Bonus part conversion data

[Alejandro]

We can gather conversion data about tossups in a straightforward manner, but we cannot do the same for bonus parts because we only record the bonus total. If scorekeepers kept track of which parts were answered we could get bonus part conversion data which would help determine the actual difficulty of bonus parts and potentially reduce scorekeeping errors during a match.

We can find out the conversion rate for tossups by going through the scoresheets and counting how many times a question at a certain number was answered in one round. We cannot do this for bonus parts because scoresheets only have a column for the bonus total. We may be able to make a good estimate of the bonus part conversion rate by looking at the bonus total and assuming that a total of 10 means that only the easy part was converted, a total of 20 means that the easy and medium parts were converted, etc. However, this would require looking at each bonus and determining which parts were easy/medium/hard, which is a time-consuming process. This estimate would also fail to catch problems with misjudged difficulty (e.g. the hard part was easier than the medium part).

Bonus part conversion data could be gathered if scorekeepers tracked which parts were converted. This could be done by simply leaving checkmarks or Os for correct parts and Xs for missed parts. Because most tournaments only allow three-part bonuses we can modify existing scoresheets to include 3 columns for each bonus part. The bonus total can then be calculated from the marks and placed into another column. This should reduce scorekeeping errors because it is easier to verify which parts were converted.

There are some possible issues with this system: it may take too much time if the moderator is also the scorekeeper and there is no way to add this data to SQBS. However, the data that can be gained by including this information should outweigh these concerns.

[Jeremy Gibbs Paradox]

Perhaps if someone designed a scoresheet where the traditional column for listing bonus points was divided into thirds so the scorekeeper (or moderator/scorekeeper as the case may be) could merely check off which third was appropriate for which bonus part was answered?

[dtaylor4]

This has been tried before, I think Brown tried this with one of the EFTs.

The issue is adding the extra space to a scoresheet while still making it useful for SQBS entry. Honestly, I think that 8.5x11 is too small for most tournaments' scoresheets, with the variants involving powers, negs, rebounds, etc.

Scorekeepers need to be told to keep the sheet CLEAN. Scoresheets should be filled out in pencil, this way when there are errors/issues, they can be fixed easily while making the scoresheet still legible. Handwriting, I've all but given up on.

[Marble-faced Bristle Tyrant]

Actually, a lot of high school tournaments I've been to recently (which is really just GATA tournaments plus Collins Hill JV) have had three "10"s in the bonus columns that you circle or cross out accordingly. Whenever I'm keeping score on one that just has a blank for bonus conversion, I keep track of each part in the margin with "X"s and "O"s, though it's mainly so I can remember how many points have been scored.

[Skepticism and Animal Feed]

I would classify this kind of data as "nice to have" as opposed to "must have". If I were a TD and somebody proposed to me that my tournament should track this data, I would be very concerned that collecting it would slow things down, would lead to many scorekeeping errors, and would just distract from my #1 goal of providing a quick, profitable tournament that is enjoyed by as many teams as possible.

In the not too distant future, when we all have iPads and will keep score using apps, I'm sure we will be able to quickly and painlessly collect this data (perhaps even the more valauble data of when somebody buzzed). Until then, I am content to channel Joe Morgan and determine how hard a bonus is by looking at it.

[Stained Diviner]

Speaking as a writer, I am not sure how useful this information would be. At the end of each tournament, you would have information on 600 or more different parts, which is a lot to keep track of unless it's put in some useful database. Even then, you have to keep in mind that the conversion depends on the field strength, and with a small amount of data it could be impacted by which teams happened to hear that bonus.

Also, the difficulty of an answer within a bonus depends on the placement within the bonus. For example, a bonus with first part Flaubert and second part Madame Bovary, and another bonus with first part Madame Bovary and second part Flaubert are each going to get better conversion on the second part than the first part. Also, while it's generally a safe assumption that a tossup in a decent high school set will have a giveaway saying one or more of the best known things about the answer, that's not a safe assumption with bonuses. I've seen well written bonuses where the hard part was Mark Twain or France. Those exact bonus parts probably shouldn't be repeated as is no matter what the conversion data is.

All that being said, I guess a list of bonus parts with 0% conversion would be worth seeing. As Bruce explained, however, I don't know if it is worth the effort.

[mtimmons]

Perhaps more useful than data on individual bonus parts would be data on conversions of entire bonuses. A number of tournaments seem to collect data on tossup conversion but I haven't seen any tournaments that release data on what percentage of teams got 30 or 20 or 10 or 0 on individual bonuses. This has the advantage of already being on regular score sheets and preserving the connectedness of bonuses.