Michigan MLK Memorial Question Memorandum

Version 2.001

We wrote the so-called "Stanford Guidelines" in 1991, when no Guidelines gave question writers basic instruction. The Guidelines, however, have not kept up with the game. Basic problems of question quality still persist four years later.

This Memorandum addresses many persistent problems with questions.It is my hope that this Memorandum will be treated as an authoritative clarification of the "Stanford Guidelines," until we update those Guidelines to reflect these additions (some of which do conflict with the original Guidelines).

It is also my hope that this Memorandum will be read by *all* team captains and Tournament Directors (as well as by "professional" question writers, who routinely violate many of the rules below). Tournament Directors who follow the Guidelines for their events should ensure that all questions conform to these instructions. Deviations or editing lapses that result in punishing players are never acceptable given the high level of competition on the circuit and the enormous expensesinvolved in competing in tournaments.

Bad questions (and the protests that follow) decide matches, decide tournament winners, and decide (as we all now know) national championships. They also spoil the enjoyment of the game for spectators and players alike, and produce unnecessary ill-will.

With all questions, ask this: if the match were tied, and the outcome of this question decided the championship, would you be proud to have this question decide the tournament? If no, then the question does not belong in the packet. Of course, any *one* question between two good teams is virtually a coinflip. The point is that every question be as fair as possible, so a misleading, poorly written, or inaccurate question does not decide the outcome. One question often *does* decide the outcome of the tournament, which is why every question must be written and edited seriously.

The quality of the competition in our activity has quickly outgrown the quality of the questions. Every week, we ask what are essentially professional players to compete on embarrassingly amateur questions, as if Major League Baseball were played in sandlots filled with broken glass and unmowed grass. This is most acutely true at recognized Regional and National Championship Tournaments.

If we do not undertake efforts to bring the quality of the questions up to the level of the competition, we will undermine the growth of the game. The fairness of the questions is the very stuff on which the integrity of the game depends. There are no more excuses for poor questions.

This document was produced with the extensive input (but not necessarily consensus) of many persons, including Pat Matthews, Eric Hilleman, John Edwards, Tom Michael, Chris Golde, Doug Bone, Gerard Magliocca, Kevin Olmstead, Mitchell Szczepanczyk, Joe Saul, John Davenport, Rob Long, Maya Kobersy, and Joel Goldberg, some of whom contributed exact wording to certain sections. No doubt, the next version will benefitfrom the comments of dozens more people who play and love the game (please email comments to dfrazee@umich.edu).

David Frazee
November 1995

I. Cardinal Rules ("Wolverine Rules"?)

These four cardinal principles underlie this Memorandum. Should question writers follow them, as well as the spirit of fairness they embody, no further instructions would be necessary.

1. All questions must contain many pieces (ideally 4+) of substantial information, with clues ordered from most obscure to least obscure. The challenge is to compact as much information as possible into 3 or fewer sentences. A well-written question will often contain 3-4 pieces of information *per sentence.*

2. All questions must clearly and succinctly move unidirectionally toward one unambiguous answer from the first word to the last. What that means is this: writers should read questions one word at a time, thinking of all possible answers at *every* point in the question. If the question at any point is narrow enough to draw a reasonable answer that is wrong, the question must be rewritten. Often it is good to have questions that are a little bit sneaky in terms of where they're going, so long as they don't invite someone to leap to a wrong answer.

3. Someone with perfect knowledge of the subject should usually know the correct answer first. Of course, that player may not have the courage to buzz-in first (or may be hung over from late night NTN, etc.), but questions should create conditions in which the most knowledgeable player is the one likely to answer the question first. In no event should knowledge be punished, though many of the best questions are more like puzzles that reward more than just "pure" knowledge.

4. When the above three rules are not violated, questions should also entertain as well as educate, as boring questions destroy the joy of the game. An important, but unnumbered, rule is that the above rules can be violated. There are always some cases in which exceptions are perfectly OK. As you gain more expertise. you will learn these exceptions. However, know what you are doing first, as even experienced question writers should approach exceptions with caution. This is a wickedly fast game played by many truly gifted players. Should one of your questions violate one of these rules and draw an unjust penalty from a good player, you justifiably deserve their wrath.

