Lederberg 2 Discussion
- Sima Guang Hater
- Auron
- Posts: 1965
- Joined: Mon Feb 05, 2007 1:43 pm
- Location: Nashville, TN
Lederberg 2 Discussion
Thanks to my cowriters, Michael Hausinger, Saajid Moyen, Mike Cheyne, and Harrison Brown. Thanks also to Matt Bollinger, Auroni Gupta, Chris Manners, Ophir Lif, and anyone else who playtested and offered helpful comments. And especially thanks to Will Alston, Jacob Reed, Rob Carson, Ophir Lif, Tanay Kothari, Nick Wawrykow, and Michael Hausinger for giving up their Saturday to staff this tournament.
Feel free to discuss the tournament here. The packets have been submitted to the database and will be available shortly.
Feel free to discuss the tournament here. The packets have been submitted to the database and will be available shortly.
Eric Mukherjee, MD PhD
Brown 2009, Penn Med 2018
Instructor/Attending Physician/Postdoctoral Fellow, Vanderbilt University Medical Center
Coach, University School of Nashville
“The next generation will always surpass the previous one. It’s one of the never-ending cycles in life.”
Support the Stevens-Johnson Syndrome Foundation
Brown 2009, Penn Med 2018
Instructor/Attending Physician/Postdoctoral Fellow, Vanderbilt University Medical Center
Coach, University School of Nashville
“The next generation will always surpass the previous one. It’s one of the never-ending cycles in life.”
Support the Stevens-Johnson Syndrome Foundation
- Excelsior (smack)
- Rikku
- Posts: 386
- Joined: Sun Jan 25, 2009 12:20 am
- Location: Madison, WI
Re: Lederberg 2 Discussion
A+ tournament; would undergo mind-wipe and play again.
My brain was kind of fried by the time we started playing Lederberg, so I don't really have anything in the way of substantive commentary other than to say that I enjoyed this tournament a lot.
(also, I will pay you the dollars you are owed once my bank verifies on Venmo.)
My brain was kind of fried by the time we started playing Lederberg, so I don't really have anything in the way of substantive commentary other than to say that I enjoyed this tournament a lot.
(also, I will pay you the dollars you are owed once my bank verifies on Venmo.)
Ashvin Srivatsa
Corporate drone '?? | Yale University '14 | Sycamore High School (OH) '10
Corporate drone '?? | Yale University '14 | Sycamore High School (OH) '10
- Quantum Mushroom Billiard Hat
- Rikku
- Posts: 435
- Joined: Tue Jun 12, 2007 1:16 pm
- Location: Midland, MI
Re: Lederberg 2 Discussion
I had a lot of fun working on this set, and I hope everyone enjoyed it. Many thanks to Eric for letting me help to write it and everyone else who helped write/playtest the set as well. It's a lot of fun to learn about the reactions and molecules that show up while researching for this type of tournament, like hexaferrocenylbenzene (my new avatar), which was used as a clue in the aromaticity tossup.
Feel free to discuss anything that caught your attention, but I would like to know how a couple of my tossups played:
Feel free to discuss anything that caught your attention, but I would like to know how a couple of my tossups played:
- 1. The alcohols tossup in packet 3 was written without any names of reactions (other than I suppose hydroboration), and instead only used description of the chemistry going on. Is this a good thing, or did it just get confusing? I don't think it would be possible to put many tossups of this type in a tournament, since it overlaps with a lot of other answer space.
2. The ethanol tossup in packet 2 was constrained to clues on production of biofuel. This is an area where there is a lot of research, but did that type of clue selection limit the answers enough to make the tossup transparent?
Michael Hausinger
Coach, Bay City Western High School
formerly of University of Michigan and East Lansing High School
Coach, Bay City Western High School
formerly of University of Michigan and East Lansing High School
Re: Lederberg 2 Discussion
I had a great time playing the tournament, thanks for doing it.
-Seth
-Seth
Seth Teitler
Formerly UC Berkeley and U. Chicago
President of NAQT
Emeritus member of ACF
Formerly UC Berkeley and U. Chicago
President of NAQT
Emeritus member of ACF
- Fond du lac operon
- Wakka
- Posts: 228
- Joined: Tue Jan 31, 2012 8:02 pm
Re: Lederberg 2 Discussion
Thanks to Eric for allowing me to write for this, and special thanks for being so understanding when I had to back out of half my commitment when real life intervened. My questions that made it into the set were on: numerical integration, a math common link on "smooth", the heat equation (in math), the Borsuk-Ulam theorem, the gamma function, Niels Henrik Abel, Andrey Kolmogorov, and a common link on "planar." Comments on any of these are appreciated.
Harrison Brown
Centennial '08, Alabama '13
"No idea what [he's] talking about."
Centennial '08, Alabama '13
"No idea what [he's] talking about."
- The Ununtiable Twine
- Auron
- Posts: 1058
- Joined: Fri Feb 02, 2007 11:09 pm
- Location: Lafayette, LA
Re: Lederberg 2 Discussion
All were of high quality. Your choice of clues was appropriate for the difficulty level. Hopefully planar algebras become a thing in quizbowl.Fond du lac operon wrote:Thanks to Eric for allowing me to write for this, and special thanks for being so understanding when I had to back out of half my commitment when real life intervened. My questions that made it into the set were on: numerical integration, a math common link on "smooth", the heat equation (in math), the Borsuk-Ulam theorem, the gamma function, Niels Henrik Abel, Andrey Kolmogorov, and a common link on "planar." Comments on any of these are appreciated.
I demand for my 10 on the Watkins tossup be changed to 15. That's my only real complaint about this tournament. That and maybe that my prediction that Cheyne would write on Tomorrowland did not come true.
EDIT: silly spelling error oops
Last edited by The Ununtiable Twine on Mon Mar 31, 2014 1:15 am, edited 1 time in total.
Jake Sundberg
Louisiana, Alabama
retired
Louisiana, Alabama
retired
- Adventure Temple Trail
- Auron
- Posts: 2764
- Joined: Tue Jul 15, 2008 9:52 pm
Re: Lederberg 2 Discussion
I'll say as a non-science player that this was still a really fun tournament/experience despite my lack of specialty. The preponderance of easy/comprehensible answer lines and basic-description clues made it fun to follow along with and play even while getting sciencexterminated by much more capable opponents. Science fiction is cool, and fictional science is even cooler.
I noticed a rash of instances where repeated prompting resulted in people getting negged for answers that didn't quite make it to what the question wanted, even from quite capable players (e.g. I buzzed in and said "lipoproteins" upon hearing "LDL, VLDL, etc." and didn't know what to do from there; another player saying "preparing a gel for electrophoresis" for the tossup on PAGE). I can't be sure how much of this to attribute to the questions (i.e. whether there were instances where players had to "read the author's mind") and how much to the players (i.e. whether we just didn't know enough to make the fine-grained distinctions between related ideas which are required in science education and science quizbowl playing, or whether people were just being silly-aggressive at 1 AM, as can happen in one's 23rd round of quizbowl in a single day), so I guess I'm curious what more legitimately science-minded people thought about prompting and clue specificity and the like.
I noticed a rash of instances where repeated prompting resulted in people getting negged for answers that didn't quite make it to what the question wanted, even from quite capable players (e.g. I buzzed in and said "lipoproteins" upon hearing "LDL, VLDL, etc." and didn't know what to do from there; another player saying "preparing a gel for electrophoresis" for the tossup on PAGE). I can't be sure how much of this to attribute to the questions (i.e. whether there were instances where players had to "read the author's mind") and how much to the players (i.e. whether we just didn't know enough to make the fine-grained distinctions between related ideas which are required in science education and science quizbowl playing, or whether people were just being silly-aggressive at 1 AM, as can happen in one's 23rd round of quizbowl in a single day), so I guess I'm curious what more legitimately science-minded people thought about prompting and clue specificity and the like.
Matt Jackson
University of Chicago '24
Yale '14, Georgetown Day School '10
member emeritus, ACF
University of Chicago '24
Yale '14, Georgetown Day School '10
member emeritus, ACF
Re: Lederberg 2 Discussion
This tournament was fantastic and I had an excellent time playing it. As with all tournaments, there were a few clunkers here and there, but they didn't affect my overall enjoyment of the set very much.
Thanks to everyone for writing!
Thanks to everyone for writing!
Cody Voight, VCU ’14.
