Prompt

From QBWiki
Jump to navigation Jump to search

Disclaimer: This article is about rules. Its contents are not authoritative. Please consult official rules for up-to-date information.

NAQT gameplay ruleseligibility rulescorrectness guidelines
ACF gameplay ruleseligibility rules
PACE gameplay and eligibility rules
The Rules of the Game

A prompt is a request for additional information given by a moderator to a player whose answer to a question is partially correct but ambiguous. In this scenario, the player is said to have been prompted by the moderator because the answer they gave was promptable - alternatively, they may say they were prompted on their answer. Receiving a prompt allows a player an opportunity to give an additional answer.

A player may receive multiple successive prompts. It is possible for a player to intentionally aim to be repeatedly prompted to stall or to attempt to gain additional information about an answer, but this is unsportsman-like.

The set of answers which warrant a prompt is enumerated in the answerline alongside the acceptable answers. Under strict readings of the rules, no answers other than those written down should be prompted on; moderators will occasionally use their discretion to determine that an answer should be prompted on or accepted, but this is to be avoided.

Examples

The most common reason for a prompt is that a player has answered with a superset of the correct answer.

ANSWER: the second Punic War [prompt on the Punic Wars]
ANSWER: polar bears [or Ursus maritimus; prompt on bears]

The standard formatting for answers which can be prompted is underlining, though in older sets it is possible to find alternative formats.

Answers which are insufficiently specific to uniquely identify a single person are typically prompted on - for instance, if multiple people in a field share a surname or if it is a monarch with a regnal number. Many of these cases are explicitly included in acceptability guidelines (like ACF gameplay rule G.3A regarding regnal numbers) and do not often receive instructions in the answerline.

History

Answer: Spirit of Ecstasy (accept Spirit of Speed; prompt on "flying lady")

(from 1989 Georgia Tech US Open - Tennessee)

Prompts have existed for a very long time, but not forever. The concept began to appear in packets by at least the end of the 80's, with the University of Tennessee leading the way: the first usage of the phrase "prompt" in an answerline on the archive comes in the Tennessee packet of the 1989 Georgia Tech US Open and the 1989 Tennessee Masters tournament features the equivalent construction "(if national banks given, ask for more information)" in a question on Banks of the United States. During this time period, prompts were often given the same underlining as the answer (if they were formatted at all) - bolding was not used there (or anywhere) because packets were written on typewriters which incapable of embedding bolded text.[1] As with other formatting decisions during the period before widespread adoption of Microsoft Word, many of the specifics vary wildly from year to year, tournament to tournament, and even packet to packet based on who was writing.

ANSWER: 2022 FIFA World Cup stadia in Qatar [prompt on soccer stadium, accept any answer line which includes both "world cup" and any other underlined portion, prompt for “world cup” unless an answer mentioned all three of FIFA 2022 Qatar stadia]

(from 2015 Chicago Open - Round 2, Tossup 17)

In the intervening decades, there was relatively little change in how prompts looked. Technological shifts around the turn of the century eventually allowed the use of bold text, which are now featured prominently in powers and in answerlines.

There have been no comprehensive surveys on the frequency of prompt instructions over time, but a cursory glance at packets from the turn of the century indicates a near complete absence - compare this with modern questions, where they appear in maybe a quarter of answerlines. Usage varies significantly between subjects, with particular topics being very important to

ANSWER: Christian conversion [accept word forms and clear-knowledge equivalents about converting to specific denominations; accept rebirth or regeneration; accept Christian testimony or testifying; accept being born again or Christian witnessing; accept turning before mentioned; prompt very generously on “proselytization”, “evangelism” and word forms, things about “missionaries”, “preaching”, “baptism”, “prayer”, “faith”, "divine revelation", "receiving a message from God", or other related answers]

(from 2015 RILKE - Round 1, Tossup 11)

The practice of underlining prompts was introduced by John Lawrence in RILKE,[2] which was a literature side event played after the 2015 Chicago Open which formatted prompts with quotation marks and underlines. Other events that weekend continued to use just quotation marks or nothing at all. This innovation was later brought to the masses in 2016 ACF Nationals, which was probably the second tournament to make use of them.

In recent sets, almost all prompts now use this format.

The antiprompt is a recent[when?] development which indicates that an answer was too specific, rather than not specific enough.

The directed prompt was introduced recently[when?] and embeds an instructions to be provided to the prompted player.