Anatomy of a Good Tossup

Consider these examples:
Example 1:
Born March 9, 1940 in San Juan, Puerto Rico, he launched his theatrical career in the New York Shakespeare Festival in 1971 as Proteus in "Two Gentlemen of Verona"; his last role was in "The Burning Season" as Chico Mendes. For 10 points--name this actor who died in October 1994.
answer: Raul Julia or Raul Rafael Carlos Julia y Arcelay

This question is a straight-forward factually-dense question. It has many substantial clues: (1) his birthday. (2) his birthplace, (3) the location and date of his career beginning, (4) the role and play involved, (5) his last role and film, and (6) the month he died.

Example 2 (modified from question by Gerard Magliocca):
This author's wife was upset when she saw his classic 1939 novel in a Japanese bookstore translated as "Angry Raisins." For 10 points--name this American author who described Rose, Sharon, Ma, and Tom as the Joad family travels during the Depression from Oklahoma to California in "The Grapes of Wrath."
answer: John Steinbeck

This question has an interesting lead-in that will appeal both to competitors and spectators alike, making the question entertaining and memorable. It also have many substantial clues: (1) the date of the novel, (2) the mistranslation, (3) the nationality of the author, (4) four of the characters in the novel and the main family name, (5) the time setting of the novel, (6) part of the plot of the novel, and (7) the title of the work. Obviously, someone who has read the novel a dozen times may be beaten by someone who is able to figure out the answer based on the mistranslation of "Angry Raisins."This question, unlike the previous example, is more like a puzzle in the beginning, rewarding more than "pure" knowledge (but not punishing it). Many of the best questions have this quality.

II. Answers

1.When more than one answer is required for a tossup, that must be stated at the outset in some form. "These two U.S. presidents..." or "Two answers required."

2.When a lead-in could lead to multiple possible answers – and this cannot be corrected by rewriting the question – a prompt must be placed to either accept the other answer(s) as correct or to prompt for the listed answer. In a creator-creation situation, this caveat need not apply, as players may already give both pieces of information, but it is recommended.

Consider the question which begins:"F. Sherwood Rowland of the University of California-Irvine..."A good player will buzz in here and answer "1995 Nobel Prize in Chemistry." However, the question goes on to say "...won the 1995 Nobel Prize in Chemistry for their research on which layer of the atmosphere." (answer: Ozone layer). A good question writer will either add a host prompt to accept "1995 Nobel Prize in Chemistry" on an early buzz or will rewrite the question. A better lead-in is "This layer of the atmosphere was the subject of research by..."In no event is the question acceptable as it was first written unless it has a host note to accept or prompt.

In general, host notes are suboptimal solutions as novice question editors tend to remove them (not understanding why they are added by experienced question writers) and since some moderators miss them. They also allow a question writer to write a weaker question. It is always better to have a question that leads unambiguously to one answer from the beginning.

3. Married and unmarried names for persons, especially women, must be supplied whenever appropriate. Birth names, as well as professional names and pen names must always be listed, along with common nicknames. When these are insufficient to receive credit, a prompt must be listed.

4.Royal titles must be listed whenever appropriate.

5.Only the *minimum* information necessary to distinguish should be underlined. "Eleanor Roosevelt" is sufficient.

6.Court case names usually require only the first party listed, though sometimes the second party alone is sufficient, especially if the first party listed is common. (e.g., Marbury v. Madison, Gitlow v. New York, United States v. Nixon.) Sometimes, legal scholars choose to refer to cases based on the latter party's name for other reasons (such as the first party having a long name or being a scumbag, e.g., Bowers v. Hardwick or Bowers), in which case either party's name should be acceptable, or may shorten the party's name (Meritor Savings Bank v. Vinson or Vinson).Sometimes, especially with older cases, the parties may have a different order at the trial, appellate, and Supreme Court levels. Players should be prompted for a full case name only if the answer is ambiguous.

7.The *family* name must be the one underlined. For languages other than English, this is not necessarily the last name listed. When in doubt, check a biographical dictionary.(One should always fact check all names of persons in a biographical dictionary before submitting a packet).

8.Country names must also be listed in the country's own language. "Kingdom of Bhutan or Druk Yul". The names of countries in their own languages are listed in almost all almanacs and geographical dictionaries.

