Academic Competition Foundation (ACF) Question Writing Guidelines

version = 1.7
date = 12 January 1995
author = Marc Swisdak
 
Acknowledgements
 
         These guidelines are originally based on the Question Writing 
Guidelines prepared by the Stanford College Bowl Club and edited by Pat
Matthews of the University of Pennsylvania, and the 10 Commandments of
Question Writing composed by Carol Guthrie for the 1993 Masters tournament.
Revisions have been made using Jim Dendy's question guidelines for the 
1994 ACF Nationals.  I would like to thank everyone who has made
suggestions/corrections.
 
Introduction
         
         By far, the most critical component of any tournament is the
quality of the questions.  Question quality, while ultimately the
responsibility of the tournament director, strongly depends on the packets
which are submitted.  As veteran tournament directors know, it is difficult
to have to rewrite a bad packet (especially one submitted at the last 
minute) while still taking care of the other details of running a 
tournament. This guide is designed to raise the quality of questions 
which are written and submitted to invitational (and specifically, ACF) 
tournaments.
  
 
                                Packets

General Notes

         Usually, teams write a new packet of questions for each
tournament.  While this may prove to be time-consuming during the school 
year, under no circumstances is it acceptable to use questions previously
submitted to another tournament.  Furthermore, most tournament directors 
are not omniscient, so the submitting team has the responsibility of making
sure every part of their packet is factually correct.  If there is any doubt 
about a particular question, it is best to leave it out of the packet
entirely.  
	The tournament director (TD) always has complete control over
the content of the submitted packets.  If the TD explicitly requests a 
specific distribution, then the submitting teams should follow that 
distribution, despite any philosophical differences they may have with 
it.  Finally, teams should make every effort to submit packets to 
tournaments by the required due date. Submitting late packets only result 
in a harder, and possibly sloppier, editing job by the tournament director.

Packet Length
 
         Since ACF games consist of 20 (and occasionally tie-breaking)
toss-ups, any submitted packet should contain at least 25 GOOD toss-ups. 
The extra toss-ups are crucial for three reasons.  First, tie-breaking 
toss-ups are occasionally needed; second, in case of similarities between 
separately submitted questions between rounds, the extra toss-ups provide
a safety margin for the TD; and third, the TD may not agree with your high
opinion of some of the first 20 toss-ups and choose to reorder the packet.
For similar reasons, a packet should also contain 25 GOOD bonuses.
 
Question Quality
 
        The purpose of writing ACF questions is not to show off the
question-writer's mastery of esoteric knowledge.  Instead, a balance must
be  struck between writing questions that realistically test the 
knowledge of the competitors and questions written to "stump the chump."
The ability to distinguish between what is reasonable and what isn't is 
best acquired by writing packets and having others play on and critique 
them.  Attending invitational tournaments also can help one get a feel 
for what is and is not an acceptable question.  It is usually difficult 
for someone who has just been introduced to the sport to write a good 
question packet.
 
Distribution
 
         The distribution of a packet is ultimately in the hands of the
TD.  If he or she requests a specific distribution, the attending teams
are obligated to follow those guidelines.  With that in mind, the following
question balance is that used at the 1994 ACF Nationals with Jim Dendy as 
TD.  It is offered as one suggestion of a balanced 30-question distribution

        Packets should be balanced with questions coming from a number of
different subject areas as shown below.   Moreover, questions should also be
balanced within disciplines, i.e. Social Science/Geography questions 
should not all be about geography.  While this distribution is not set in 
stone, changes should be kept to a minimum, unless the TD suggests otherwise.
 
Literature 
	English/American				4
	World/Ancient				1
	Miscellaneous				1
Science
	Physics/Astronomy/Engineering		2
	Biology/Earth Sciences			2
	Chemistry/Biochemistry			1
	Miscellaneous				1
History
	American					3
	European					2
	Miscellaneous				1
Fine Arts/Music/Film
	Classical Music/Dance			1
	Painting/Sculpture/Architecture		1
	Miscellaneous				1
Social Science/Geography				3
Religion/Mythology/Philosophy			3
Current Events/Sports/Entertainment		1
Miscellaneous
	(Usually from lit/science/history)	2

 
Literature
 
         Literature questions should cover a broad range of geographical
areas and chronological periods.  Authors, literary movements, characters,
plot summaries, and titles are all valid subjects for questions.  Note 
that literature is not confined to novels.  Drama, poetry, and short stories 
are all included here .  A number of good references for this category are
available at any bookstore. 
 
Science
 
         This subject covers everything from botany, zoology and other
biological sciences to chemistry, physics, and other physical sciences.
Care should be taken to achieve a balance between the different branches 
of science, as is done in the above breakdown.  Questions should generally
focus on the science itself instead of the scientists who made the 
discoveries - i.e, asking about the photoelectric effect instead of 
Einstein.  Questions of the latter type are more concerned with the 
history of science which only marginally falls under this heading.  
Sources for these types of questions include textbooks and various 
scientific journals aimed at the general public.
 
