This memorandum is intended as a supplement to the Swisdak ACF Question-Writing Guide. All teams planning on submitting a packet for ACF-format nationals should be familiar with both documents, and should carefully observe all the rules they contain. However, in the event of a conflict between this document and the Swisdak guidelines, the supplement will control.
Few of the rules in here are absolute. Consistency between packets is absolutely necessary for any format, but preserving the individual style of different team-written packets is also important. With the exception of the category distributions (which are fairly loose to begin with) and the deadlines (which are not), minor deviations from these rules will be tolerated so long as they comply with the underlying spirit of the format.
This is a very preliminary document. Any suggestions or criticisms are very welcome.
1. No packet should contain more than one question from the same narrow subcategory of knowledge, unless both are parts of a single bonus. Narrowness should be understood with reference to the frequency with which such subcategories appear in academic competition: for example, Japanese literature is a narrow subcategory, but American literature is not. However, American transcendentalist literature would be a narrow subcategory, as would detective fiction. Along the same lines, it is fine to have multiple questions on Physics, but there should not be multiple questions on particle physics, electricity & magnetism, etc.
Note also that two questions can fit into different category distributions, and still be considered members of the same narrow subcategory. For example, a question on Greek mythology should not be included in the same packet with a question on the plot of Homer's _Iliad_.
2. Every packet should contain as diverse a range of subject matter as possible. No packet should fill its entire history requirement with military-political history, American history, or 20th-century history. Science questions should be balanced between the history of science and scientific concepts, as well as between theoretical and applied branches of science. The detailed distribution shows one possible means of meeting this diversity requirement.
3. Questions should be scattered evenly by category through the packet; all of the literature questions, for example, should not be in the beginning or the end. This is especially important for the bonus section, as not all 20 bonuses will necessarily be reached.
| Category | First 20 questions | Last 10 questions |
|---|---|---|
* At least one of the ten reserve questions should be either art/film or music.
** Should be from different categories. In addition, if there were five questions written on Literature, History, or Science already, neither of these miscellaneous questions should be used on another question from that category.
There will be gray areas in all of these categories; problems that arise from this are left to the writer's best judgment. In addition, many questions (particularly bonuses) will straddle more than one category. In these cases, the writer should assign the question to a category based on its dominant element.
No more than 2 history questions should be about military history; no more than 3 should be about political leaders.
No more than 3 of these should be about writers; no more than 4 should be about novels or novelists. Questions about King Arthur, Trojan war characters, etc. should be treated as literature if they relate to a specific text, but as myth if they do not.
Social science includes (but is not limited to) psychology, economics, anthropology, linguistics, sociology, and political science. In a combined packet (tossups and bonuses), each social science question should relate to a different discipline.
No more than 3 of these should be on the history of science or biographies of scientists.
A combined packet should not contain more than one question each on architecture and film. By contrast, it is permissible to fill this quota entirely with questions on visual art.
No packet should contain more than one "name the composer" tossup.
These should not be too historical.
This may include purely interdisciplinary questions, extra questions from the above categories, or questions that do not fit into any of the above categories.
Also, please remember that this exception is for current _history_, not current pop culture.
Some examples might help here:
1. ... Name this battle of July 1863, the bloodiest battle
of the Civil War.
Answer: _Gettysburg_
This is a knowable question for a novice team, as the giveaway clue is common knowledge to most educated Americans.
2. Name this novel about a three-year-old who refuses to
grow up, the most famous work of Gunther Grass.
Answer: The _Tin Drum_
This is probably neither knowable nor guessable for a novice team, as few new players will have heard of Gunther Grass. However, it is probably knowable by a median team.
3. ... Name this Eugene O'Neill tragedy, whose main
characters are Harry Hope and Theodore Hickman.
Answer: The _Iceman Cometh_
This is probably knowable by neither the novice nor the median team, since the contents of Eugene O'Neill plays are a bit less familiar to most of us than their titles. However, it is guessable by both, since almost everybody has at least heard of this work (even if they have no idea what it is actually about).
At least 80% of tossup giveaways should be knowable by the median team, and at least an additional 10% should be guessable.
It should be emphasized that, in the case of tossups, these are minima, not maxima. Hard tossups should never be written for the sake of writing hard tossups alone.
Substantive knowledge is preferable to fluff knowledge or trivia. Questions should try to favor players who actually understand a topic over those who have merely memorized a list of facts. For example, knowing the substance of Bertrand Russell's ideas is far more important than knowing the year in which he won the Nobel Prize, and a question on Bertrand Russell should reflect this.
ACF-format questions should be oriented towards the well- informed layperson, not the specialist. For example, questions about law should cover that material which a well-educated non- lawyer should know, and which she has a reasonable opportunity to learn without going to law school. Science questions may assume familiarity with introductory-level college course material, but more advanced topics should be included only to the extent that a layperson could learn about them through a magazine like "Scientific American." Questions that ask for technical jargon not easily learnable by a layperson should be avoided, regardless of the category.
Note that this is not a general limitation on difficulty-- only on certain types of difficulty. For example, a bonus part about an obscure composer like Henryk Gorecki is acceptable, since some classical music fans may know about him simply by virtue of being classical music fans. By contrast, a more technical question about counterpoint theory is less acceptable, even though it might be far easier than the Gorecki question to a person who is studying composition.