9.Works written in a language other than English must have their original language title listed, as well as all English translations."L'Etranger or The Outsider or The Stranger". Original titles are always listed in the various Benét's Reader's Encyclopedias, and most other reputable reference source.

Ethnocentric (if not overtly racist) English-only rules are never acceptable for this activity. If foreign language alternatives are too messy, the writers should never write questions dealing with other parts of the world. This is not acceptable, though. English speaking nations are not the entire world, though question writers often act as if this were the case. The world is a diverse place, and this game is played by persons from all different parts of the world, as well as by persons from English-speaking nations who have an expert knowledge of other cultures. Someone with a native or expert knowledge of a foreign language and culture should be rewarded for this knowledge, not actively punished. More importantly, the question is about something that was originally written in another language. Indeed, if only one answer must be preferred, the foreign language title is the *only* correct answer. The English translation of that work is only a derivative of the work, but one that we choose to accept for ease of administering the game.

10.Whenever the answer given is substantially equivalent to the listed answer, but the given answer is later used as a clue in the question, the given answer must be accepted or a prompt must be given. For example, if an abbreviation for a group is given as a giveaway clue, but the player buzzes in early and gives the abbreviation, that should be taken as correct, or the player should be prompted for the group's full name. Players should not be punished for not being 100% clairvoyant.

11.When an exact answer is given that identifies the information in the question, that answer must be accepted, even if the listed answer is only an approximation of the given answer. For example, if a question begins, "You want to find the area of a regular hexagon with sides 6..." and a player answers "24 radical 3," that answer should be taken as correct, even if the question finishes with "is the area closest to 31.4, 41.4, or 51.4?"Of course, no competent question writer would ever write such a terrible question (see also the section below on multiple choice answers). This point, though, is this: never should an exact answer to a question be counted as incorrect. When this occurs, it absolutely undermines the entire philosophy of academic competitions and constitutes a serious breech of ethics – compromising the trust that players must have in the integrity of the questions and in the game itself, penalizing the very knowledge which is the basis of the game.

12.When the information in the question is actually performed by a closely related group, or a sub-group of the listed answer, that other answer must also be listed, or ruled out. For example, many writers may confuse the precise roles or relationships between of the Federal Reserve Board and the Federal Reserve Open Market Committee or the Department of the Interior and the Bureau of Indian Affairs.

13. In a calculation question, the units of the final answer should be specified in the beginning, or equivalent answers must be listed to handle otherwise correct, but early answers.

14. Scientific names must be listed as alternate answers whenever appropriate, especially for animals and plants.

15. Answers must also list commonly used symbols and letters that represent the same answer, such as the letter "c" for the speed of light. Note, though, that questions that ask what certain letters stand for in science are almost always awful, as the hundreds of major scientific fields all have their own abbreviations and symbols.

16. Chemical symbols as well as element numbers should be accepted, or should be listed as answers that require prompts for more information. On chemical formulas: if the question is going to be reasonably gettable by most players who have been through introductory chemistry, the formula should not be too hard in the first place. A possible problem with the formulae are when the same empirical formula applies to different chemicals. For example, a question leading the answer "fructose" should NOT take C6H12O6 as the sole correct response (prompt, though), since that formula also applies to a host of 6-carbon sugars (glucose among them). If there is a question about whether a moderator can interpret the formula, spell it out (e.g., C-6-H-12-O-6 or equivalent rearrangements, although for this example, a chemist who states it as "O-6-C-6-H-12" is a sick puppy). A knowledgeable chemist will shift from chemical formula to common name to uncommon odd name (e.g., "muratic acid" for hydrochloric acid for HCl (H-C-L)) to chemical (IUPAC) name with reasonable facility. For this reason, all of these reasonable answers should be listed.

17. Official postal abbreviations should be accepted for states, or should be listed as answers that require prompts for more information.

18. Transliterated answers must be treated carefully, as they often generate a number of correct possible English translations. This is especially true of titles, such as "Night on the Bare Mountain," "Night on Bare Mountain," "Night on the Bald Mountain," and "Night on Bald Mountain," all of which are common English translations for one Mussorgsky work.

