of that attitude ;-)

View on HN

This does not seem like a viable scheme in typical multiple choice tests. In a typical test question, there are often (almost always?) obvious non-answers, so it is significantly easier to guess what is wrong than to guess what is right. When you make a multiple choice test question with, say, 4 choices, it is actually hard to make all 4 choices look reasonable.

I guess the deterrent to taking advantage of such a strategy is the risk involved. You are wiped out if you miss just one. So you have to be supremely confident or a real gambler to knowingly adopt such a strategy.

Reminds me of an interview I saw with one of the co-authors of "The Disaster Artist," the book about the making of "The Room," a famously terrible movie. He said that was what made "The Room" so fascinating - it's rare to find a creative work where every single decision was wrong.

That's overstating it IMO - The Room's production values are pretty adequate. The sets are OK, the sound's decently recorded, it has a proper soundtrack by a real composer, the camera's mostly in focus and properly exposed etc. There are plenty of shitty DSLR films with equally bizarre dialogue, terrible acting, flawed structure etc but the unique joy of The Room is that the decent-ish production values sucker you into thinking you're watching a "real" film.

Similarly, this automobile is well built despite having every design decision be wrong. There's some distinction between the architecture and the craftmanship - a poorly crafted and poorly architected anything is just a pile of rubble, but a well crafted thing with poor architecture is fascinatingly bizarre.

Great point. You don't see issues with basic filmmaking like with "Birdemic" - inconsistent audio levels, etc.

Ha interesting. OK taking this a bit too seriously -- it seems like there should be a better reward than that.

Suppose you actually know all the answers on the test, and you know you know them. You have two choices:

1) Mark them all correctly

2) Mark them all incorrectly, based on your knowledge of the correct answers

There's no incentive for doing #2, because you get an A either way. But if they offered a better reward, like being able to skip the next test with an automatic A, then #2 would look appealing.

It was more like a badge of honor. When you looked at the results of an exam, it was a few As, then a bit more Bs, then bit more Cs etc. then a huge number of Fs (fail) and at the very bottom were 1-2 As of those crazy people that took the "all wrong" option. It was the most difficult university of one of EU countries, so those exams weren't trivial and designed to kick most students out. When you got such an "inverse A", it was something to brag about in a pub later as just one correct answer would make you fail the exam...

Bit like shooting the moon in hearts.

I had a history class that offered this for the multiple-choice section of exams. I was tempted to do it, but never dared, because there was always a question or two I didn't know and would've been guessing at the wrong answer...

Reply on news.ycombinator.com