Difference between revisions of "Fairness"
Kevin Wang (talk | contribs) |
Kevin Wang (talk | contribs) m |
||
| (8 intermediate revisions by 2 users not shown) | |||
| Line 1: | Line 1: | ||
<onlyinclude>The concept of '''fairness''' is the idea that the two [[teams]] competing in a [[game]] of quizbowl should have equivalent opportunity to receive points and ultimately win. For a game to be fair, both the [[questions]] and the [[format]] must be fair. | <onlyinclude>The concept of '''fairness''' is the idea that the two [[teams]] competing in a [[game]] of quizbowl should have equivalent opportunity to receive points and ultimately win. For a game to be fair, both the [[questions]] and the [[format]] must be fair. | ||
| − | Fairness is a key tenet of [[good quizbowl]]; conversely, it is often absent in [[bad quizbowl]].</onlyinclude> | + | Fairness is a key tenet of "[[good quizbowl]]"; conversely, it is often absent in "[[bad quizbowl]]".</onlyinclude> |
In a perfectly fair setting, the team which has more [[knowledge]] should always win the game, but this would require teams to play an infinite number of perfect questions to remove all variance. Thus, things are typically considered fair if there are no structural problems which produce unfairness. | In a perfectly fair setting, the team which has more [[knowledge]] should always win the game, but this would require teams to play an infinite number of perfect questions to remove all variance. Thus, things are typically considered fair if there are no structural problems which produce unfairness. | ||
| − | ==Sources of | + | ==Potential Sources of Unfairness== |
| − | |||
| − | The standard for a fair tossup is that it is well-written and [[pyramidal]] - this allows the maximum opportunity for a team with superior knowledge to [[buzz]] first and answer correctly. Conversely, an unfair question can be poorly written, apyramidal, or both. Things like factual errors, [[hoses]], and [[swerves]] are inherently unfair, because they disadvantage teams which know more. | + | ===Tossup Questions=== |
| + | The standard for a fair [[tossup]] question is that it is well-written and [[pyramidal]] - this allows the maximum opportunity for a team with superior knowledge to [[buzz]] first and answer correctly. Conversely, an unfair question can be poorly written, apyramidal, or both. Things like factual errors, [[hoses]], and [[swerves]] are inherently unfair, because they disadvantage teams which know more. | ||
| − | + | ===Bonus Questions=== | |
| + | Before the standardization at 30 points, it was possible for different [[bonuses]] in a given round to be worth different amounts of points. A team could thus lose entirely because the bonuses they happened to receive were worth fewer points than the bonuses their opponent received. | ||
| − | + | ===Difficulty variation=== | |
| + | One major source of unfairness is [[difficulty variation]] between questions or between rounds, though it is rare that this is pronounced enough for an entire tournament to be considered unfair. This is especially pronounced when there are systematic variations in the difficulty of a specific [[category]] or specific rounds, but is also true when it affects random questions. Because there are a finite number of questions in a round of quizbowl and a finite number of rounds in a [[tournament]], unusual spikes in difficulty for certain questions or rounds can affect the outcome of a tournament. For instance, if a team that is normally strong in [[history]] encounters a packet with extremely difficult history questions, then that team might have a disadvantage for that round entirely due to the questions in that packet rather than the strength of their opponent. | ||
| + | |||
| + | Because it is not possible to write questions which are all exactly the same difficulty (and because it is not always clearly defined what any particular difficulty is), it is typically accepted that the questions in a [[set]] will take on some distribution in terms of difficulty, but that the goal should be to ensure they are on average similar in difficulty with limited variation from round to round. Enabling the Round Report option in [[SQBS]] or other [[quizbowl software]] when reporting statistics can be one way for editors and writers to get a better sense of the variation between rounds. | ||
| + | |||
| + | ===Match Formats=== | ||
| + | Some may view certain match formats in which teams are presented with different questions such as during [[60 second]] rounds or [[fanfare]] rounds as potentially unfair, especially if certain rounds are much easier than others. The lack of the ability for each team to have the same opportunity to answer the same questions can be exacerbated during these rounds. | ||
| + | |||
| + | ===Tournament Formats=== | ||
| + | There is explicit unfairness in some [[bad quizbowl]] formats. [[College Bowl]] was infamous for playoff formats which disadvantaged particular teams. [[Single-elimination]] is often considered to be unfair, with [[double-elimination]] being the de facto standard for finals of sufficient size. Other tournaments have, in the past, done things like deliberately mis-seed teams (such as putting all the A-teams in one prelim bracket and all the B-teams in another prelim bracket) or group teams by arbitrary factors (such as all the teams from one state in one bracket and all the teams from another state in another), which arbitrarily limits which teams can advance to the playoffs and/or finals. | ||
==Things that are not inherently unfair== | ==Things that are not inherently unfair== | ||
Latest revision as of 13:36, 6 January 2022
The concept of fairness is the idea that the two teams competing in a game of quizbowl should have equivalent opportunity to receive points and ultimately win. For a game to be fair, both the questions and the format must be fair.
