In high school I had a class where your final grade was based on a total number of points earned through the semester. The final exam was worth a large portion of the total semester points—let’s say it was 500 of your semester’s possible points. Your percent-score on the exam determined how many of those points you received. (Ace the exam, and you get all 500 points.)

The exam was many hundreds of multiple-choice questions; The exam was so long that no one could ever finish it. The questions had to be shuffled to mix the material taught in the course. Every year the questions were identical, but each year the teacher made a copy of the master list, cut up (yes, with a scissors) the questions, shuffled the strips, and then taped the questions onto a sheet with question numbering, to create a unique Frankenstein-exam every year. This Franken-xam was then photocopied (via a Volkswagen Beetle sized behemoth in the main office) to produce the actual exams.

In the days before the exam, we were told to work at our own pace, to answer each question (skips counted as wrong answers) and to simply stop when time was called. Afterwards, the teacher would calculate the average number of questions attempted by the class. That average was then used as the possible number of questions for calculating our exam scores. (Thus the shuffling to create an exam that is however-long we made it as we took it!) If you went farther than the class’s average attempted number, then you could score some extra points (if you get the answers right, of course) to offset any wrong answers you had along the way. A lot of work to shuffle it every year, but it was a neat idea.

I think it had always worked because kids just didn’t care enough to think it through. We weren’t told the total number of questions, nor what previous classes had attempted. But, for discussion here, let’s say the class’s average-attempted is 200. And let’s say I were to answer 227 questions, but I get 24 wrong. That feels like an 89%, right? No, actually I end up with 203 correct answers, which is more than the class’s average-attempted of 200. I actually score 101.5% and I would get all of the exam’s 500 points towards my semester total. Wait, there’s more: As extra credit, my 3 extra correct answers (my 203 against the 200 attempted average) become extra credit points just added right to my semester total. I’d get 503 points towards my semester!

After the exam was announced, two of my friends and I, realized…

- Do not tell another soul about this or everyone will fail the exam.
- When you get the test, go as fast as you can. Our goal is to attempt as many questions as possible.
- The goal isn’t to get every question right— The goal is to get a lot right.

For example, if we could get just 60% right—normally a really poor performance on an exam—while attempting twice as many as the class average, we win big. Say, 200 average-attempted, against our 400 attempted, at 60% correct (240 correct answers of 400)… we’d score 120% on the exam, plus 40 extra points (our 240 correct above the 200 needed) That’s 540 points towards the semester. And, if we could get 75% correct, while attempting 3 times as many questions, then our exam score is 225% (that’s our 450 correct answers, while needing only 200) plus an extra 250 points (that’s our 450, minus the 200 to ace the exam) That’s 750 points towards the semester! Now do you see the attack? :)

I never understood why no one else ever tried that.

I know this is a minor thing in the universe of problems with secondary education and grading, but I found the hack interesting.

~ Bruce Schneier from, https://www.schneier.com/blog/archives/2023/10/hacking-the-high-school-grading-system.html

slip:4usebo19.

…and I’m actually not sure if what we tried even worked. You thought I was going to have a clear take-away about my actual scores, or the test never being given again?! No the take-away is: Oh, I’ve been thinking like a hacker for a Long. Long. Time.

ɕ