History
 
         As with literature, these questions should cover a large range of
geographical areas and chronological periods.  European and American 
history are the most common areas of focus, however. 

Fine Arts/Music/Film

         The fine arts include painting, sculpture, engraving, ballet, 
architecture, film and other similar categories.  Music questions do not 
have to be constrained to classical music, although the emphasis is generally 
on this area.  Other types of music from Gregorian chants to jazz are 
also viable questions sources. 

Social Sciences/Geography

         These questions focus on areas such as anthropology, economics,
sociology, and psychology which are traditionally separated from the
natural sciences.  As with the science category, question should be evenly
distributed within various subfields.  Geography questions are also included
here.
 
Religion/Mythology/Philosophy
 
         A variety of topics are included in this category.  Mythology
questions can range over any mythology (Babylonian, Celtic, American Indian)
but the mainstay of questions have been about Greek, Roman, and Norse myths.

Current Events/Sports/Entertainment
 
         Current events questions focus on recent news stories and events.
Daily newspapers and weekly news magazines are good sources for these
questions.  In keeping with the academic nature of ACF, questions of this 
sort should focus on items of long-term importance.  A general rule of 
thumb is that if the question would be unanswerable 1 year after it was 
written, it should probably not be included. 
	Sports and entertainment questions focus on popular culture
including music, fashion, television, and movies.  In keeping with the
academic nature of ACF, questions from this area are generally few in
number.


                                Toss-ups
 
Structure
 
         Toss-ups should be ordered like 30-20-10 bonuses (see below) --
the most difficult clues are given at the beginning of the toss-up while
the easier "give-away" clues are given nearer the end.  This rewards teams
with deeper knowledge, while still allowing teams that are not as familiar
with the topic to have a chance at answering the question.

        The first pronoun in a toss-up should always refer to the answer. 
Also, that pronoun should occur as close as possible to the beginning of 
the question.  This eliminates any ambiguity for the players and generally
improves the flow of the question.  Note that toss-ups should almost always
ask for one piece of information.  If two pieces of information are needed,
the question must begin with "2 answers are required," or an equivalent
phrase.

        To avoid monotony, toss-ups should not all follow the exact same 
blueprint.  For instance, a packet which has ten toss-ups starting "Born 
in [...] he went to school at [...] and later ..." is unacceptable.  
While some toss-ups of this type are perfectly acceptable, too many can make a
packet very dull.

         All toss-ups should include the magic words "for ten points"
(sometimes abbreviated FTP).  These words should appear IMMEDIATELY before
the last clue is given.  The reason for this convention is that allows 
the player to know when the question is completed, and thus when it it safe to
buzz without risking a penalty.
 
Ex.     ... for ten points, identify this author of _Great Expectations_.
 
         ... FTP, name this largest U.S. state.

         The answer to a toss-up should be clearly visible beneath the
question.  The pertinent part of the answer should be marked (usually
underlined).  Full names should be required only when there is the
possibility of confusion.  All alternate answers should be anticipated
and noted.
 
Ex.     Possible Answers       _William Harrison_
                 
                               _Benjamin Harrison_

				Thomas _Jefferson_
                         
                 _FDIC_ or (_Federal Deposit Insurance Corporation_)
 
Difficulty
 
         Deciding the difficulty of a toss-up can be challenging since
different people possess different bodies of knowledge.  However, a
general rule is that in a game between two medium-level teams, 80-90% of the
toss-ups should be answerable BY THE END OF THE QUESTION.  Notice that this
does not mean that everyone should be able to answer every question. 
However at least one member of a well-balanced team should usually have the
answer when the question finishes.
 
Misleading Toss-ups
 
        The following types of toss-ups should be avoided at all costs.
 
a. Toss-ups which are written in such a way that they encourage teams to
jump in early with the wrong answer.
 
Ex:  1, 1, 2, 3, 5, 8 are the first few terms of the Fibonacci series. 
FTP, spell Fibonacci.
 
b. Toss-ups which use pronouns incorrectly.  The first pronoun used in a 
question should always refer to the answer.  
 
Ex:  _His_ theorem stating that one cannot find a, b, c integers so that 
a^n + b^n = c^n for n>2 enabled _Fermat_ to go down in history...
        The question could be rewritten to start "Fermat's Theorem which
states that  ...)
 
c. Questions with irrelevant information.
 
Ex:  Paris is the capital of France, but FTP who killed Paris in the Trojan
War.  
  
Length
 
         While the ACF format does allow questions to be longer than other
formats (due to the untimed nature of the game) this should not be taken as
a license to write unnecessarily long questions.  A balance should be struck
between giving various clues (arranged from hardest to easiest) which reward
deeper knowledge and giving extraneous information that merely adds to the
word count.  Good toss-ups are generally more than one sentence long since
they contain a variety of clues leading to the correct answer.  As a general
guide, questions should be kept between 2-5 sentences in length.

Spelling Toss-ups

        Because spelling toss-ups are so easy to abuse, they should be 
avoided when writing packets.   If a spelling question must be part of a
round, make it a bonus. 
 
