First, some of my mistakes
- October SMEC Machines: I got to learn a very valuable lesson here: Write the key before the event gets run. Impossible problems, confusing wording, conflicting diagrams, and more problems happened here! Writing the key gives you a chance to look over your exam and fix mistakes, and slightly more importantly, helps you make sure that a problem isn't really, really awful.
- Pearl City Machines: The exam here also had its issues, but it wasn't as bad as the October SMEC, so I won't focus on those. Instead, here's an error that is pretty much only encountered online. It starts with a usual error, one question didn't have a correct answer choice. So, for Division C, I just threw that question out. However, Division B had that same question with that same error, and that event wasn't going to run yet for another couple of hours, so I figured I'd jump in and try to change the answer choice to be correct for them. Now, the major issue turned out to be that the answer choices didn't update for everyone. Rather than having teams refresh, I just asked teams to chat me their answer if it didn't show up. BAD IDEA. And to make the situation worse, I only changed one answer choice, so if a team had one partner with the updated choices and another partner with the incorrect ones, it was obvious what the answer was. In the future, I would (a) just throw the question out, or (b) change ALL answer choices so I can just have teams refresh and/or I can post what the choices are supposed to be in the chat box without accidentally giving an advantage to a team who it wasn't updated for. Now, granted, this was just one question, and I threw it out, but I now know for future reference what I should do.
- Regionals Codybusters: The event overall was too short, 13 questions, but that wasn't the big problem here. My main issue was in the good ol' Cursed Spanish Caesar Cipher. It was set up like a xenocrypt, but supposedly was a shift cipher, except that the ~Ns were all autofilled to that letter. In the same question, the English translation had also been appended after the Spanish version with a different shift key or something crazy like that, and it was just a super confusing and completely unexpected. Creative, but way outside what the rules allow and a little bit out of left field.
- Regionals Astronomy: Astronomy is a hard event for everyone, ES and competitor alike. I appreciate all of the effort the ES put into this event. DSOs from last year were used (the Three Quasars showed up...), giving someone like me, who did it last year, an inherent advantage. As I went through the exam, I noticed a couple of questions that seemed vaguely familiar and I couldn't put my finger on why. Then, I read one particular one, and it clicked. These were the exact questions from the SSSS Astronomy test I wrote, and that gave us an advantage for very clear reasons. Just another reason that direct copying from exams isn't a good idea. On top of this, there were a couple of questions that did not have correct answer choices (we just did our best and got lucky), and a couple of questions that had two identical answer choices. And one last really minor thing I'll bring up. There was a free-response question that would be graded only in the event of a tie. It was long and took effort to answer. We had plenty of time to answer it, but for the effort that was put into it, I would have appreciated some credit for the problem. Minor issue, though, and I realize that there are reasons for not wanting to grade it. Just thought I'd bring that up.
- Regionals Circuit Lab: Nothing big at all. Actually pretty well-run altogether. I just wanted to mention that there were quite a few extremely vague questions. In the end, though, it didn't turn out to actually be too awful.
- State Machines: Way, way too easy. It only had a small handful of calculation questions, a small handful of definition questions, and the rest of the questions wound up being a lot of Machine Identification. Machine ID is extremely subjective for lots and lots of reasons, and I know my partner and I talked through a lot of technicalities of the examples they gave for the examples we needed to identify. Machines isn't an ID event, I enjoy the physics of it, not trying to debate whether or not something is technically a wedge, inclined plane, or lever. Again, I appreciate the contribution of the ES, and the effort they put into it. It wasn't copied, and the event actually got run, so again, I have gratitude for the ES. But for someone running machines in the future (if it ever gets run again... I have suspicions, but that's another story), try to avoid putting a ton of Simple Machine ID questions in there, because (a) mainly that they are subjective, and (b) the event has so many way cooler parts to it.