-
- Rikku
- Posts: 308
- Joined: Mon Aug 10, 2009 12:05 am
Re: Lederberg 2 Discussion
Signing up for both this and History Bowl (the weekend before a big exam day, no less) turned out to be not as good of an idea as I'd thought when this one didn't start until 8pm and didn't finish until past 1am... but Lederberg itself being super fun more than made up for that. Thanks very much to Eric and his co-writers/staffers for this rare opportunity to play a lot of science questions in one night!
Kenneth Lan, ASU '11, '12, UIC '17
The University of Illinois at Chicago
-stranger in a strange land (2013-)
The Sonoran Desert quizbowl ecosystem
-activist/advocate (2010-2013)
The Arizona State University Quizbowl Club
-elder statesman (2011-2013)
-coach (2009-2011)
-club president (2008-2011)
-founder (2007-)
The University of Illinois at Chicago
-stranger in a strange land (2013-)
The Sonoran Desert quizbowl ecosystem
-activist/advocate (2010-2013)
The Arizona State University Quizbowl Club
-elder statesman (2011-2013)
-coach (2009-2011)
-club president (2008-2011)
-founder (2007-)
- Excelsior (smack)
- Rikku
- Posts: 386
- Joined: Sun Jan 25, 2009 12:20 am
- Location: Madison, WI
Re: Lederberg 2 Discussion
I liked all of these questions (well, Jake buzzed on "planar" before anything I could comprehend came up, but I assume that was going to get good, too). I haven't re-looked at any of the questions, but I felt that the questions on numerical integration, the heat equation, and the gamma function did a particularly good job of giving me useful and suggestive context before they dropped a clue I knew for sure. (The same would have been true of the question on Abel if only I had realized that "abelian" is derived from "Abel"...)Fond du lac operon wrote:Thanks to Eric for allowing me to write for this, and special thanks for being so understanding when I had to back out of half my commitment when real life intervened. My questions that made it into the set were on: numerical integration, a math common link on "smooth", the heat equation (in math), the Borsuk-Ulam theorem, the gamma function, Niels Henrik Abel, Andrey Kolmogorov, and a common link on "planar." Comments on any of these are appreciated.
I will note that it offends my sense of quizbowl aesthetics to have common links on "THIS WORD" - e.g. "smooth", where it looks like there aren't any conceptual similarities between e.g. smooth numbers and smooth functions that lead to them getting that designation. That aside, that tossup was still pretty nifty.
Aside: did anybody buzz on the Watkins tossup before "grand vizier of fuckfaces"?
Ashvin Srivatsa
Corporate drone '?? | Yale University '14 | Sycamore High School (OH) '10
Corporate drone '?? | Yale University '14 | Sycamore High School (OH) '10
- Chimango Caracara
- Wakka
- Posts: 151
- Joined: Mon Jan 09, 2012 11:29 pm
- Location: Madison, WI
Re: Lederberg 2 Discussion
I thought there were a lot of cool answerlines in this tournament. I especially liked the questions on gene duplication, the TCR, calmodulin, hybrid sterility and C-H bonds.
Nick Jensen
UC Davis Microbiology Graduate Group 2017-2022
Dartmouth College 2011-2015
UC Davis Microbiology Graduate Group 2017-2022
Dartmouth College 2011-2015
- Fado Alexandrino
- Yuna
- Posts: 834
- Joined: Sat Jun 12, 2010 8:46 pm
- Location: Farhaven, Ontario
Re: Lederberg 2 Discussion
I got this tossup later; I didn't think the answer line could be this silly.Excelsior (smack) wrote: Aside: did anybody buzz on the Watkins tossup before "grand vizier of fuckfaces"?
I really enjoyed this tournament, 10/10 would play again. There was a good mix of harder clues with easy answer lines and answer lines (especially in the biology) that you normally wouldn't hear. Although, some of the easier answer lines (like the one on Java, or the one on salinity) that many people sat on and could not believe was so easy.
Thanks to Eric and all the writers, and all the staffers, who unlike the players, had to actually stay awake through the night to read for this event.
Joe Su, OCT
Lisgar 2012, McGill 2015, McGill 2019, Queen's 2020
Lisgar 2012, McGill 2015, McGill 2019, Queen's 2020
Re: Lederberg 2 Discussion
I wrote all of the trash questions. I wasn't really thinking of difficulty here, so a lot of them were pretty hard, so sorry about that. I also apologize to Andy Watkins for defaming his--ah, go fuck yourself.
Mike Cheyne
Formerly U of Minnesota
"You killed HSAPQ"--Matt Bollinger
Formerly U of Minnesota
"You killed HSAPQ"--Matt Bollinger
- The Ununtiable Twine
- Auron
- Posts: 1058
- Joined: Fri Feb 02, 2007 11:09 pm
- Location: Lafayette, LA
Re: Lederberg 2 Discussion
Hmmm, in looking at the results there's no evidence of my anticlimactic tournament-winning tossup as it seems I have been credited with 0 0 0 in that game.
Jake Sundberg
Louisiana, Alabama
retired
Louisiana, Alabama
retired
- Auks Ran Ova
- Forums Staff: Chief Administrator
- Posts: 4295
- Joined: Sun Apr 30, 2006 10:28 pm
- Location: Minneapolis
- Contact:
Re: Lederberg 2 Discussion
I did, in playtesting!Excelsior (smack) wrote:Aside: did anybody buzz on the Watkins tossup before "grand vizier of fuckfaces"?
Rob Carson
University of Minnesota '11, MCTC '??, BHSU forever
Member, ACF
Member emeritus, PACE
Writer and Editor, NAQT
University of Minnesota '11, MCTC '??, BHSU forever
Member, ACF
Member emeritus, PACE
Writer and Editor, NAQT
Re: Lederberg 2 Discussion
Fun tournament. I personally felt that the CS was problematic at times - for example that Java question wasn't ideal, and that shading tossup is kind of like a color change kecleon, for every new clue that was introduced the answer line changed. If Saajid, Eric, or friends would like substantive critiques on CS, just ask and I'll post it here or email you, otherwise, I don't want to clog this thread.
Ike
UIUC 13
UIUC 13
- Sima Guang Hater
- Auron
- Posts: 1965
- Joined: Mon Feb 05, 2007 1:43 pm
- Location: Nashville, TN
Re: Lederberg 2 Discussion
I'd be happy to hear it.Ike wrote:Fun tournament. I personally felt that the CS was problematic at times - for example that Java question wasn't ideal, and that shading tossup is kind of like a color change kecleon, for every new clue that was introduced the answer line changed. If Saajid, Eric, or friends would like substantive critiques on CS, just ask and I'll post it here or email you, otherwise, I don't want to clog this thread.
Eric Mukherjee, MD PhD
Brown 2009, Penn Med 2018
Instructor/Attending Physician/Postdoctoral Fellow, Vanderbilt University Medical Center
Coach, University School of Nashville
“The next generation will always surpass the previous one. It’s one of the never-ending cycles in life.”
Support the Stevens-Johnson Syndrome Foundation
Brown 2009, Penn Med 2018
Instructor/Attending Physician/Postdoctoral Fellow, Vanderbilt University Medical Center
Coach, University School of Nashville
“The next generation will always surpass the previous one. It’s one of the never-ending cycles in life.”
Support the Stevens-Johnson Syndrome Foundation
Re: Lederberg 2 Discussion
Okay here are the individual critique of questions of CS that I said I would give. I’m posting it here, because it could be the case I’m just providing an anomalous view on some of the questions, if anyone thought some of the questions played horribly but I said they played well or vice versa, feel free to chime in. Also, WARNING NERDS ahead.
Round 1:
Stephen Kleene - This is a perfectly fine tossup on Kleene that covers important material. Nitpick: The clue about the self-loop is not technically correct if you’re talking about a self-loop to one state; the entire DFA must have a self-loop to return an instance of an entire string being Kleene starred as opposed to a Kleene starred single alphabetical symbol.
cache - Problematic. The first clue is fine. The clue about the stampede is unfortunate, as it doesn’t specifically apply to caches; it happens a lot in something called memcache, which is not a cache as you think of it, but a program you run on servers to prevent databases from repeatedly serving a request over and over using cache-like principles, which caused me to think “oh these must be virtual pages of some sort” because pages are loaded into memcache when I worked with it. No one associates memcache with caches. Also, that LRU and MRU clue isn’t very good - as that is also how you replace pages. The rest of the tossup looks fine though.
