Science Olympiad at MIT Invitational 2019

[venules]

Inexperienced ES coming through to add on to dragonfly’s point: sometimes you just can’t predict what’s going to happen if one thing goes wrong. During an event at my school’s invite, I noticed that some of the teams were rotating in the wrong order (there were more stations than teams). However, I didn’t call them out on it because I thought that as long as everyone went to all the stations, there wouldn’t be an issue. In the last 15 or so minutes of the event, there ended up being multiple teams piled up at one station because they all needed the same station. Luckily, my partner and I were able to defuse the situation, but that was something that neither of us anticipated to happen. Whether it’s a tournament with less than ten teams or a tournament of MIT’s caliber, sometimes things don’t flow as smoothly as they could, and how you respond to it matters most.

Accessibility of information is something that SO can improve a lot on, not only with invitational tests but scores as well (even some state scores are never posted publicly). I know lumosityfan likes to compare to quizbowl a lot (maybe a bit too much ), but SO really could take a lot from them in terms of sharing of information and scores. People should not have to scour Twitter or Google News for state results, and teams should not have to find and engage in an extensive underground "black market" of tests to find sufficient test resources.

Seconded. Even if that information isn’t released immediately, it should eventually make its way into the public sphere.

[pikachu4919]

I think the difficulty in getting the sheet volume of volunteers is another reason to partner with Harvard. It doesn't have to be a split tournament like I was suggesting, that was more of a hypothetical compromise in case both universities wanted to host a tournament. If it was a GGSO setup instead where the host rotated, you could pool the volunteers from both organizations into one tournament.

It's worth noting that GGSO has always been held at Berkeley despite intent to rotate, because Stanford's policies are too restrictive for them to host a tournament.

Thanks, didn't realize that. Harvard did just start a B division tournament, so perhaps they can join up with Harvard running B and MIT running C. Regardless, my thinking is a larger volunteering pool is better than a smaller one!

There are several internal reasons why a Harvard/MIT collaboration attempt won’t work and in fact might actually be disastrous. I know what they are, but I won’t say much about them here.

[EastStroudsburg13]

There are several internal reasons why a Harvard/MIT collaboration attempt won’t work and in fact might actually be disastrous. I know what they are, but I won’t say much about them here.

This is part of the "compromise" aspect. I suspected there were internal issues involved between the two organizations, but my hope is that eventually they can put differences aside for the benefit of the community (unless the issues are of the "university administration" persuasion, in which case, the scepter of red tape strikes again!).

[primitivepolonium]

Any thoughts on Chem Lab? Hope y'all liked the fun facts.

Thoughts of the (main) writer:
I admit I aimed for an exam that was easier than last year's--last year's exam was insanely long, but this year's was comparable in length to 2015/6/7. I competed at MIT Chemistry Lab in 2015 and 2016 and had no problem finishing the exam although I certainly needed every allotted second. (And given that I got 4th or 3rd both years, I assumed that most of the top 6-top 10 felt similarly.) As for the questions: when looking through my question bank for questions to use, I eliminated a lot of the more esoteric and labor-intensive ones (eg: magnetic susceptibility, crystal field theory, phosphoric acid questions).

Ultimately, I think what went wrong for a lot of teams was either 1) question placement or 2) team management. For 1: Page 2 had a question about ferromagnetic systems inspired by the recent discovery of a gaseous ferromagnetic system. At least a few teams got some points on it, but it certainly isn't your run-of-the-mill magnetism question. Similarly, Question 2 had some tricky acids reactions, and Question 5 (PbF2) as difficult unless you did something like it (and in my experience, I only got familiar with it after getting a whole section of its kind on a particularly nasty exam). If a team got stumped on these difficulties, instead of flipping through the entire exam, they would have wasted a lot of time that they could have spent on the easier trivia questions or on Question 7, which was basically freezing point depression and specific heat calculations. And for 2: I saw teams that got full points on page 1, nearly full points on pages 2-3...and then things just went completely downhill. I can't be certain what happened there, but I would assume it was time pressure.