19.Words with the same root that show knowledge of the correct answer should usually be listed also. For example, a question that begins ""this word is derived from..." which asks for something like "jingoism" should also accept "jingoistic." This would normally be an example of exact and unambiguous knowledge, except that many writers at the *end* of the question include something like "...name this 8-letter word." Again, players should not be punished for lacking 100% clairvoyance.

20. Questions that do not indicate a level of geographic specificity early must accept other answers that show exact and unambiguous knowledge, or must prompt for more information. For example, "The Headquarters of the International Labor Organization, GATT, the UN High Commissioner for Refugees, WHO,..." should ideally be answered "Geneva." However, if the question finishes "...are all located in what country," then the question is misleading.

21. As a final note, many movements have been made to permit blitzing or shotgunning of answers. These rules are only necessary if questions are not written correctly, as correctly written questions lead to one unambiguous answer. Permitting blitz answers is misguided because it permits the existence of poorly written questions ("well, they could have given more information if they really knew it"). The improvement of the game (as well as the aesthetic beauty of it) demands that question writers be *forced* to produce tightly worded questions that lead only to one answer.

III. Types of questions

1. Spelling tossups are inappropriate for competition except in extremely rare circumstances in which it is clear to *everyone* that the answer is a spelling question. The presence of spelling tossups destroys players' reflexes, as they cause players to buzz in and spell difficult words (thinking falsely a question is a spelling question), or misleads players into answering with the word instead of spelling it. When spelling is used in bonuses, the spelled word must never be a transliterated word. Alternate spellings must also be listed, including spellings used in other English-speaking countries.

2.Math or calculation tossups are acceptable, but they must be carefully written not to mislead players into buzzing before they have all the operations they must perform. In addition, they should never be exercises in brute, mechanical calculation speed.

3. Too often, question writers end questions with "name this 11-letter word," which punishes someone who gives the adjective form instead of the noun form. This is not a game of counting letters, but of recall of exact factual information.

4. Questions are too often not gauged to the appropriate difficulty. Please see the note on the distribution of difficulty attached as the appendix.

IV. Lead-ins

1.Most writers are too lax with question lead-ins. Questions must move unidirectionally toward one unambiguous answer. As stated above: writers should read questions one word at a time, thinking of all possible answers at every point in the question. If the question at any point is narrow enough to draw a reasonable answer that is wrong, the question must be rewritten. Period. Read this point again.

2.Similarly, non-unique lead-ins that narrow the field of possible answers to just a few are no longer acceptable. Lead-ins must uniquely pinpoint one possible answer as soon as possible, or must be general enough not to draw an incorrect early answer. "Maya Angelou wrote a poem for this event..." is not acceptable, as she wrote poems for (among other events) both Clinton's inauguration and the Million Man March. Instead, the question lead-in must limit the range of answers to one (e.g., "In October 1995, Maya Angelou...) Often, if you could not use it as a 30 point clue in a 30-20-10 bonus, it is not a good lead-in. In contemporary competitions, most early incorrect answers are the fault of the question writer. Many of the best players are also the most penalized.

3. Question writers still use pronouns poorly. "It was inspired by Sauk Center, Minnesota..." is not acceptable, as it could draw an answer of both "Main Street" as well as "Gopher Prairie." A better lead-in is "This city was inspired by Sauk Center..." Another example is "President McKinley was assassinated here...," to which the answer could be "Buffalo" or "Pan American Exposition."

4. A related problem is the use of passive voice. "A new color was added in 1995..." could lead to either "blue," "M&Ms," etc. This is a bad lead-in.

5.Too often, pronouns are used correctly, but do not distinguish between characters and actors. For example, "He uttered the famous line, 'Frankly my dear...'" does not distinguish between Clark Gable or Rhett Butler.

6.A related problem is recitation of quotations without specifying whether the speaker of the quote or the subject of the quote is wanted. For example, "Tell me what brand of whiskey he drinks. I would like to send a barrel of it to my other generals" could lead to either Grant or Lincoln.

7.Introductory phrases that modify a noun are not acceptable. For example, consider the lead-in "The author of the 1914 collection of short stories 'Dubliners,' James Joyce was also famous for..."When someone hears "The author of the 1914 collection of short stories 'Dubliners,'" they will justifiably buzz in with "Joyce" and draw a penalty. Such lead-ins are not acceptable.