Round 2:
Machine Learning - I don’t know a huge amount about machine learning, but this tossup seems okay. I question dropping “neural networks” with 2 lines to go, as if you don’t know what neural networks are, you aren’t going to know what machine learning is. Other than that, a solid tossup.
Java - This tossup isn’t good on a practical scale. I don’t know anything about the PhantomReference, but that may be an okay clue. However, that clue about enum isn’t good, since a lot of languages have enums, meaning that you’re expecting a player to know that enums were introduced in specifically Java 5, which is a ridiculous thing to expect. With regards to the next clue, C++ just got the ability to implement closures as well, so that clue about closures is unfortunately non-unique. Also, this tossup then proceeds to drop Swing, which from my understanding, was where pretty much everyone buzzed, because that’s the first thing in the tossup that engineers who aren’t Java gods can pinpoint to Java. This tossup pretty much exemplifies the problem with most programming language tossups…which is that its very hard to uniquely specify what people may potentially know, because very few people ever know one particular language so well or intermediately well that you can produce a good distribution of buzzes.
Round 3
trees - This tossup is pretty good. Everything seems to be solid, except for some of the language used in the question. No one is able to figure out what you are talking about in your description of the recursive definition of the tree. Also, tree rotations are simply used to maintain the height balance of the tree, there is no need to wrap it up in coy language, such as “maintain the invariants,” why not just say “the height of them can be reduced using rotations” or something similar.
polynomial factorization - is this the CS? This seems like a fine idea, though I didn’t cover any of this material in class, and am unable to pass judgment on this topic.
Round 4
shading: Not good. To start, I have to say that shaders aren’t used only for shading. When shaders were first developed, they were used explicitly for shading, hence their name, but now that is no longer the case since they turned out to be incredibly useful for a whole multitude of tasks. Assuming that engineers will associate all shading languages with “shading” exclusively is kind of like asking a musicologist to associate the violin with Paganini exclusively. What I’m saying on a quizbowl practical level is that you can’t give some vague clues about a shading language and say “this task” and people will jump to shading. The leadin isn’t very good, what you’re trying to say is that OPENGL has just included new tessellation evaluation and tessellation control shaders, why not just explicitly state that fact instead of giving a coy enum? I had a hard time figuring out what you’re trying to say. The next clue is a hose for modeling reflection, since the Lambert model is a model of reflection. The problem with this clue is that you don’t typically do any shading while building a Lambert model; these are two completely different parts of the graphics pipeline; you don’t combine shading with diffuse reflection, you performing shading after you model the Lambertian reflection, if you’re going to render your scene at all. I can see numerous people buzzing in here and saying “modeling reflection” or potentially “ray tracing” on that clue.. The next clue about Renderman isn’t good, as the answer is now “image synthesis,” even an answer of “rendering” here is actually correct plausible, since that is what Renderman does. The descriptions of Gouraud and Phong shading are not entirely correct, but I can’t imagine anyone complaining about that hampering them from answering the tossup; however, that clue about Phong shading you are describing can also apply to Phong illumination or Phong reflection, which makes use of Phong shading, so the name drop of Phong isn’t the best clue in the world either. The last few lines, however point to shading explicitly, and points to shading explicitly.
DFS - What is that first sentence trying to say? Also, I’m not sure what algorithm you are actually describing for finding SCCs - why not just state Robert Tarjan used this algorithm as part of an algorithm to find a graph’s SCCs? Also, it is possible to implement BFS recursively as well.
Round 5
O(n). A fine tossup. I do enjoy unrolling recurrences and solving them. The rest of the clues are solid and correct.
linear programming - Calling linear programming a field is misleading. No one goes out and says “I’m going to become a linear programmer;” it’s really a technique. That confused the hell out of me. Other than that, the cluing is solid.
Round 6
string searching - This tossup is solid - I don’t have any substantive commentary to offer though.
compilers - I think this tossup is a bit antipyramidal. The first few clues drop the names of programming languages which makes the answer obvious. It then proceeds to the relatively obscure Sethi-Ullman algorithm, before just dropping abstract syntax trees. Saying scripting languages generally do not include these tools is confusing at best, as there are many things people think of when they hear scripting languages, but the fact they lack a compiler is not the first or second thing they will think. Nitpick: the pronoun in this tossup switches from adjective (this type) to noun, (These tools) to verb (they can use tools like PyPy to do this.) That’s just flat-out confusing.
Round 7
design patterns - I think this tossup is a bit easy, though I freely admit that I have been engaged in the world of software engineering now and this material is second nature to me. I think the clueing itself is fine
convex hull - This tossup is mostly good. I wouldn’t put in that clue about collision detection into the tossup. There are literally dozens of different constructs you can use to make collision detection easier / possible, no one is going to say convex hull on that clue, or even think convex hulls. If you absolutely have to use that clue, why not explicitly state why you a convex hull is computed for collision detection?
Round 8
Multiplication - This tossup is solid, like the string searching tossup. I'm going to use this tossup to just point out how vanilla some of this tournament's CS was. Is anyone really that much of an expert on Karatsuba's algorithm or the Schonhagen-Strassen algorithm that they really felt awesome about this tossup? Even if they were, I just think there were so many questions on these relatively boring algorithms that no one who isn't a theoretician is going to be giddy about. A couple of questions like these are fine, but I think you can spend some time writing questions that will hit other software engineer's wheel.
stable marriage - Is this CS? I am not sure, I think this tossup was okay, though I was not able to recall its name, nor do I know much about this, so I can’t provide much commentary.
Round 9
error correction - This tossup is solid. I’m amused by this quantum error correction procedure.
Emil Post - This tossup seems to get to the point pretty quickly. Is there not something else that people know Emil Post for besides the PCP? I think tossing up Post is fine for a tournament, but it would help to provide another set of clues before just providing what the PCP actually is.
Round 10:
cryptographic keys - This tossup would be okay if only engineers and quizbowlers weren’t playing the tossup, since the Diffie-Hellman protocol has been beaten into the canon. The rest of the clueing is good though.
permutations of a set - How are you supposed to arrive at the word permutation if you haven’t encountered those algorithms in a math context? This answer and prompt lines should be expanded to include all possible things a player might say. A lot of players who hear the Knuth shuffle clue will arrive at something other than the word “permutation.”
Last impression: I think for at a tournament at this level, it is okay to include tossups on the other subfields of CS, including networking / distributed systems, the formal theory of program verification, etc, instead of having so many tossups on fundamental algorithms or basic software engineering such as compilers, Java, and patterns. There is a whole bevy of information that people can potentially know at this level, and I wished this tournament waded into more experimental areas.
- Ike
Round 1:
Stephen Kleene - This is a perfectly fine tossup on Kleene that covers important material. Nitpick: The clue about the self-loop is not technically correct if you’re talking about a self-loop to one state; the entire DFA must have a self-loop to return an instance of an entire string being Kleene starred as opposed to a Kleene starred single alphabetical symbol.
cache - Problematic. The first clue is fine. The clue about the stampede is unfortunate, as it doesn’t specifically apply to caches; it happens a lot in something called memcache, which is not a cache as you think of it, but a program you run on servers to prevent databases from repeatedly serving a request over and over using cache-like principles, which caused me to think “oh these must be virtual pages of some sort” because pages are loaded into memcache when I worked with it. No one associates memcache with caches. Also, that LRU and MRU clue isn’t very good - as that is also how you replace pages. The rest of the tossup looks fine though.
Round 2:
Machine Learning - I don’t know a huge amount about machine learning, but this tossup seems okay. I question dropping “neural networks” with 2 lines to go, as if you don’t know what neural networks are, you aren’t going to know what machine learning is. Other than that, a solid tossup.
Java - This tossup isn’t good on a practical scale. I don’t know anything about the PhantomReference, but that may be an okay clue. However, that clue about enum isn’t good, since a lot of languages have enums, meaning that you’re expecting a player to know that enums were introduced in specifically Java 5, which is a ridiculous thing to expect. With regards to the next clue, C++ just got the ability to implement closures as well, so that clue about closures is unfortunately non-unique. Also, this tossup then proceeds to drop Swing, which from my understanding, was where pretty much everyone buzzed, because that’s the first thing in the tossup that engineers who aren’t Java gods can pinpoint to Java. This tossup pretty much exemplifies the problem with most programming language tossups…which is that its very hard to uniquely specify what people may potentially know, because very few people ever know one particular language so well or intermediately well that you can produce a good distribution of buzzes.