In addition, a disappointingly small number of people attempted the lab even though we told them to start it right away. It was easier than the majority of the exam and would have taken ~20 minutes tops. The teams that did attempt it did fairly well on it; 2 teams got close to full points.

Anyway, yeah. I'll stop talking. Let me know what you all think. I aim for averages of ~50%, and I severely undershot, so I welcome all suggestions and feedback.

Also just wanted to add to the Harvard MIT conversation: my understanding is that MIT normally gets quite a few Harvard volunteers (~5-10 people) and Harvard gets a few MIT volunteers when HUSO swings around. 5-10 doesn't sound like a lot, but if they sign up, they usually come and stay the whole day, where they are basically able to staff 2-3 events which is a lot of help. The problem lies in schedules. Harvard doesn't start school til the 21st (roughly). Normally this isn't a problem since people return a few days early, but this year, the vast vast majority of Harvard people were out.

[apc1234567]

Any thoughts on Chem Lab? Hope y'all liked the fun facts.

About chem lab, I thought it was pretty fun, and the lab was pretty cool (although it took like 5 mins.) Just curious, how accurate were you supposed to be for the buffers? We got like 1-2% error on the pH, but you can't do that well with grad. cylinders.

There was definitely a massive time crunch though. Information from previous sections carried forward, which made it a pain to split up the test. I also like CFT:(

Also not sure what particles in boxes have to do with physical properties or acids, but that question was kind of fun (although I think I forgot to square n).

[primitivepolonium]

Any thoughts on Chem Lab? Hope y'all liked the fun facts.

About chem lab, I thought it was pretty fun, and the lab was pretty cool (although it took like 5 mins.) Just curious, how accurate were you supposed to be for the buffers? We got like 1-2% error on the pH, but you can't do that well with grad. cylinders.

There was definitely a massive time crunch though. Information from previous sections carried forward, which made it a pain to split up the test. I also like CFT:(

Also not sure what particles in boxes have to do with physical properties or acids, but that question was kind of fun (although I think I forgot to square n).

Thanks so much for your feedback!

Yeah, the lab was supposed to be simple and fast to compensate for the exam (which surprises me that so many teams didn't do it, since for many it could have doubled their final scores). In terms of precision--it is definitely tricky with the graduated cylinders and pipettes though it's possible. I considered giving y'all micropipettes instead since they're pretty precise, but when I tested it out using the grad cylinders/disposable pipettes, I got 2.93 while aiming for 2.94 and 2.42 while aiming for 2.44. With the micropipettes (where you can select specific volumes) we got 2.91 and 2.43, so we decided the precision was comparable and more than enough for the lab. We didn't expect anyone to get precisely the desired pH but within 0.3 was definitely reasonable.

RIP--I also like CFT but they said it would be too out of scope :/ and it was tested last year.

As for particles in a box--yeah, in hindsight that question should have either been removed (few people got to it anyway) or heavily rephrased. The connection to the rules was that it was "colors" (under physical properties), since the HOMO/LUMO gap directly determines the color in many substances, and in conjugated molecules, said gap depends a lot on the size. But again, this wasn't the most direct connection, and if I could do it again, I'd eliminate the calculation parts and replace it with color theory.

In terms of splitting--splitting by question was probably the only real way to split the exam.

Let me know if you have any questions or concerns! Really appreciate your reply.

[will926ok3645]

Might be a bit late but here's my MIT experience with my events:

Experimental: I feel like it was a bit too easy for a competition such as MIT, but then again, you can't do something so obscure that no one knows what the topic is. There were a lot of different ways to take the materials given and incorporate it with the topic. Fairly shocked that a large kitchen knife was available, but it was certainly something to look back on and laugh about. Overall it was fun, and the event supervisors were really nice.