Applications

Theoretical

The strictest interpretation of prompts dictates that they should only be awarded to answers whose ambiguity is the only barrier preventing them from being accepted - answers which are overly specific/general or are a superset/subset of the correct answer are simply incorrect. In this view, prompts are used purely to allow players to clarify that the answer they gave was the correct answer. This is, in fact, stricter than any ruleset currently used in quiz bowl but is still interesting from a theoretical standpoint: questions written to fulfill this definition would require extraordinarily precise clue choice and language in order to ensure that players could determine the correct answer precisely enough. This specificity means that the set of players who knew the clues in these questions would be exactly the set who knew the answer.

At the other end of this spectrum is an incredibly permissive concept of a prompt which awards them to effectively anything that is even roughly correct. This is so broad that the utility of prompts approaches zero - if anything can be prompted on, the player has no way of knowing whether their answer is too specific, too general, or something else entirely. Situations can approach this if there is a mismatch between the scope of the answer and the scope of the clues, which can happen in vanity sets and/or bad questions.

The current state of prompts lie in between these two extremes, with the rules mandating prompts for various forms of inexactitude. Additionally, writers and editors are given room to grant prompts and thus sets can be more permissive than this.

Practical

Prompts allow players an opportunity to receive points for knowledge that is not precisely what is asked for. That they are not as strict as they could be is a concession to players and writers alike, granting room for imperfect play and less tightly written questions. The need for some prompts of some questions could be removed by editing to a higher standard of uniqueness, but conversely the ability to add "extraneous" prompts allows for a broader range of topics and approaches to be used in questions.

Both ACF and NAQT provide formal definitions for what prompts are, but the criteria that an answer must satisfy to warrant a prompt are not super strictly defined:

ACF NAQT
If a player gives an answer which is not outright incorrect, but is nonetheless not specific enough to precisely identify the answer, the moderator must prompt, asking the player for a more specific answer (e.g. “I need more,” “More specific, please”, “More,” or simply “Prompt”), rather than ruling on it outright.
-gameplay rules section F.4
A moderator may prompt for a clarified response in three cases:
a. If a player gives an answer that could be interpreted as compatible with the clues that have been read (in the case of a tossup) or with the entire part (in the case of a bonus), but which is ambiguous, the moderator will prompt by saying something like “more information, please” (e.g., a player says Roosevelt, and the answer sought is Eleanor Roosevelt).
b. If a player gives an answer that is compatible with the clues that have been read (in the case of a tossup) or with the entire part (in the case of a bonus), but which is too general or too specific to be considered precise, the moderator may also prompt. For example, a question beginning, “During World War II, the primary target of the Blitz was…” may have the moderator prompt on responses of United Kingdom or England if it is looking for London. Some responses may be too general to warrant prompting (e.g., a response of Northern Hemisphere for a question beginning, “This region is home to the Yenets people, and it…”).
-rule I.3a/b

The process of giving an answer in response to a prompt is different from simply answering a question because the player has additional information: they now know that their original answer was "close enough" to a correct answer. Because of this advantage, players are generally assumed to have similar, if not the same, chance of converting a tossup after being prompted and at the end. This assumption is central to many conventions of protest resolution, e.g. ACF gameplay rule H.5E states that in situations where a player "should have been prompted" their team will be awarded an uncontested tossup. This has been challenged as unfair,[3] although no strong alternatives have been considered.

Much of the utility of prompts to the player arise from the well-defined and relatively restrictive conventions for which answers can be prompted on. The player typically only has consider a limited range of ways in which they can modify their original answer to get the correct answer. This can be compromised in situations with atypically permissive criteria for prompting, where players may be unable to determine exactly how their initial answer (or answers, as is often the case in questions with many prompts) was incorrect. To give an absurd example, a ruleset which obeyed conventional guidelines for prompting but added in prompts if the answer given rhymed with the primary answer would be much harder to play despite more opportunities for players to be prompted.

References

  1. Bold and italic fonts are different typefaces - though the character shapes are reminiscent of those of the regular font, they are still different. Though invisible on modern text editors, the process of switching between boldface and regular text on a typewriter involved switching the typewheel or typeball, as well as owning a typewriter capable of such a change.
  2. Re: ACF Nationals 2016 Discussion by ThisIsMyUsername » Wed Apr 20, 2016 6:10 am
  3. "Should have been prompted" replacement policy by Mike Bentley » Mon Jan 13, 2020 10:33 am