Round 3
trees - This tossup is pretty good. Everything seems to be solid, except for some of the language used in the question. No one is able to figure out what you are talking about in your description of the recursive definition of the tree. Also, tree rotations are simply used to maintain the height balance of the tree, there is no need to wrap it up in coy language, such as “maintain the invariants,” why not just say “the height of them can be reduced using rotations” or something similar.
polynomial factorization - is this the CS? This seems like a fine idea, though I didn’t cover any of this material in class, and am unable to pass judgment on this topic.
Round 4
shading: Not good. To start, I have to say that shaders aren’t used only for shading. When shaders were first developed, they were used explicitly for shading, hence their name, but now that is no longer the case since they turned out to be incredibly useful for a whole multitude of tasks. Assuming that engineers will associate all shading languages with “shading” exclusively is kind of like asking a musicologist to associate the violin with Paganini exclusively. What I’m saying on a quizbowl practical level is that you can’t give some vague clues about a shading language and say “this task” and people will jump to shading. The leadin isn’t very good, what you’re trying to say is that OPENGL has just included new tessellation evaluation and tessellation control shaders, why not just explicitly state that fact instead of giving a coy enum? I had a hard time figuring out what you’re trying to say. The next clue is a hose for modeling reflection, since the Lambert model is a model of reflection. The problem with this clue is that you don’t typically do any shading while building a Lambert model; these are two completely different parts of the graphics pipeline; you don’t combine shading with diffuse reflection, you performing shading after you model the Lambertian reflection, if you’re going to render your scene at all. I can see numerous people buzzing in here and saying “modeling reflection” or potentially “ray tracing” on that clue.. The next clue about Renderman isn’t good, as the answer is now “image synthesis,” even an answer of “rendering” here is actually correct plausible, since that is what Renderman does. The descriptions of Gouraud and Phong shading are not entirely correct, but I can’t imagine anyone complaining about that hampering them from answering the tossup; however, that clue about Phong shading you are describing can also apply to Phong illumination or Phong reflection, which makes use of Phong shading, so the name drop of Phong isn’t the best clue in the world either. The last few lines, however point to shading explicitly, and points to shading explicitly.
DFS - What is that first sentence trying to say? Also, I’m not sure what algorithm you are actually describing for finding SCCs - why not just state Robert Tarjan used this algorithm as part of an algorithm to find a graph’s SCCs? Also, it is possible to implement BFS recursively as well.
Round 5
O(n). A fine tossup. I do enjoy unrolling recurrences and solving them. The rest of the clues are solid and correct.
linear programming - Calling linear programming a field is misleading. No one goes out and says “I’m going to become a linear programmer;” it’s really a technique. That confused the hell out of me. Other than that, the cluing is solid.
Round 6
string searching - This tossup is solid - I don’t have any substantive commentary to offer though.
compilers - I think this tossup is a bit antipyramidal. The first few clues drop the names of programming languages which makes the answer obvious. It then proceeds to the relatively obscure Sethi-Ullman algorithm, before just dropping abstract syntax trees. Saying scripting languages generally do not include these tools is confusing at best, as there are many things people think of when they hear scripting languages, but the fact they lack a compiler is not the first or second thing they will think. Nitpick: the pronoun in this tossup switches from adjective (this type) to noun, (These tools) to verb (they can use tools like PyPy to do this.) That’s just flat-out confusing.
Round 7
design patterns - I think this tossup is a bit easy, though I freely admit that I have been engaged in the world of software engineering now and this material is second nature to me. I think the clueing itself is fine
convex hull - This tossup is mostly good. I wouldn’t put in that clue about collision detection into the tossup. There are literally dozens of different constructs you can use to make collision detection easier / possible, no one is going to say convex hull on that clue, or even think convex hulls. If you absolutely have to use that clue, why not explicitly state why you a convex hull is computed for collision detection?
Round 8
Multiplication - This tossup is solid, like the string searching tossup. I'm going to use this tossup to just point out how vanilla some of this tournament's CS was. Is anyone really that much of an expert on Karatsuba's algorithm or the Schonhagen-Strassen algorithm that they really felt awesome about this tossup? Even if they were, I just think there were so many questions on these relatively boring algorithms that no one who isn't a theoretician is going to be giddy about. A couple of questions like these are fine, but I think you can spend some time writing questions that will hit other software engineer's wheel.
stable marriage - Is this CS? I am not sure, I think this tossup was okay, though I was not able to recall its name, nor do I know much about this, so I can’t provide much commentary.
Round 9
error correction - This tossup is solid. I’m amused by this quantum error correction procedure.
Emil Post - This tossup seems to get to the point pretty quickly. Is there not something else that people know Emil Post for besides the PCP? I think tossing up Post is fine for a tournament, but it would help to provide another set of clues before just providing what the PCP actually is.
Round 10:
cryptographic keys - This tossup would be okay if only engineers and quizbowlers weren’t playing the tossup, since the Diffie-Hellman protocol has been beaten into the canon. The rest of the clueing is good though.
permutations of a set - How are you supposed to arrive at the word permutation if you haven’t encountered those algorithms in a math context? This answer and prompt lines should be expanded to include all possible things a player might say. A lot of players who hear the Knuth shuffle clue will arrive at something other than the word “permutation.”
Last impression: I think for at a tournament at this level, it is okay to include tossups on the other subfields of CS, including networking / distributed systems, the formal theory of program verification, etc, instead of having so many tossups on fundamental algorithms or basic software engineering such as compilers, Java, and patterns. There is a whole bevy of information that people can potentially know at this level, and I wished this tournament waded into more experimental areas.
- Ike
Ike
UIUC 13
UIUC 13
Re: Lederberg 2 Discussion
I think if I were to go away from the single question commentary and talk about the CS as a whole, my most substantive comment would be that I sometimes found the pronouns imprecise or confusing.
For example, I learned about the VC dimension (lead-in to the machine learning tossup) wrt neural networks & [linear] classifiers. While it's true to say that classification algorithms are an example of machine learning algorithms, there's nothing in the tossup that really tells me to buzz with machine learning over classification. Now, if you instead said something like "algorithms in this field" [this is just me spit-balling, I'm not going to guarantee it would work with the question], this would probably change the way I thought about the clue.
Another example, the linear programming question (sidenote: I went to look up the lead-in, and Wikipedia claims it is a result in nonlinear programming which would seem to be...wrong). As Ike notes, linear programming is less a field and more a collection of methods to solve a certain set of problems. Which certainly sounds like a field, I'll admit, but I don't think anyone conceptualizes it as such.
Another example, keys. I didn't buzz on Diffie-Hellman (i.e. before "exchange of these tools") (well, I didn't buzz at all, actually, but point stands) because I was (1) astonished it was there (2) could not figure out what "tools" meant. Granted, I played this packet after 1am, so my processing faculties were severely deficient, but I still don't think most people would connect "tool" to "cryptographic key". A better choice would've been "things", which is a bit vague but does fit the answer precisely.
Lastly, I strongly agree with:
For example, I learned about the VC dimension (lead-in to the machine learning tossup) wrt neural networks & [linear] classifiers. While it's true to say that classification algorithms are an example of machine learning algorithms, there's nothing in the tossup that really tells me to buzz with machine learning over classification. Now, if you instead said something like "algorithms in this field" [this is just me spit-balling, I'm not going to guarantee it would work with the question], this would probably change the way I thought about the clue.
Another example, the linear programming question (sidenote: I went to look up the lead-in, and Wikipedia claims it is a result in nonlinear programming which would seem to be...wrong). As Ike notes, linear programming is less a field and more a collection of methods to solve a certain set of problems. Which certainly sounds like a field, I'll admit, but I don't think anyone conceptualizes it as such.
Another example, keys. I didn't buzz on Diffie-Hellman (i.e. before "exchange of these tools") (well, I didn't buzz at all, actually, but point stands) because I was (1) astonished it was there (2) could not figure out what "tools" meant. Granted, I played this packet after 1am, so my processing faculties were severely deficient, but I still don't think most people would connect "tool" to "cryptographic key". A better choice would've been "things", which is a bit vague but does fit the answer precisely.
Lastly, I strongly agree with:
Ike wrote:Last impression: I think for at a tournament at this level, it is okay to include tossups on the other subfields of CS, including networking / distributed systems, the formal theory of program verification, etc, instead of having so many tossups on fundamental algorithms or basic software engineering such as compilers, Java, and patterns. There is a whole bevy of information that people can potentially know at this level, and I wished this tournament waded into more experimental areas.