Fairness is a key tenet of "good quizbowl"; conversely, it is often absent in "bad quizbowl".
In a perfectly fair setting, the team which has more knowledge should always win the game, but this would require teams to play an infinite number of perfect questions to remove all variance. Thus, things are typically considered fair if there are no structural problems which produce unfairness.
Potential Sources of Unfairness
Tossup Questions
The standard for a fair tossup question is that it is well-written and pyramidal - this allows the maximum opportunity for a team with superior knowledge to buzz first and answer correctly. Conversely, an unfair question can be poorly written, apyramidal, or both. Things like factual errors, hoses, and swerves are inherently unfair, because they disadvantage teams which know more.
Bonus Questions
Before the standardization at 30 points, it was possible for different bonuses in a given round to be worth different amounts of points. A team could thus lose entirely because the bonuses they happened to receive were worth fewer points than the bonuses their opponent received.
Difficulty variation
One major source of unfairness is difficulty variation between questions or between rounds, though it is rare that this is pronounced enough for an entire tournament to be considered unfair. This is especially pronounced when there are systematic variations in the difficulty of a specific category or specific rounds, but is also true when it affects random questions. Because there are a finite number of questions in a round of quizbowl and a finite number of rounds in a tournament, unusual spikes in difficulty for certain questions or rounds can affect the outcome of a tournament. For instance, if a team that is normally strong in history encounters a packet with extremely difficult history questions, then that team might have a disadvantage for that round entirely due to the questions in that packet rather than the strength of their opponent.
Because it is not possible to write questions which are all exactly the same difficulty (and because it is not always clearly defined what any particular difficulty is), it is typically accepted that the questions in a set will take on some distribution in terms of difficulty, but that the goal should be to ensure they are on average similar in difficulty with limited variation from round to round. Enabling the Round Report option in SQBS or other quizbowl software when reporting statistics can be one way for editors and writers to get a better sense of the variation between rounds.
Match Formats
Some may view certain match formats in which teams are presented with different questions such as during 60 second rounds or fanfare rounds as potentially unfair, especially if certain rounds are much easier than others. The lack of the ability for each team to have the same opportunity to answer the same questions can be exacerbated during these rounds.
Tournament Formats
There is explicit unfairness in some bad quizbowl formats. College Bowl was infamous for playoff formats which disadvantaged particular teams. Single-elimination is often considered to be unfair, with double-elimination being the de facto standard for finals of sufficient size. Other tournaments have, in the past, done things like deliberately mis-seed teams (such as putting all the A-teams in one prelim bracket and all the B-teams in another prelim bracket) or group teams by arbitrary factors (such as all the teams from one state in one bracket and all the teams from another state in another), which arbitrarily limits which teams can advance to the playoffs and/or finals.
Things that are not inherently unfair
Random chance is not unfair, though it may produce unexpected results.
There is significant debate with how fair existing rules for protests are, especially with regards to prompts. Existing rules typically award the team which won the protest with an uncontested tossup, which begs the question of whether that is an equivalent scenario to a player being prompted.