Sample Questions
 
a.  One of only 2 Americans wounded at the Battle of Trenton, he served 
in a Virginia regiment during the Revolutionary War.  He was a US Senator 
in 1790, minister to France 1794-96 and governor of Virginia 1799-1802.  
After helping negotiate the Louisiana Purchase, he served as both 
Secretary of State and War.  FTP, name this President who William Plumer 
prevented from being unanimously elected.

                     James _MONROE_
 
The question starts with a difficult clue (the Battle of Trenton) and 
then gets easier (governor of Virginia ... Louisiana Purchase ...  
Secretary of State and War).  Finally, it ends with a clue that should b
recognizable to most teams (President almost unanimously elected).
 
b.  Born in 1885, she and her husband owned a coffee plantation which she ran
after his death.  The collapse of the coffee industry lead her to sell the
property after which she began writing.  Her first work was _Seven Gothic
Tales_ followed by others such as _The Angelic Avengers_, _Winter's 
Tales_, and _Shadows on the Grass_.  FTP, name this author of _Out of 
Africa_.
 
             Isak _DINESEN_ (or Baroness Karen _BLIXEN_)
 
Again, the question starts with a difficult clue and proceeds through 
more and more well-known works until it reaches the final clue of _Out of 
Africa_.  Also, the writer was sure to include possible alternate answers 
to the question (Dinesen's real name).  Noting alternate answers is a
necessity for question writers.
 
                                Bonuses
 
Structure

         While all toss-ups follow the same general structure, bonuses allow
for some creativity on the part of the question writer.  That creativity does
have some limits however, since all bonuses should be worth 30 points.  By
making all bonuses the same value, teams can be certain of the number of 
points available in a game.  Of course, this means that since all bonuses 
are worth the same amount, the question writer must take care to write 
all of the bonuses for a round on an equal level.

        Bonuses are subdivided into smaller parts, usually worth 5, 10, 
or 15 points each.  One part bonuses should not be used.  In fact , one 
part bonuses already exist in any packet - they are called toss-ups.  
Bonuses are intended to award the team who gets them, so they should 
usually have at least one part which is answerable by most teams (even 
if only worth 5 points).

	Even though all bonuses are worth 30 points, there are a number 
of different ways of reaching this total.  These include 6 parts worth 5 pts,
3 parts worth 10, 30-20-10 (see below), two parts each 15-10-5, three 
parts each 10-5, three increasingly difficult parts worth 5, 10, and 15,
and many others.  However, just as with toss-up, it is extremely boring 
if a packet contains 10 bonuses of the same type.  So, care must be taken 
to vary the presentation as well as the subject material.
 
Cautions

        One type of popular bonus is the 30-20-10.  This question offers 
30 points if a team can get the answer off the first, fairly difficult clue, 
20 off the second, and 10 off the third.  Variants of this type include 
the 30-20-10-5 and other permutations.  These types of questions should
be used sparingly however since a 30-20-10 bonus is essentially a 
toss-up in disguise.

        Writers should avoid bonuses which reward random guessing by teams.
These include so-called binary bonuses where the range of possible 
answers is small, i.e., True/False and Yes/No bonuses. Ordering 
bonuses that require a list of things to be put in a sequence are 
obviously more difficult, but should be used sparingly since misplacing 
one item can cause a team to earn 0 points, despite knowing a good 
deal about the subject.

        Finally, packets should never contain all-or-nothing bonuses (of 
which one-part bonuses are a subgroup).  Bonuses where "a miss stops you" 
should also be avoided since they can penalize teams even if they know 
the subject well. 
 
Difficulty
 
         The difficulty of bonuses, like toss-ups, can be hard to judge.  
However, since teams are allowed to confer on bonuses and since they are 
potentially worth more points, they should be more difficult than 
toss-ups.  Two medium-level teams should expect to convert somewhere 
around 50% of available bonus points.  Combining this guideline and the 
one given above for toss-up difficulty, two evenly-matched average teams 
should expect to have a combined score between 400 and 500 points.
 
Sample Question
 
a.  (30) The promiscuity of Edward III of England caused major problems 
for the lines of royal succession.  His tangled bloodlines were not
straightened out until 100 years after his death.  Identify these 
descendants of Edward III for the stated number of points.
 
 1.  For 5 points, this eldest son of Edward defeated the French at 
Poitiers and was the father of Richard II.
                _EDWARD_ the _BLACK PRINCE_
        
 2.  For 10 points, his father was the bastard son of the widow of Henry 
V and a clerk of her wardrobe, while his mother was a descendant of a mistress 
of John of Gaunt, fourth son of Edward III.  As the earl of Richmond, his 
claim to the throne was tenuous at best.
 
                        _HENRY VII_
 
3.  For 15 points, he was both the grandson and great-great-grandson of 
Edward III.  He began the War of the Roses at the Battle of St. Albans and 
would have become king on the defeat of Henry VI, except for the fact 
that he had died 3 months before the war's end.
 
                  _RICHARD_, Duke of _YORK