Cody Voight, VCU ’14.
- Excelsior (smack)
- Rikku
- Posts: 386
- Joined: Sun Jan 25, 2009 12:20 am
- Location: Madison, WI
Re: Lederberg 2 Discussion
Adding some brief thoughts onto what Ike said:
Machine learning - pretty solid.
Java - concur with Ike that the first half of the tossup is kind of meh. As Ike points out, tossups on individual programming languages are not the greatest idea.
polynomial factorization - the bits and pieces of this that I'm familiar with look pretty solid. Nice idea for a tossup.
O(n) - great tossup, and I do appreciate the clarificatory note at the beginning.
compilers - agree with Ike's criticisms. More generally, I am increasingly convinced that one can't really write a good question on "compilers".
stable matching - very interesting idea; seems to be implemented well, though I won't claim to be an expert on any sort of marriage.
error correction - solid tossup.
While I agree that the CS could have been more "exciting", it was perfectly solid nonetheless (modulo Ike's criticisms), so kudos to Saajid/Eric/whoever else.
Machine learning - pretty solid.
Java - concur with Ike that the first half of the tossup is kind of meh. As Ike points out, tossups on individual programming languages are not the greatest idea.
polynomial factorization - the bits and pieces of this that I'm familiar with look pretty solid. Nice idea for a tossup.
O(n) - great tossup, and I do appreciate the clarificatory note at the beginning.
compilers - agree with Ike's criticisms. More generally, I am increasingly convinced that one can't really write a good question on "compilers".
stable matching - very interesting idea; seems to be implemented well, though I won't claim to be an expert on any sort of marriage.
error correction - solid tossup.
While I agree that the CS could have been more "exciting", it was perfectly solid nonetheless (modulo Ike's criticisms), so kudos to Saajid/Eric/whoever else.
Ashvin Srivatsa
Corporate drone '?? | Yale University '14 | Sycamore High School (OH) '10
Corporate drone '?? | Yale University '14 | Sycamore High School (OH) '10
Re: Lederberg 2 Discussion
The set was pretty great in general. But, I have some complaints.
#1) The subdistributions at least for physics and math were heavily skewed.
It seemed to heavily focus on stat mech/solid state and QFT in the physics and algebra in the math. The one semi-analysis question (smooth) was even transformed into an algebra problem with its early clues. There were exactly 0 questions on mechanics, which does not seem good.
#2) Some of the clues were incredibly not helpful. I'll give a few examples and then explain what I think the reason behind them is:
I think that all the examples I talked about show a general problem in quizbowl writing that I've only now thought about. Questions are written from the perspective of a quizbowler mining clues. Clues utilized are often myopic in nature by not considering the fact that many techniques you may talk about in clues are used in contexts other than in your question. Even though Lederberg did in general avoid these problems, it too was unable to avoid this overly narrow focus that made many clues very unhelpful as per my examples above.
#1) The subdistributions at least for physics and math were heavily skewed.
It seemed to heavily focus on stat mech/solid state and QFT in the physics and algebra in the math. The one semi-analysis question (smooth) was even transformed into an algebra problem with its early clues. There were exactly 0 questions on mechanics, which does not seem good.
#2) Some of the clues were incredibly not helpful. I'll give a few examples and then explain what I think the reason behind them is:
This first clue is not useful at all. Multiple scales is often considered a generalization of both boundary layer theory and WKB in addition to being a better perturbative theory. Furthermore people studying this would hardly find the clue useful. Instead a clue about how (why it was developed) it is a generalization of perturbation theory would have been highly helpful.Packet 9 wrote:6. One extension to this method defines a function Y in terms of two different values of t; that is the method of multiple scales.
ANSWER: perturbation theory
Again these clues are not useful because the first 2 clues in this case refer to multiple things. Both these things could be used in something like solving a differential equation see like here: http://ursa.as.arizona.edu/~rad/phys205/ode/node12.html.Packet 2 wrote:13. One approach to doing this involves constructing a truncation of the tensor product of a one-dimensional multilevel basis. That approach was introduced by Smolyak and is known as “sparse grids.” One class of methods for this involves subdividing an interval when two approximations of the output differ greatly; those methods are known as adaptive algorithms.
ANSWER: numerical integration [or obvious equivalents; accept numerical quadrature; accept numerical discretization before “approximations”; prompt on “integration” alone]
So, these clues are very bad because they apply to many different geophysical waves such as water waves. Example: http://oceanworld.tamu.edu/resources/oc ... r12_02.htmPacket 4 wrote:These entities can pushed into forced stationary modes caused by a meridonal velocity field. Equations governing them usually begin by considering the equation d-dt of quantity zeta plus f, all over h, is equal to zero, which is the principle of potential (*) vorticity conservation.
ANSWER: Rossby waves
While this clue is technically correct, it entirely misses the essence of what is the Langer correction. Instead it's basically a synopsis of the Wikipedia entry. It is never limited to spherically symmetric systems. I'll quote from one of my books (emphasis theirs): In 1935 Langer made the amazing observation that [the connection formulae and turning point approximation] may be replaced by a single formula which is uniformly valid.When using the WKB approximation on a system with this configuration, a constant of one-fourth is added to L times L plus 1 terms; that trick is named for Langer.
ANSWER: spherical [or sphere or spherically symmetric or radially symmetric; prompt on “shell”]
I think that all the examples I talked about show a general problem in quizbowl writing that I've only now thought about. Questions are written from the perspective of a quizbowler mining clues. Clues utilized are often myopic in nature by not considering the fact that many techniques you may talk about in clues are used in contexts other than in your question. Even though Lederberg did in general avoid these problems, it too was unable to avoid this overly narrow focus that made many clues very unhelpful as per my examples above.
Libo
Washington '14, Michigan '13, Troy High School '09
Washington '14, Michigan '13, Troy High School '09
Re: Lederberg 2 Discussion
This may be true of some questions at Lederberg but I think it is a bigger problem in other tournaments. You can expect me to make a post about this later in general, because I think pretty much every tournament where I bitch about the sciences suffers from this. To me, the ability to write in such a way is what differentiates a good science writer from a passable and / or bad science writer. (I'm looking at you, my least favorite science writers.) All I can say is that the best way to fix this problem is to source your phrasings and clues from textbooks and problem sets and not a more general source, since you'll know that some people will have thought about your clue in the capacity you describe, and not in some other format.I think that all the examples I talked about show a general problem in quizbowl writing that I've only now thought about. Questions are written from the perspective of a quizbowler mining clues. Clues utilized are often myopic in nature by not considering the fact that many techniques you may talk about in clues are used in contexts other than in your question. Even though Lederberg did in general avoid these problems, it too was unable to avoid this overly narrow focus that made many clues very unhelpful as per my examples above.
Ike
UIUC 13
UIUC 13
- The Ununtiable Twine
- Auron
- Posts: 1058
- Joined: Fri Feb 02, 2007 11:09 pm
- Location: Lafayette, LA
Re: Lederberg 2 Discussion
Harrison Brown write analysis? Pshaw!Ringil wrote:The one semi-analysis question (smooth) was even transformed into an algebra problem with its early clues.
Jake Sundberg
Louisiana, Alabama
retired
Louisiana, Alabama
retired
- Sima Guang Hater
- Auron
- Posts: 1965
- Joined: Mon Feb 05, 2007 1:43 pm
- Location: Nashville, TN
Re: Lederberg 2 Discussion
Finally. This love-in was getting tiring.Ringil wrote:The set was pretty great in general. But, I have some complaints.
I wrote three questions on mechanics (action, normal modes, rigid bodies). I've actually taken mechanics (the Thornton and Marion version, not like Physics C mechanics), and try to write about it whenever I can; I've found it particularly difficult to find more than ~6 answerlines, since there aren't very many named things in my experience, and those that have names have been done to death (Lagrangian, calculus of variations, Hamiltonian, the Coriolis force, Noether's theorem, etc). I also tried to balance optics, thermo/statmech, fluids, quantum, particle physics, E/M, relativity etc; I won't say I did a perfect job, but I tried to use the textbooks in each of these disciplines pretty liberally.Ringil wrote:#1) The subdistributions at least for physics and math were heavily skewed.