Questions should be entertaining. Considering the audience that you will be writing for, the easiest way to do this is to make them informative-- try to write your questions in such a way that they stick in people's minds, and make your facts memorable enough that even if your audience can't get a particular question right, they will learn and remember something new. Boring questions are often boring simply because the author doesn't try to tell a good story with his material.
Questions should not ask for "essay answers." This is simply too great a burden on the moderator.
For purposes of this section, tossups are divided into three parts: lead-in, development, and giveaway. The lead-in is the introductory information for the question; although it does not necessarily have to be a single sentence, it should not extend past the middle of the question's third line. The giveaway is the last piece of substantive information in the question, it should begin no earlier than the middle of the second-to-last line. The development is everything in between.
All lead-ins should contain information that is unique, substantive, and reasonably identifiable by a player with exceptional knowledge of the question's subject. The lead-in should also be difficult enough that it will not be answered by a player with only average knowledge of the subject.
Lead-ins should not contain "list" information. Examples of list-information include Nobel Prize years, atomic numbers, or birth years.
Where possible, lead-ins should contain a pronoun that signals the type of information that the answer is seeking. If considerations of style make a pronoun-flag impossible, the question should allow for reasonable alternate answers to be accepted if answered early.
Lead-ins should avoid dry recitations of biographical facts, especially where they give no reasonably identifiable information about the answer. In particular, lead-ins that begin, "Born in X, he was the son of a Y, and was educated at Z" should not be used.
Lead-ins should not encourage random guessing, and should not suggest one obvious answer to those with less-than-exceptional knowledge. For example, "This Norwegian playwright" is a terrible lead-in, since many players will (wisely) ring in at this point with "Ibsen," since he may be the only Norwegian playwright that they know.
Lead-ins should not be written with the intention of provoking incorrect buzzes. "His operas include _La Boheme_" is a bad lead-in if the answer is Leoncavallo (who wrote the lesser- known version); for other reasons, it is a bad-lead in even if the answer is Puccini.
To the greatest extent possible, the information contained in the transition should be a logical continuation of the story told in the lead-in. Transitions should not read like an abrupt non- sequitur.
Transitions should contain at least two distinct nuggets of information, from which a player might reasonably be expected to answer the question.
Transitions should be strictly pyramidal: transitional facts should be easier than the lead-in fact(s), and more difficult than the giveaway fact. Within the transition, facts should be arranged in strict descending order of difficulty.
Giveaway clues should be preceded by the words, "for ten points, ..." (or FTP, to use packet-writing shorthand).
Giveaway clues should be designed to make the question answerable to as many players as possible. "FTP, name this author of _Anna Karenina_" is a better giveaway clue than "FTP, name this author of _Resurrection_."
All bonuses should have a maximum of thirty points available.
All bonuses should award points in multiples of five.
There should be a general consistency in the difficulty of bonuses within the same packet.
Single-part, all-or-nothing bonuses should not be written.
Although bonuses would ideally test related pieces of information, the parts should not be so closely related that a multiple-part bonus effectively works out to an all-or-nothing. (That is, that knowledge of one part is so strongly correlated with knowledge of the other parts that any team playing it is likely to get full points, or no points).
No bonus should take longer than 45 seconds to complete. (Remember, under ACF rules, teams are entitled to 5 seconds per bonus part to prepare an answer).
Bonus introductions should be as short and to-the-point as possible.
All bonus questions should clearly indicate the point values assigned to each part.
Trash questions. As a general rule of thumb, any question that is appropriate for a trash tournament is inappropriate for an ACF tournament, just as any appropriate ACF question will usually be inappropriate for a trash tournament. (BTW, this is not a slam against trash players or trash questions-- only that we think that there is some virtue in treating trash as an entirely separate format). There is one extremely narrow exception to this rule: serious jazz and serious film questions might have a place in both formats, although this exception should not be extended to Kenny G or the filmography of Pauly Shore.
Straight-out-of-Benet's questions. Although Benet's is probably the outstanding question-writing source in print today, its wide popularity has caused some problems. In particular, questions should be written so that a person who actually has outside knowledge of writer X will have an advantage over the player who has merely memorized the Benet's entry for Writer X. As a result, every literature question should try to contain some information that cannot be found in Benet's. (Of course, subject to this rule, there is nothing wrong with continuing to use Benet's as a writing resource).
This same rule also applies to An Incomplete Education.
Spelling questions.
Questions that are entirely based on memorization of the Nobel Prizes, or of the Periodic Table.
Name-the-element questions of the form: "Discovered in 1878 by von Tronka and Kohlhaas, its alloys include Revulurium, Boorsteinium, and Edwardsium. Its atomic mass is 467.55, and its melting point is 683 degrees Celsius. FTP, name this element with atomic number 115, whose chemical symbol is Hp." There is nothing wrong with an interesting or substantive question whose answer happens to be a chemical element, however.
Question-writers are responsible for backing up the facts in their packets with reliable sources. We don't expect you to footnote your questions, but if we write back to you and ask for a source on a particular fact, you should be able to give us one.
Do not write questions from memory without double-checking your facts with a reliable source. Note that not all popular reference books are factually reliable. In particular, be careful of: The Book of Lists and all its sequels; anything else written by the authors of the Book of Lists, anything which markets itself as an imitation of the Book of Lists; An Incomplete Education (science sections particularly), and Ann Landers columns.