ACF and Powers (POLL)

[Carlos Be]

One problem with quiz bowl writing is that we cannot see how our tossups play every time they are read. To (partially) resolve this issue, some sets collect data on buzz-points. Unfortunately, collecting buzz-point data takes a lot of work, and the data is rarely analyzed in depth. Luckily, there is an alternative: simple conversion statistics. For sets where recording buzz-points is infeasible, such as ACF Nationals, conversion statistics may be valuable. One issue remains— without powers, simple conversion stats don't tell you anything at all about where a question is converted.

If ACF had powers, simple conversion statistics would be a lot more valuable. For instance, if power marks are fairly late (say ~70% through) and still only 2-3 teams power a given question, then we would know that the clues are far too hard. This information may not be reflected in powerless conversion statistics, since hard questions often cliff near the giveaway. On the other hand, if power-marks were earlier (say ~40-50% through) and questions consistently have 5-6 powers, then the set editors could shut down any player complaining that the set is too hard with facts and logick.

Admittedly, power-marking is not trivial. Estimating the point in a question by which a given number of will buzz is very difficult. But writers should be estimating how many people will buzz at a given point in a question whether the question is power-marked or not, so the additional effort of labelling one of these points is not too significant.

If you're afraid of giving extra points for power, then these powers do not need to be worth extra points. The point is to record objective information so that we can say things about a set that aren't biased by whatever biases are biasing things. For this reason I recommend that ACF should have powers.