It seemed to heavily focus on stat mech/solid state and QFT in the physics and algebra in the math. The one semi-analysis question (smooth) was even transformed into an algebra problem with its early clues. There were exactly 0 questions on mechanics, which does not seem good.
The math, however, I will cop to. I know almost nothing about analysis, and originally had more analysis answers before the time crunch hit me/Harrison had to bail. Sorry about that.
http://www.damtp.cam.ac.uk/user/ps57/BenderOrszag.pdf page 544. The only place I've seen MSA used is the canonical example of solving the Duffing equation, and it seemed to me to be the thing to do when secular terms get all up in your perturbation expansion. I'm happy to be corrected on this point, and acknowledge I probably could have written that clue better (though once I use the word "term" it becomes much more transparent of a question).Ringil wrote:#2) Some of the clues were incredibly not helpful. I'll give a few examples and then explain what I think the reason behind them is:This first clue is not useful at all. Multiple scales is often considered a generalization of both boundary layer theory and WKB in addition to being a better perturbative theory. Furthermore people studying this would hardly find the clue useful. Instead a clue about how (why it was developed) it is a generalization of perturbation theory would have been highly helpful.Packet 9 wrote:6. One extension to this method defines a function Y in terms of two different values of t; that is the method of multiple scales.
ANSWER: perturbation theory
You're completely correct about these, mea culpa.Ringil wrote:Again these clues are not useful because the first 2 clues in this case refer to multiple things. Both these things could be used in something like solving a differential equation see like here: http://ursa.as.arizona.edu/~rad/phys205/ode/node12.html.Packet 2 wrote:13. One approach to doing this involves constructing a truncation of the tensor product of a one-dimensional multilevel basis. That approach was introduced by Smolyak and is known as “sparse grids.” One class of methods for this involves subdividing an interval when two approximations of the output differ greatly; those methods are known as adaptive algorithms.
ANSWER: numerical integration [or obvious equivalents; accept numerical quadrature; accept numerical discretization before “approximations”; prompt on “integration” alone]
So, these clues are very bad because they apply to many different geophysical waves such as water waves. Example: http://oceanworld.tamu.edu/resources/oc ... r12_02.htmPacket 4 wrote:These entities can pushed into forced stationary modes caused by a meridonal velocity field. Equations governing them usually begin by considering the equation d-dt of quantity zeta plus f, all over h, is equal to zero, which is the principle of potential (*) vorticity conservation.
ANSWER: Rossby waves
Is that single formula always adding a constant factor of 1/4 to the L(L+1) term? I can't seem to find any instances in which this method is used to solve anything other than the Schrodinger equation for a radially symmetric potential.Ringil wrote:While this clue is technically correct, it entirely misses the essence of what is the Langer correction. Instead it's basically a synopsis of the Wikipedia entry. It is never limited to spherically symmetric systems. I'll quote from one of my books (emphasis theirs): In 1935 Langer made the amazing observation that [the connection formulae and turning point approximation] may be replaced by a single formula which is uniformly valid.When using the WKB approximation on a system with this configuration, a constant of one-fourth is added to L times L plus 1 terms; that trick is named for Langer.
ANSWER: spherical [or sphere or spherically symmetric or radially symmetric; prompt on “shell”]
Yeah this is something I've complained about in the past (viewtopic.php?f=21&t=13405&p=247534&hil ... up#p247534) and I'm saddened to see that in at least two cases I failed to do it correctly (esp in the numerical integration one - that one popped out instantly when I searched).Ringil wrote:I think that all the examples I talked about show a general problem in quizbowl writing that I've only now thought about. Questions are written from the perspective of a quizbowler mining clues. Clues utilized are often myopic in nature by not considering the fact that many techniques you may talk about in clues are used in contexts other than in your question. Even though Lederberg did in general avoid these problems, it too was unable to avoid this overly narrow focus that made many clues very unhelpful as per my examples above.
Eric Mukherjee, MD PhD
Brown 2009, Penn Med 2018
Instructor/Attending Physician/Postdoctoral Fellow, Vanderbilt University Medical Center
Coach, University School of Nashville
“The next generation will always surpass the previous one. It’s one of the never-ending cycles in life.”
Support the Stevens-Johnson Syndrome Foundation
Brown 2009, Penn Med 2018
Instructor/Attending Physician/Postdoctoral Fellow, Vanderbilt University Medical Center
Coach, University School of Nashville
“The next generation will always surpass the previous one. It’s one of the never-ending cycles in life.”
Support the Stevens-Johnson Syndrome Foundation
Re: Lederberg 2 Discussion
I didn't see the question on rigid bodies, and I forgot the normal modes question. However, the questions on action while on answer line from mechanics didn't use substantial clues from mechanics. This is similar to writing a question on Alexander the Great based only on art clues and calling it history. I don't think just because a questions are hard to write means that we shouldn't write about important things. There are plenty of engineering things that could have been asked about as mechanics, perhaps even my infamous contact tossup.The Quest for the Historical Mukherjesus wrote:I wrote three questions on mechanics (action, normal modes, rigid bodies). I've actually taken mechanics (the Thornton and Marion version, not like Physics C mechanics), and try to write about it whenever I can; I've found it particularly difficult to find more than ~6 answerlines, since there aren't very many named things in my experience, and those that have names have been done to death (Lagrangian, calculus of variations, Hamiltonian, the Coriolis force, Noether's theorem, etc). I also tried to balance optics, thermo/statmech, fluids, quantum, particle physics, E/M, relativity etc; I won't say I did a perfect job, but I tried to use the textbooks in each of these disciplines pretty liberally.Ringil wrote:#1) The subdistributions at least for physics and math were heavily skewed.
It seemed to heavily focus on stat mech/solid state and QFT in the physics and algebra in the math. The one semi-analysis question (smooth) was even transformed into an algebra problem with its early clues. There were exactly 0 questions on mechanics, which does not seem good.
The math, however, I will cop to. I know almost nothing about analysis, and originally had more analysis answers before the time crunch hit me/Harrison had to bail. Sorry about that.
The question on action:
By my count there are two mechanics clues in that question: the classic orbit clue and Hamilton's principle function.Packet 1 wrote:In quantum field theory...In QED...In a scalar field theory...Feynman formulation
Yes, MSA is used in dealing with the Duffing equation. But quoting from B&O pg. 544: Multiple-scale analysis is a very general collection of perturbative techniques that embodies the ideas of both boundary-layer theory and WKB theory. In practice it can is often used in boundary layer things for example section 8.8 of Applied Asymptotic Analysis by Miller.The Quest for the Historical Mukherjesus wrote:http://www.damtp.cam.ac.uk/user/ps57/BenderOrszag.pdf page 544. The only place I've seen MSA used is the canonical example of solving the Duffing equation, and it seemed to me to be the thing to do when secular terms get all up in your perturbation expansion. I'm happy to be corrected on this point, and acknowledge I probably could have written that clue better (though once I use the word "term" it becomes much more transparent of a question).Ringil wrote:#2) Some of the clues were incredibly not helpful. I'll give a few examples and then explain what I think the reason behind them is:This first clue is not useful at all. Multiple scales is often considered a generalization of both boundary layer theory and WKB in addition to being a better perturbative theory. Furthermore people studying this would hardly find the clue useful. Instead a clue about how (why it was developed) it is a generalization of perturbation theory would have been highly helpful.Packet 9 wrote:6. One extension to this method defines a function Y in terms of two different values of t; that is the method of multiple scales.
ANSWER: perturbation theory
The clue could have instead been worded as such:
The naive application of this method can result in blowup as time goes to infinity despite the system having energy conserved; that can be dealt with by considering two time scales.
See B&O you linked above pg. 510 where they use it for a non-symmetric system as is more typically done.The Quest for the Historical Mukherjesus wrote:Is that single formula always adding a constant factor of 1/4 to the L(L+1) term? I can't seem to find any instances in which this method is used to solve anything other than the Schrodinger equation for a radially symmetric potential.Ringil wrote:While this clue is technically correct, it entirely misses the essence of what is the Langer correction. Instead it's basically a synopsis of the Wikipedia entry. It is never limited to spherically symmetric systems. I'll quote from one of my books (emphasis theirs): In 1935 Langer made the amazing observation that [the connection formulae and turning point approximation] may be replaced by a single formula which is uniformly valid.When using the WKB approximation on a system with this configuration, a constant of one-fourth is added to L times L plus 1 terms; that trick is named for Langer.