Water Quality: Jeffery (the event supervisor), if you're reading this, both teams loved how peppy you were and respected the amount of time and effort you invested into your test. Making Water Quality into a story was incredibly impressive, and the test had enough difficult questions to separate the top tier teams from some of the lower quality teams (no pun intended). The easier questions were made more difficult by tying it back to the story, which made for an interesting dynamic. Some of the easier questions were also longer and you had to know how to do it in order to get it right. Overall a solid test, maybe a bit on the easier side. The only complaint was that one of the solutions was outside of the 1%-10% range in the rules.

Picture This: Trial event that lasted four minutes. Not much more to say about that beyond the fact that the girl who administered the words would occasionally zone out and miss when we got a right answer, but she was incredibly nice!

Forensics: Probably the poorest run event of the day, but I attribute very little of it to the event supervisors. They were both very nice and and seemed genuinely excited. Preparing individual samples for everyone must have been an extremely laborious task, and I have a lot of respect for both supervisors and any volunteers that set up the samples. As mentioned above, the two Bunsen burners was a hindrance, and a few with a butane base would have been nice, but the supervisors did what they could with the resources given. Our station also did not have a candle or matchbox, and the teams around us did not either. It wasn't until after the event finished and I was talking to the other team from my school that we realized they had a candle and matchbox. You could see a layer of film on the top of nearly every solution for plastic density tests by the time it got to our time block (we were near the end) and there were times that I would go to pick up a sample to wipe it off and i would knock it, making it sink. It's nearly impossible to keep the samples uncontaminated, but these solutions were very poorly kept. The room was also way too small. Me and my partner were crowded out by the teams next to us, and talking with other teams it seems that they had the same problem. Some of our samples ended up getting mixed, so I never got to test the bonus powder given. The test itself was almost too long for its own good. I wish I knew what the crime was about, but we frankly didn't have time to read the scene, let alone finish the test. This also goes for the other team from my school, who finished second in the event. There's nothing wrong with a long test, but I feel that having to choose between a decent analysis or doing the content questions doesn't reflect the skills of the event fully. That being said, it was still a good event and all of the misfiring didn't take away from the fact that the test content was extremely high quality and the event supervisors did a fantastic job supervising what is arguably the hardest event to be an event supervisor for

[pikachu4919]

Forensics: Probably the poorest run event of the day, but I attribute very little of it to the event supervisors. They were both very nice and and seemed genuinely excited. Preparing individual samples for everyone must have been an extremely laborious task, and I have a lot of respect for both supervisors and any volunteers that set up the samples. As mentioned above, the two Bunsen burners was a hindrance, and a few with a butane base would have been nice, but the supervisors did what they could with the resources given. Our station also did not have a candle or matchbox, and the teams around us did not either. It wasn't until after the event finished and I was talking to the other team from my school that we realized they had a candle and matchbox. You could see a layer of film on the top of nearly every solution for plastic density tests by the time it got to our time block (we were near the end) and there were times that I would go to pick up a sample to wipe it off and i would knock it, making it sink. It's nearly impossible to keep the samples uncontaminated, but these solutions were very poorly kept. The room was also way too small. Me and my partner were crowded out by the teams next to us, and talking with other teams it seems that they had the same problem. Some of our samples ended up getting mixed, so I never got to test the bonus powder given. The test itself was almost too long for its own good. I wish I knew what the crime was about, but we frankly didn't have time to read the scene, let alone finish the test. This also goes for the other team from my school, who finished second in the event. There's nothing wrong with a long test, but I feel that having to choose between a decent analysis or doing the content questions doesn't reflect the skills of the event fully. That being said, it was still a good event and all of the misfiring didn't take away from the fact that the test content was extremely high quality and the event supervisors did a fantastic job supervising what is arguably the hardest event to be an event supervisor for

You're right - Forensics is a tricky event to supervise. Bumping that up on a scale of running it for 76 teams makes it even harder. And getting a lab that has as many amenities as possible for an ideal iteration of the running of the event is just as hard if not harder. All your concerns are certainly valid, but in response to it, I'll give you a little insight into running the event that I've gained from doing it as many times as I have.