8. Introductions that do not have anything to do with the final answer are rarely acceptable. Anything in the form "X is Y, but for 10 points, what it Z" must be rewritten (e.g., "'A Tale of Two Cities' was written by Charles Dickens, but who wrote 'Moby Dick.'")Similarly, the question "Adelbert Ames served as governor of Mississippi from 1874....For 10 points--name the current governor" is unacceptable.Someone hearing "Adelbert Ames served as governor..." will justifiably buzz in with Mississippi.

9. Giving one half of a well-known pair in a lead-in, when not asking for the other half as the answer, is often misleading. For example, "One type of cholesterol is LDL..." would lead to an answer of "HDL." If the question goes on to ask "what does LDL stand for" or any other information, it is misleading. Similarly, if the question begins "Leander swam the Hellespont nightly to be with his love...," a player would buzz in with "Hero." If the question then goes on, however, to ask about Lord Byron swimming the Hellespont to re-enact this scene, the lead-in is misleading and must be rewritten.

10. The lead-in "for a quick 10 points" is usually not acceptable, nor are the questions that usually follow it. Questions that are only valuable to differentiate the split-second reactions of 8 simultaneous buzzes have *no* place in competition. Sometimes, writers use this lead-in to signal that what follows is a single sentence, which may have many pieces of information, ordered from most to least obscure. Varying the length of the questions keeps the rhythm of a packet lively. What is *never* acceptable, however, is one or two fact questions. In games like Jeopardy! in which no one can interrupt, it is not that important whether questions have multiple levels of information.*That is not the case here.*

11. Describing a work or a scene from a work when not asking for the title or author is misleading. A question beginning "A rat is killed with a frying pan is thrown across the room in a small Chicago apartment..." is acceptable if the answer is "Native Son," but not if the answer is "Bigger Thomas."

12. Questions should be written in parallel structure when possible. For example, a question that begins, "Until 1997, the tallest building in the world will be located in Chicago..." could draw an answer of "Kuala Lumpur", the city in which the new Petronas Towers are being built. It is just misleading if the question goes on to ask "... in what country...," wanting Malaysia. If the question wanted the country, it should not have dangled Chicago in the lead-in. Of course, the lead-in above, while it illustrates this point, is misleading for other reasons, as players could buzz in after "Until 1997, the tallest building in the world..." with either "Sears Tower" or "Petronas Towers," depending on where they think the question is going--the player should never be required to guess in this fashion.

13. Many lead-ins are good, but do not lead to the desired answer. If a question begins, "Manzanar in Lone Pine, California..." and a player answers "Japanese Internment Camps," that answer should be accepted as it *exactly* identifies what was asked, even if the question (stupidly) goes on to ask "what group of Americans were interned there during WWII," and the listed answer is "Japanese Americans." This is a case in which the question is acceptable, but the answer listed is the wrong one.

14.Questions which list events and ask for which came first (or last) are unacceptable. Whenever someone hears of an event, they justifiably buzz in with the year of the event, as that is a stylized type of question. Upsetting that pattern is misleading. Of course, *any* tossup that asks the players to select one answer from a list is simply unacceptable. This point against multiple choice questions will be emphasized more below.

15. A question of the form "which does not belong" is always unacceptable. For example, "'Innocents Abroad,' 'A Connecticut Yankee in King Arthur's Court' and 'The Origin of Species' – which one was not written by Mark Twain" is simply unconscionable.

V. Style

1. Sexist language as well as questions that assume certain social roles for women are unacceptable. Women should be identified primarily by their accomplishments, not their relationships to men. Inappropriate comments and questions about women's weight, bodies, sexual roles, etc. must be removed from the game.

2. Racist language or racially insensitive questions must similarly be avoided.

3.Questions must be easy to read aloud. We recommend that writers read the entire packet aloud as a final edit.

4. The use of the word "[blank]" in questions is often hokey and confusing, especially in tossups.

5.Questions must be tightly edited to remove all extraneous words and syllables. Correct editing can reduce the length of questions by up to 50%.In a timed game, this means more questions may be asked. In an untimed game, tight editing prevents the game from deteriorating into drudgery and boredom.