ANSWER: spherical [or sphere or spherically symmetric or radially symmetric; prompt on “shell”]
Well that's unfortunate.The Quest for the Historical Mukherjesus wrote:Yeah this is something I've complained about in the past (viewtopic.php?f=21&t=13405&p=247534&hil ... up#p247534) and I'm saddened to see that in at least two cases I failed to do it correctly (esp in the numerical integration one - that one popped out instantly when I searched).Ringil wrote:I think that all the examples I talked about show a general problem in quizbowl writing that I've only now thought about. Questions are written from the perspective of a quizbowler mining clues. Clues utilized are often myopic in nature by not considering the fact that many techniques you may talk about in clues are used in contexts other than in your question. Even though Lederberg did in general avoid these problems, it too was unable to avoid this overly narrow focus that made many clues very unhelpful as per my examples above.
Libo
Washington '14, Michigan '13, Troy High School '09
Washington '14, Michigan '13, Troy High School '09
- Sima Guang Hater
- Auron
- Posts: 1965
- Joined: Mon Feb 05, 2007 1:43 pm
- Location: Nashville, TN
Re: Lederberg 2 Discussion
Yeah this is definitely something I forgot about, and will be writing more out of in the future.Ringil wrote:There are plenty of engineering things that could have been asked about as mechanics, perhaps even my infamous contact tossup.
EDIT: Also, if there are other questions that suffered from this, please let me know, either by posting here or getting in contact with me some other way.
Eric Mukherjee, MD PhD
Brown 2009, Penn Med 2018
Instructor/Attending Physician/Postdoctoral Fellow, Vanderbilt University Medical Center
Coach, University School of Nashville
“The next generation will always surpass the previous one. It’s one of the never-ending cycles in life.”
Support the Stevens-Johnson Syndrome Foundation
Brown 2009, Penn Med 2018
Instructor/Attending Physician/Postdoctoral Fellow, Vanderbilt University Medical Center
Coach, University School of Nashville
“The next generation will always surpass the previous one. It’s one of the never-ending cycles in life.”
Support the Stevens-Johnson Syndrome Foundation
- grapesmoker
- Sin
- Posts: 6345
- Joined: Sat Oct 25, 2003 5:23 pm
- Location: NYC
- Contact:
Re: Lederberg 2 Discussion
I enjoyed this set; thanks to Eric for writing it and reading it online for us tonight. I'll have more substantive comments tomorrow, but for now, let me just go back to the complaints about the tossup on "machine learning" and say that I think, based on my grad level machine learning class from a year ago, that was the most "correct" answer. The VC dimension is a fundamental aspect of learning algorithms as such; that said, I'd probably accept any type of machine learning algorithm, since the VC theorem is omnipresent in all of them.
Jerry Vinokurov
ex-LJHS, ex-Berkeley, ex-Brown, sorta-ex-CMU
presently: John Jay College Economics
code ape, loud voice, general nuissance
ex-LJHS, ex-Berkeley, ex-Brown, sorta-ex-CMU
presently: John Jay College Economics
code ape, loud voice, general nuissance
-
- Lulu
- Posts: 68
- Joined: Sun Nov 04, 2012 10:15 am
Re: Lederberg 2 Discussion
I really enjoyed playing this set, thanks to Eric for writing and to Ramapriya for reading last night. Thanks also to Andrew and Shreyas for putting up with my spirited negging. Lots of interesting questions, and it was great to get to play against a new bunch of people- I hope I haven't damaged the UK's quizzing rep too much, despite getting pasted repeatedly by Steven Hines...
I just had a couple of gripes with the chemistry. The clue on agostic bonding in the tossup on C-H bonds isn't uniquely identifying as C-C bonds can also give rise to an agostic interaction, so that was a bit confusing. I attach this abstract from a paper by a guy at Oxford: http://pubs.rsc.org/en/content/articlel ... ivAbstract
Also, I really wish that iminium salts/iminium ions/just iminiums were an acceptable answer for that tossup on enamines. Although they are formally different groups, every interesting thing about enamine chemistry exists because they exist as iminiums at some point during the reaction. I think quite a few clues in that TU made reference to steps where the iminium salt was the actually interesting part. Just at the start, the cycle for the HPESW reaction shows an iminium intemediate (Fig 7 in this paper http://ac.els-cdn.com/S0040402001912426 ... deeeb5ebf5) which tautomerises to give the enamine, so there's a bit of confusion there. Admittedly, I may just have been confused because I was read this question at 2 in the morning...
Most of it, however, was great fun, despite the fact that I was basically asleep by round 8. Bloody time difference. Still, would definitely be up for playing another interation next year!
EDIT: Also, for the record, in the chromium tossup, the first Fischer carbene had it bonded to five carbonyl groups, not four. Didn't stop me from negging with tungsten at that point anyway... I think the Chromium one might have been the first one to be characterised fully, but I could have sworn that the tungsten carbene was prepared first.
I just had a couple of gripes with the chemistry. The clue on agostic bonding in the tossup on C-H bonds isn't uniquely identifying as C-C bonds can also give rise to an agostic interaction, so that was a bit confusing. I attach this abstract from a paper by a guy at Oxford: http://pubs.rsc.org/en/content/articlel ... ivAbstract
Also, I really wish that iminium salts/iminium ions/just iminiums were an acceptable answer for that tossup on enamines. Although they are formally different groups, every interesting thing about enamine chemistry exists because they exist as iminiums at some point during the reaction. I think quite a few clues in that TU made reference to steps where the iminium salt was the actually interesting part. Just at the start, the cycle for the HPESW reaction shows an iminium intemediate (Fig 7 in this paper http://ac.els-cdn.com/S0040402001912426 ... deeeb5ebf5) which tautomerises to give the enamine, so there's a bit of confusion there. Admittedly, I may just have been confused because I was read this question at 2 in the morning...
Most of it, however, was great fun, despite the fact that I was basically asleep by round 8. Bloody time difference. Still, would definitely be up for playing another interation next year!
EDIT: Also, for the record, in the chromium tossup, the first Fischer carbene had it bonded to five carbonyl groups, not four. Didn't stop me from negging with tungsten at that point anyway... I think the Chromium one might have been the first one to be characterised fully, but I could have sworn that the tungsten carbene was prepared first.
Last edited by Ewan MacAulay on Thu Apr 24, 2014 6:29 am, edited 2 times in total.
Ewan MacAulay
Oxford 2015
Cambridge 2018
Oxford 2015
Cambridge 2018
- Sima Guang Hater
- Auron
- Posts: 1965
- Joined: Mon Feb 05, 2007 1:43 pm
- Location: Nashville, TN
Re: Lederberg 2 Discussion
Fascinating; most of the textbooks I have say that it's solely C-H bonds (wikipedia does too). I learned something today. I think the question's still unique because the first half was entirely about C-H activation, though.Ewan MacAulay wrote:I just had a couple of gripes with the chemistry. The clue on agostic bonding in the tossup on C-H bonds isn't uniquely identifying as C-C bonds can also give rise to an agostic interaction, so that was a bit confusing. I attach this abstract from a paper by a guy at Oxford: http://pubs.rsc.org/en/content/articlel ... ivAbstract
Ewan MacAulay wrote:Also, I really wish that iminium salts/iminium ions/just iminiums were an acceptable answer for that tossup on enamines. Although they are formally different groups, every interesting thing about enamine chemistry exists because they exist as iminiums at some point during the reaction. I think quite a few clues in that TU made reference to steps where the iminium salt was the actually interesting part. Just at the start, the cycle for the HPESW reaction shows an iminium intemediate (Fig 7 in this paper http://ac.els-cdn.com/S0040402001912426 ... deeeb5ebf5) which tautomerises to give the enamine, so there's a bit of confusion there. Admittedly, I may just have been confused because I was read this question at 2 in the morning...
I understand why that question could have been confusing, but this is the List-Houk model (which is what the HPESW reaction clue was referencing):
The intermediate and the transition state are two different things; the transition state is an enamine, but the intermediate is an imine. All of the other stuff was named stuff about enamines (Stork, transiently produced electron-rich enamines, etc). I can understand the confusion (I've made the analogous mistake on tossups on 'enolates' that were well-written).