First off, there's a story with the room itself. I ran this event for the 2018 MIT Invite as well, and I was literally told throughout my writing process that there was a real chance that I would possibly not be able to run any lab at all if none of the departments at MIT would let them reserve a chemistry lab for the event. It's kinda reasonable, considering we're college undergrads (albeit, not even being an MIT undergrad as well) watching minors in a university lab (which then involves complications such as insurance and lawsuits and whatnot), and there was a real issue with that too when in the 2017 MIT Invite, some parents tried to break into the Forensics lab to take pictures of their kids (yeah, that doesn't go over well with university departments/administration).

Honestly, at that point, I was grateful that there was at least a department that was willing to let us use their lab so that I could even write a lab portion for the test instead of writing it entirely dry lab. And we had the same lab again this year, which at least gives us a sense of security that we could know that we would definitely have the amenities that we had last year to run that lab portion at all instead of worrying over possibly not being able to do that. I remember I also ran Princeton 2018 Forensics, and the lab we got for that was bigger but had no gas outlets (and thus no places to plug in bunsen burners), the executives didn't feel like buying portable bunsen burners since they were the more expensive option over alcohol burners, so the competitors all ended up using those. Yeah, I got quite a few complaints about that. Like I said in an earlier post, the lab used this year and last for MIT Forensics only has two gas outlets - thus, only two bunsen burners. There's nothing wrong with buying portable bunsen burners with butane bases, and it's something that could be looked into for future tournaments, but main point is, it's hard to fix something that you can't really have much control over, for sure.

Then, about prepping the materials for the event itself. Yes, we did have multiple REALLY late nights to get everything done by that Friday night before the tournament day. As I have seen it in my experience competing and supervising, there are a handful of ways you can approach preparing materials, ranked in order of degree of laziness of the proctor from high to low:

1) You can create common stations with everything in sort of giant sample banks (i.e. having large containers of every single sample at the front of the room or something, having large containers of all your reagents at another part, etc). This is the worst because of the near certainty for contamination of samples, plus it's harder to keep track of during the day.

2) You can create a small set of samples for however many teams are in a block, and then have all the blocks share these common sets of samples and reagents (so basically, if we did this, we would have created only 13 sets of materials with higher quantities of samples and made everyone share these from timeblock to timeblock). This is the method that the national supervisor uses, and I personally dislike it a lot (my coach does too, and she's known the national supervisor for a long time too) since although it reduces some potential for contamination, it really doesn't take it down by much depending on how many people before you have used those (I remember it ramped up my paranoia levels at Nats 2016 when my partner and I were competing in the very last timeblock of the day, and we somehow still medalled for some reason).

3) To improve upon #2, make small little sample packets for every single team that comes. When you have a ton of teams, this is a massive pre-tournament task, especially for the supervisors and volunteers, but it does have insanely good benefits in that every team's sample is more or less fresh and not contaminated by other teams.

As you know, we chose option 3, and our payment for making it happen was in terms of sacrificed sleep. I can definitely say that there's no certainty that we'll get every single packet perfect. We all make mistakes, and that's why we allotted the time before starting the test for everyone to check if they had everything.

You make a good point about the density solutions. They're tougher to make than you think though - the 46% IPA and each of the salt solutions have to be carefully measured and mixed to make sure they have the correct amounts to yield the correct densities. Even if we wanted to switch them out during the day if they got too messy, we probably wouldn't be able to be sure that we could do it quickly enough in between timeblocks to still save time to assess cleanup penalties and reset the stations altogether. I will say that I did try to switch out the distilled water density solution quite a few times throughout the day, but it didn't work out too much since the existing corn oil in the water from other testing definitely really messed with that. Sure, you can partially dodge larger scale contamination by creating 13 sets of density solutions for each station, but that's also quite hard if you have to be really careful with your ratios, and there's also the consideration that they'll probably still get contaminated anyways, just instead by whichever team had just used them. Bottom line, there's no really good way to avoid those density solutions getting contaminated by other teams unless you also make 76 sets of density solutions for all the teams (which I can definitely say would be a huge pain on top of the setup we already do for the samples). We made it clear that teams take their plastics out after dropping them in and be careful with handling of them, but I guess some teams don't listen.