6. Writers must use correct grammar. Though many strict grammatical rules interfere with the flow of questions meant to be read aloud, basic grammar rules must be followed whenever possible.

7. Inside jokes are almost always inappropriate, as are negative comments about other schools.

8.Simple past tense should be used when appropriate. Past perfect should be used only to refer to events completed at or before a past time spoken of. Simple past also removes the extra word "had."

9. Use of the phrase "politically incorrect" and other meaningless stock-phrases should not be used, especially if they usually are used derogatorily toward underrepresented players in the activity.

10.Eliminate redundant wording.

VI. Substantial Information

1. Too many questions contain vague, weak, fluffy, or insubstantial information. Others have substantial information, but take too long to get to it, meandering for 2-3 lines before the meat of the question begins. Just ask the question.

2. Clues must be ordered from most to least obscure. Writers still do not apply this basic rule, perhaps because many lack the experience and knowledge to anticipate what other players will and will not know.

3. Too many questions have very difficult clues followed by very easy clues, with no intermediate levels of information. If a question topic lacks intermediate information, the subject may be inappropriate for a tossup. When a potential 80+ point swing hinges on every question, every professional attempt must be made to make the tossups multi-leveled and favorable toward players with more knowledge.

4.Largest does *not* mean most populous. The largest city in the United States is not New York City, but Anchorage. Questions that ask for population figures must specify the measured entities (cities, SMSAs, etc.) as well as what measurement is being used (e.g., the 1990 Census).

5. Inexplicit Criteria must be avoided. Qualifiers like the costliest, the most expensive, etc. invariably lead to protests. The criteria need to be explicitly stated. What is meant by the "largest insurers"? The insurance companies with the most employees, the most office space, the greatest net profits, most insured, greatest gross revenues? What is meant by "the most expensive film ever made"? Adjusted for inflation by the CPI? In unadjusted dollars? Does it include advertising? Depending upon the qualification, this could lead to different answers (and could become out of date quickly, which requires date specification.)

6. When questions refer to events that occurred in geographical entities that no longer exist, the question must carefully pinpoint which answer is desired. A question referring to a 1970 natural disaster in what is now Bangladesh would likely cause a protest if East Pakistan is not an acceptable answer, since that was the entity's name in 1970.The same goes for the USSR, Yugoslavia, etc. If the desired information is not pinpointed very early in the question, a prompt must be placed in the answer should a player give the older (or, as appropriate, newer) name for an entity.

7. Factual errors occur with far too much frequency. Some occur because writers do not consult reference sources carefully when writing their questions; it is *always* inappropriate to write questions from memory. Others occur because some reference sources are not reliable.

8. More errors occur because misinformation permeates the information/reference source industry, as writers and publishers attempt to come up with the most outrageous, scintillating "facts" to sell their books. If something seems suspicious, check it with a reliable source. If it cannot be verified, strike the fact from the question. Overly dramatic questions with unverifiable assertions about history must be avoided.

9. Some information differs because the sources differ. When a conflict among reputable sources differs, the information must be omitted, or the conflict of the sources must be noted.

10. Word and phrase origins cause persistent protests. Take the example of the notoriously unreliable history of "OK." I have one source that lists over a dozen possible derivations. Unless this conflict is the meat of the question, it is inappropriate to write a question about the derivation of "OK" as it will just lead to protests.

11. Dubious honorifics must be avoided. A question that begins with "He has long been considered the supreme lawgiver of the English language..." is not acceptable (answer:Samuel Johnson). Just *who* has "long considered" him "the supreme lawgiver of the English language"? Though it makes questions slightly longer, if writers want to include such dubious honorifics, they must include who bestowed them for a semblance of credibility.

12. Subjective judgments asserted as fact must be avoided. Lead-ins like "the most important book of the 19th century..." lead to debates about The Origin of Species, Uncle Tom's Cabin, the Communist Manifesto, and many other works.

13. Whenever debatable judgments are made, the criteria must be specified. For example, what is meant by "the major persecuted ethnic minority" of a country – the most oppressed? Oppressed by the most people? The biggest group of people who are oppressed? The most noticeably oppressed? The group most in a position to bring its oppression to public debate? This is not the time to generate political, theological, or other endless debates. Use specific, verifiable facts in your questions.