If you can find someone to write another one, I'll play it too!Ewan MacAulay wrote:Most of it, however, was great fun, despite the fact that I was basically asleep by round 8. Bloody time difference. Still, would definitely be up for playing another interation next year!
http://www2.chemistry.msu.edu/faculty/w ... rbenes.htm seems to suggest otherwise, but it wouldn't surprise me if you were right.EDIT: Also, for the record, in the chromium tossup, the first Fischer carbene had it bonded to five carbonyl groups, not four. Didn't stop me from negging with tungsten at that point anyway... I think the Chromium one might have been the first one to be characterised fully, but I could have sworn that the tungsten carbene was prepared first.
Eric Mukherjee, MD PhD
Brown 2009, Penn Med 2018
Instructor/Attending Physician/Postdoctoral Fellow, Vanderbilt University Medical Center
Coach, University School of Nashville
“The next generation will always surpass the previous one. It’s one of the never-ending cycles in life.”
Support the Stevens-Johnson Syndrome Foundation
Brown 2009, Penn Med 2018
Instructor/Attending Physician/Postdoctoral Fellow, Vanderbilt University Medical Center
Coach, University School of Nashville
“The next generation will always surpass the previous one. It’s one of the never-ending cycles in life.”
Support the Stevens-Johnson Syndrome Foundation
-
- Lulu
- Posts: 34
- Joined: Tue Jan 22, 2013 10:18 pm
- Location: Austin
Re: Lederberg 2 Discussion
I too really enjoyed this -- thanks to Eric and Ramapriya for setting this up and giving me a chance to play this. There were some very interesting clues and answerlines, and I'm looking forward to going through this set at a more leisurely pace.
Corin Wagen
Kealing '12
LASA '16
MIT '20
Kealing '12
LASA '16
MIT '20
- grapesmoker
- Sin
- Posts: 6345
- Joined: Sat Oct 25, 2003 5:23 pm
- Location: NYC
- Contact:
Re: Lederberg 2 Discussion
I don't think it's terribly worthwhile to get into a question-by-question breakdown here, because I thought most of the questions were pretty good. I agree with Ike's take on some of the CS stuff, especially that "shading" question (the clues on that were just confusing to me), but overall I was really happy with what I heard. I have on small complaint about this set as a whole, which is that I felt that the clues were overly full of the kind of eponymous "extensions of this theory/equation/whatever" stuff that is usually less than helpful. I picked one example here:
I don't want to say that this question is terrible or anything like that, but I think in the places that I'm pointing to, it could have really used some explanation of just what these names were supposed to designate. I felt like a fair number of questions had this problem of referencing relatively obscure secondary theories and not giving much detail about them. I'm more fine with that for this set than I'd be for something like Nationals, but I do think some of the named clues could have been eliminated in favor of a bit more explanatory detail.
Anyway, that's my small gripe. Everything else was cool and the answer selection was excellent.
I don't know the multiple scales method and that seems like a useful clue (with caveats that people have expressed above), but the explanation of what that method is makes very little sense to me. All I can tell from that is some function is expressed as a function of two different parameters, but that's not very helpful. The next clue tells me about Weinberg's chiral theory, which is in some way connected to QCD, but that's about it. A much more useful piece of information for those of us who know something about QCD but don't have "Weinberg + chiral = perturbation" in our heads would be to actually describe something about the regime in which perturbative QCD is valid; that's something that I think physics people could understand even without knowing about Weinberg's specific work. "Constructing an effective Lagrangian" is something that you do almost anywhere in field theory, so that's mostly filler. The next clue again asks you to know something very specific about the Rabi formula, but again doesn't describe any aspect of it that could give you a handle on what's going on. This is a standard exercise in most perturbation theory textbooks, but just from the name it's hard to pull exactly what is being asked. The next clue is the first one that I think is genuinely useful in terms of pointing to what's actually happening, but in the one place where a named thing would be well-used, the question doesn't say anything about the Dyson series decomposition, which is used for finding the propagator in TDPT. I buzzed on the next clue after that; the stuff that comes later seems fine to me.One extension to this method defines a function Y in terms of two different values of t; that is the method of multiple scales. Weinberg developed the “chiral” version of this method to apply it to nucleon-nucleon interactions in QCD, where it can be used to construct an effective Lagrangian. This approach can be used to find the Rabi formula and the rate of spontaneous emission when applied to a forced two-level system. The interaction representation has to be used alongside this method to calculate the time-propagator when the system is time-dependent. When applying this method to a (*) degenerate state, a re-diagonalization of a certain operator in terms of degenerate states is required to solve the vanishing denominator problem; that method is used when this method is applied to the linear Stark effect. This method involves splitting the Hamiltonian into a “bare” Hamiltonian and another term, V, and then expanding in terms of a power series of some small parameter. For 10 points, name this method of approximating solutions to physical systems by considering the effects of a small namesake external force on the eigenvalues of the Hamiltonian.
ANSWER: perturbation theory
I don't want to say that this question is terrible or anything like that, but I think in the places that I'm pointing to, it could have really used some explanation of just what these names were supposed to designate. I felt like a fair number of questions had this problem of referencing relatively obscure secondary theories and not giving much detail about them. I'm more fine with that for this set than I'd be for something like Nationals, but I do think some of the named clues could have been eliminated in favor of a bit more explanatory detail.
Anyway, that's my small gripe. Everything else was cool and the answer selection was excellent.
Jerry Vinokurov
ex-LJHS, ex-Berkeley, ex-Brown, sorta-ex-CMU
presently: John Jay College Economics
code ape, loud voice, general nuissance
ex-LJHS, ex-Berkeley, ex-Brown, sorta-ex-CMU
presently: John Jay College Economics
code ape, loud voice, general nuissance
Re: Lederberg 2 Discussion
I think the clues are fine, but I probably would have used an exclusionary clue as well. When I see LRU and MRU, I think of page replacement algorithms.Also, that LRU and MRU clue isn’t very good - as that is also how you replace pages. The rest of the tossup looks fine though.
Matt Hayes
Unaffiliated
Unaffiliated
Re: Lederberg 2 Discussion
I really enjoyed listening to this set, it really was a great experience to sit in and try to answer all these science. I really liked how all the questions really seemed cohesive and (to me) really painted a story about what was going on - not just stated facts about the answer, which really kept my interest even though I wasn't really answering much. I felt like I could follow what was going on in each of the answer lines based on the descriptions in the question, even though I had never heard of a good number of the answer lines.
Thanks to all the writers for the great experience.
Thanks to all the writers for the great experience.
Drayer the Slayer
currently unaffiliated
currently unaffiliated
- Sima Guang Hater
- Auron
- Posts: 1965
- Joined: Mon Feb 05, 2007 1:43 pm
- Location: Nashville, TN
Re: Lederberg 2 Discussion
I think this is a good illustration of how (not) to write evocative clues for a science question. I'm recalling my thought process for this question, and I think the issue was whenever I tried to describe a particular instance of perturbation theory, it always devolved into mentioning a series expansion or something, which I thought would be too fraudable; obviously in this case I overdid it.Jerry wrote:Stuff about perturbation theory
EDIT: Here's an attempt at rewriting that question (comments welcome/encouraged):
Weinberg developed the “chiral” version of this method to apply it to nucleon-nucleon interactions in QCD, where it is typically applied in the high-energy regime to predict the ratios of the products of electron-positron annihilation. The interaction representation has to be used alongside this method to calculate the time-propagator when the system is time-dependent, and in such systems this method is used to define the evolution operator in terms of the time-ordering operator via a Dyson series. When applying this method to a (*) degenerate state, a re-diagonalization of a certain operator in terms of degenerate states is required to solve the vanishing denominator problem; that method is used when this method is applied to the linear Stark effect. This method involves splitting the Hamiltonian into a “bare” Hamiltonian and another term, V, and then expanding in terms of a power series of some small parameter. For 10 points, name this method of approximating solutions to physical systems by considering the effects of a small namesake external force on the eigenvalues of the Hamiltonian.
ANSWER: perturbation theory
Eric Mukherjee, MD PhD
Brown 2009, Penn Med 2018
Instructor/Attending Physician/Postdoctoral Fellow, Vanderbilt University Medical Center
Coach, University School of Nashville
“The next generation will always surpass the previous one. It’s one of the never-ending cycles in life.”
Support the Stevens-Johnson Syndrome Foundation
Brown 2009, Penn Med 2018
Instructor/Attending Physician/Postdoctoral Fellow, Vanderbilt University Medical Center
Coach, University School of Nashville
“The next generation will always surpass the previous one. It’s one of the never-ending cycles in life.”
Support the Stevens-Johnson Syndrome Foundation