Sorry about the candles and matches situation - that was just unfortunate in general because both were stolen or misused from stations randomly throughout the day (I noticed one team use an entire box of matches to do flame tests on the non-ignitable ends....sigh). That could have been solved with more backups, which we unfortunately overlooked and didn't realize until it was too late, and we could have also been more clear that we did have cigarette lighters as a substitute for matches in case or that we could light peoples' candles for them if they wanted us to. I'm not sure if we weren't clear enough, or if we did say it clearly, but things just got too crazy. We were just maybe a little desperate on trying to save the candles and matches as the day progressed to last until the end when we realized that these items were accidentally being stolen by teams throughout or being used excessively in a manner that just wasted them.

Finally, tackling the test. Another competitor from this weekend messaged me asking how I would personally approach the test if I were taking it with a partner. So, I will share what I said to them here. First thing's for sure, more often than not, this event in general, due to the way it's structured, is impossible to do without splitting the test and putting complete trust in your partner to do their thing without you looking over their shoulder. It is true that the IDs are worth the most points, also a result of the way the event is structured.

So, if I were doing this, I would have split those ID's with my partner depending on what we were most comfortable with, and with one of us starting the chromatography right after the time starts. Then, we would make sure to get as much of the lab portion done as possible (in which it is completely doable to get everything done in a relatively short amount of time if you practice a ton, first for accuracy and then for speed), then move on to the free response if you finish early, and then right at 20-ish minutes, maybe a little over that, left, one partner focuses on the analysis portion (or the essay, whatever the test decides to call it), and then the other partner continues doing the other free response questions until the 5 minute mark, in which that partner then cleans up while the other partner is still writing the analysis. That would be how I would approach it ideally, but it takes a lot to pull something like that off.

I will say, the nationals tests are also very long as well, but they generally pressure you more in terms of speed due to the generally large amount of easy trivia in the supplement rather than actual difficulty of the content (unless things have dramatically changed ever since I was competing).

I know I've written a long response, but I hope it gives some insight into how many things are taken into consideration when running this particular event that explains why some things happened the way they did. Looking back on my experience as a competitor and reflecting on my more recent experiences as an event supervisor, I can definitely say that I would have not realized the magnitude of difficulty it takes to run an event over actually preparing it as a competitor, but that has become so much more clear to me for every year that I've been in Division D. You can shift as much to the mindset of a competitor while thinking about how to be a good event supervisor, but ultimately, event supervising is a whole other animal to deal with and is an art that takes much practice to master. I can always think of new ways to improve things while considering the constraints I may have, be it material or facility constraints, but I know there are always ways in which I can improve my event supervisorship and I'm glad you thought we did a decent job with the given constraints. We'll definitely see what we can do about that for future years.

[jaggie34]

Thoughts on my events at MIT:

Fermi: Very well written test with a good level of difficulty, there were some oddities in the format (i.e. a question being asked differently in the italics vs. the actual question as a kind of trick?) but overall a very good test.

Circuit: This was the best Circuit Lab test I've encountered, there was a great mix of simpler questions and more complex questions, as well as a good mix of math and concepts. I also feel the same way about the lab with some more difficult questions and some less difficult. The ESes were very nice and managed the event very well.

Biophysics: This test was very difficult. A 25/192 was a top 15 place and 18/192 was top 25. One thing I did notice was that to a certain degree there was calculus involved, however, it didnt matter much since it was a trial event.

Thermo: This event was not run as well as my others in my opinion. Although I definitely wouldn't say that it was run badly by any means, some of the instructions given were confusing or unclear. The test was fairly difficult, again with a few questions that I found the wording of odd.

[primitivepolonium]

"First off, there's a story with the room itself. I ran this event there was a real issue with that too when in the 2017 MIT Invite, some parents tried to break into the Forensics lab to take pictures of their kids (yeah, that doesn't go over well with university departments/administration)."

A quick correction--I believe the parents tried to break in the friday before to scope out the lab which is still pretty bad.