14. In an effort to make questions creative and colorful, writers often speculate about motives, ideas, and thoughts of the question subject, or offer interesting interpretations of limited factual data. Such coloration, however, often makes questions factually inaccurate. Any "creativity" which butchers factual accuracy for cleverness isn't worthy of the label. A creative question will express information in a different, surprising, or thoughtful way, but not in a way that renders the information absurd or wrong. If speculation will make a question more interesting, it must be noted as speculation with qualifiers like "perhaps."

15. Writers must correctly distinguish between percent and percentage. "Percentage" is a mass unit number which carries the unit of the original, while "percent" is a fraction of 100.  If 18 of 24 tossups are answered correctly, the "percentage" answered correctly is 18 tossups, or 75 percent.

16. Questions with answers that are academic disciplines or broad fields of thought should be avoided. Such questions with answers as existentialism, anthropology, sociology, etc. are invariably inane, limited, and wrong. Modern academic disciplines, with dozens of subfields, do not allow for such questions. Ask something about the *content* of the field instead.

17. Questions with weak premises should be avoided. A common such premise is "he is the last entry in the biographical dictionary." Depending upon the dictionary, this could be Zwingli, Zworykin, or perhaps others. This is not a constant fact, and the ordering of someone's name in a dictionary is just stupid information compared to their accomplishments.

18. Works performed by teams should not be attributed solely to the group's leader, as it is factually inaccurate.

19. Stock clues should be avoided. It is always acceptable to ask new and interesting questions about familiar subjects. Using tired, often-repeated information as clues, however, makes for boring games and provides an unnecessary advantage for warhorse players over novices.

20. Information that is time dependent or may become out of date must be date-stamped. Usually, a month and year is sufficient.

21. Calculation of the number of years that elapse between dates should usually be avoided, especially if rounding could yield conflicting answers (e.g., "If the next landing of humans on the moon occurs in the year 2025, how many years will have elapsed since the first?" as it depends on whether the landing takes place after July 20, 2025.) Asking for the current year in other calendar systems has the same problem. The exact reference points must be specified.

22. Questions should not assume that the United States or Europe is the world. Information like "the most influential musician of the 1800s..." begs the question "influential to *whom*"? It is also important to specify humans when appropriate (e.g., "I will give you the name of an organ, you tell me..."), as many biologists play this game who know information about quite a number of species.

VII. Bonuses

1. Questions that stop when the players miss an answer should be avoided. This is a qualifier that is a crutch to prop up bonuses that are too easy or too sloppily written.

2. All or nothing bonuses should also be avoided (except for single answer bonuses, of course), for the same reasons.

3. Questions that require the ordering of a list or matching of items from two lists should be presumed invalid unless they have some exceptional redeeming merit. These questions reward luck and guessing, and may actually punish knowledge, as a team that places only one item out of order may receive no points. They also permit a team *with no knowledge of the subject whatsoever* to win a tied match by guessing the same answer for each slot. Write instead about something substantial about the subject. Content-based questions requiring exact factual recall are always better.

If you write a matching question, you should follow these rules:
(a) No matching of pairs from two different lists.
(b) Keep list to 5 items *MAX*.
(c) There must be redeeming merit in knowing order the list.
For example, this is an acceptable list to order chronologically: "execution of Robespierre, assassination of Marat, storming of the Bastille, execution of Louis XVI, and Tennis Court Oath is taken."
(d) In a list like the above, players should only need to give short bits of information that uniquely identify the items to get points--such as "Robespierre," "Tennis Court oath" – though "the first one" and the like are not acceptable.
(e) There should be no more than one of these questions in a packet.
(f) This type of bonus should be used earlier in the packet to prevent teams from having a real incentive to guess the same answer for each position just to get 5 points (such as a tied score with time running out).

4. Multiple choice questions are *never* acceptable. If you expect teams to know nothing about the question—so much so that you must give choices--then the question is inappropriate. Write a different question. Again, it is never acceptable for teams to get points by guessing when the team *has no knowledge of the subject whatsoever.*The same applies to true/false and other limited choice questions (even with 4+ choices).A question that does not lend itself to a unique, verifiable answer is inappropriate.

5. Bonuses that call for players to derive a list, then order it, are punitive and in extremely poor taste (e.g., "name the first six U.S. Supreme Court Justices--in order.") If generating the initial list is sufficiently difficult, then it should stand in its own right. If it is too easy, it should not be asked at all. It is acceptable to ask for a list and give a bonus if all are in correct order, but denying points for what may be just one ordering mistake is outrageous.

6. Questions should have no more than 3 parts, 4 if they are short. Instead of having 6 parts, choose 3 parts with a range of difficulty. If four-parters are used, they should be limited, and used earlier in the packet so that teams may not use them to stall the game.

7.Questions should not repeat the point value in each sub- part when it is the same. Just say "for 10 points each" at the beginning and ask the questions. Repeating the point value wastes time.

8.Questions that call for players to guess numbers, usually within a range, are often inappropriate for competition. Such questions invariably result when someone opens an almanac and writes trivial information into a question. Anything reeking of "guess which number I'm thinking of" usually have no place in this activity, with this caveat: asking for players to name x year within y years is reasonable, as long as the x year is reasonable and worth knowing.

9.Progressive, or 30-20-10 questions are widely abused. First, when the answer itself is so difficult after all clues have been read that it could be a single part 20 or 25 point bonus, it is inappropriate to use a 30-20-10 format. Second, 30 point clues are too often insufficient to generate an answer (e.g., "This writer was born in 1907 in New York City" could apply to literally dozens of people). 30 point clues must *always* pinpoint one unique gettable answer. Third, sometimes the 30 point clue is either impossibly difficult or inanely easy, neither of which is appropriate. Finally, too often the 10 point clue is far too easy. Teams are not entitled to 10 points automatically. *Good* teams should be able to get the question at some point 90% of the time, though.

10. When a single answer bonus is so difficult that even good teams will have little chance of answering it, it should usually not be asked. Sometimes it is acceptable to make such a questions into a tossup. If a team misses it as a bonus, they have been robbed of a reasonable chance to earn bonus points. If it is a tossup early in the game, then neither team is much harmed if it is unanswered (This does not apply to set-question rounds, however, as unanswerable questions punish both teams.) On the other hand, if a team gets the tossup, then it has well earned its bonus. In any event, the difficulty range of the tossups as a whole must conform to the range specified in the appendix.

11. Periodic table chess and the like are forever banned. Memorization of lists and tables is not something to be encouraged. These questions are almost always the result of last-minute question writing.

12. 10-5 doubles should only rarely be used, as they waste too much time (i.e. "if you can get this on the work, you get 10 points, 5 if you need the author"). Usually, it is better to choose 10 point clues that have a range of difficulty instead. If you do use these questions:
(a) they should be limited to two 10-5 (or 15-5, or 15-10) pairs (no more than 4 part for the *entire* bonus),
(b) the first clues should be very short, with the secondary clue being excruciatingly short,
(c) the form should be used only once or twice a packet, and
(d) the form should only be used toward the beginning of the packet so that teams do not use it to stall the game.

13. Too many single answer bonuses should not be used. As a rule, no more than 25% of the bonuses should be single- answer converted tossups. They have a value in the beginning of the packet to speed the game up, as well as near the middle and end of the second half, to allow the team behind to stage a quick comeback and to prevent the team ahead from killing the clock on its bonuses. But they should not be overused.

14.Trite, insipid lead-ins to bonuses should not be used. Anything resembling the following is off limits:

"My roommate Al has a complete Beatles collection..."
"Everyone has heard of..."
"We all know that..."
"Let's play that game: match the artist..."
"It's time again for everyone's favorite African capitals quiz..."
"You might know (or not know) about X, but this question is about Y..."
"For example, if I said X, you'd say Y..." (If the question is that difficult to understand, you might consider not asking it.)
"How well do you know your Athenian history?..."
"The breakup of Yugoslavia has created a bunch of new world capitals for any good academic competitor to memorize. How far along are you."
"We all grew up listening to early 80s rock and roll..."
"Let's hearken back to the not so distant past..."
"It's time for your favorite biology game..."
"It's time again for everyone's favorite question..."

Just ask the question. The same goes for tossups.

15. Variable value bonuses are right out, the element of luck should never play into a team's victory that they got a bonus worth more than the opposing team. All bonuses should now be worth 30 points in the end.