Monday, June 13, 2005

Maha!Quizzer - Bombay Round



Venue : Sophiya College,Mumbai
Date : 12th June, Sunday.
Time : 10 to 11:30 am
Organizers : Karnataka Quizzing Association (KQA)

Maha!Quizzer was a unique format consisting of 150 written questions to be answered in 1.5 hours at the same time at 5 locations around the country(Bangalore, Chennai, Hyderabad, Trivandrum and Mumbai). There were prizes for local winners and the title of Mahaquizzer was to be bestowed upon the person with the highest score in all 5 cities.

The only problem was that when a quiz is conducted across geographical boundaries care needs to be taken about the scope being truly fair to people from all the places involved. There were many questions which were very south-centric - for eg. we were asked about a certain mountain range you pass while going from trivandrum to bangalore or something like that. Another complaint was the sport. All the questions involved niche sports and their rules. There were no cricket-nostalgia questions or sport questions that did more than scratch the surface. Most questions were like - "this sport involves an arena of 2 x 15 m and is governed by ...."

The high point of the 'quiz' were the hindi film questions. They were really well set. Also lots of questions on literature-arts-architecture-language. Some of these questions were really good but some however bordered on the arbit. Overall a written 'exam' is never fun and KQA needs to come up with a better format if they really want their Mahaquizzer. Maybe take some tips from Siddharth Basu and Co. at UC.

(Also KQA seems to be getting afflicted with the problems of big associations. When you hear the words 'it's a policy decision not to hand out empty elims sheets' you know that something is wrong)

Friday, June 03, 2005

Anand & Vibhendu's Elims classifications

Anand and Vibhendu did some preliminary classification of *Elimination* rounds from four quizzes: Chakravyuuh 2005, JBIMS, PiQue 2005 and Quiz-O-Mania 2005.

After some application of MS Excel pivot chart magic (thus recouping their MBA investments), they have the following graphs to show. The base data is available with me in case anyone's interested. Each image has been preceded by a small note from A & V on how they arrived at those classifications.


Category

[Anand]: Simple [classes], we used a school subject kinda classification consciously trying to nail everything: less on Misc. etc, Social Sc includes History, Geo, Politics, Eco etc and hence appears to have more weightage. (No entry for Yoga, SUPW (Ramanand's note: SUPW stands for "Socially Useful Productive Work" (and has other variations too)))
[Vibhendu]:Mythology, etymology are clubbed under Lit for most cases.


CategoryPosted by Hello

Provenance

[Anand]: Sadly, this is divided into India & Western only--we have left out further subdivisions as that domain belongs to Llosa, Jospin, Chomsky etc--so a Home or Away format.


ProvenancePosted by Hello

Quality/Originality

[Anand]: The criterion used is if we feel it has appeared before in the same form (i.e. worded the same circa John Sutter), or is a basic regurgitation of a standalone fact that has appeared in the same form ( as in the q), or is derived from a fact/incident well chronicled.( This is for Banality)( Like also Lara Croft's stats.) Originality - anything that we felt required the QM to drill down on facts, contort incidents, derives from a reworking of known or even unknown events, and we felt was essentially framed by QM solely for the purposes of asking.( i e the SRT Clapham q, chimes )


QualityPosted by Hello Summary

[Vibhendu]: The classification was done in less than ideal conditions, some of it on our way back from Pune. Hence most of it was done on-the-fly and has not been revisited even once. The exercise is mostly exploratory in nature and was never intended to be a critique of the quizzes, hence no classification on the lines of good, not so good and bad etc. Provenance & category classification was again done to detect any mojor trends in the present day quizzes, can be improved with certain sub-topics institutionalised as independent categories by themselves;-) As for the contentious issue of originality-banality classification, the criterion is here the thought process followed by the question solver rather than the qn-setter. Chestnuts are classified as banal straightaway. Awell known fact making a quizzing debut may still run the risk of being dubbed as banal. e.g, north eastern capitals, new jalpaiguri & Geeta Dutt. Noble intentions behind incorporating quiz-virgin topics, notwithstanding ;-) Again, if the question solver is aided by an existing thread/chestnut, the qn could have been branded banal. e.g. A connect qn on last Male Singles Grand slam winners from the host nations finding an echo in a Ladies singles winners connect.


SummaryPosted by Hello

Regular disclaimers of errors in judgement, subjectivity and lack of consistency follow. Surely, there is enough scope for refining.The reason behind bringing it up was that it can be used as a template/suplement when you do a classification exercise of your own.
:: Anand and Vibhendu



Wednesday, June 01, 2005

On Fairness - Hirak's Blade

After the recent series of posts on the maddening methods, one of our old prabhaaris (the man who almost had the Inquizzition quiz named after him and the orator of the lost Mood-I monologue) Hirak has this to say: (Hirak, pardon me for the usual digs and for including the note before your actual "post" - I included it because I found it relevant)


I was reading the latest series of posts and I think that this discussion is in serious need of Occam's razor. I posted a rather long comment and I have removed it and I request you to post it on my behalf.

You had started a thread about what question setting is all about and that discussion has got subverted on the way. It seems to me that this discussion on systems is quite futile and even if someone does simulate them, the core assumption is that ALL questions will be fair and equal.

Such an assumption can never be made so every simulation suffers from a serious flaw. I, like you, hold a good question very close to my heart and I wish we talk about that more. I also hope that people on notesandstones still believe that quizzing is about fun and learning and not about showing how much you know.

I am currently in stuck in a Third World country (when it comes to quizzing), where quizzing is a rare sport. I am quite jealous of the fact that you guys have enough quizzes to discuss the best quizzing method. Filled with nostalgia here are my two paisa thoughts and a few axioms which which do not seem to be universally agreed anymore.

The discussions have been interesting and the different ideas were refreshingly original. Still, no method has, bar IR, has been tested or simulated. With each correction for unfairness there is an increase in complexity, and hence the chances of adoption of that method decrease.

My hypothesis is: The simplest and best method for fairness will converge to a written final.
It has 'ALL' the elements of a good system:

1) All teams get the same number of attempts.
2) Teams get to attempt every possible question set by the QM.
3) Points in direct proportion to what each team knew.
4) Also takes care of PMQ's and other artifacts of question-framing since everybody is on the same ground.

Of course, despite its technical strengths it is universally rejected because it fails in one vital respect which is 'lack of drama'.

As Kunal noted and which I rephrase as: Quiz Axiom 1: "Quizzes are slightly unfair competitions."

IR is not fair but slightly unfair. A really good system should have the slight unfairness. It is that slight unfairness or element of chance that adds drama and allows a slightly weaker team to win. I still am a proponent of a few buzzer rounds and joker rounds in quizzes.

Quiz Axiom 2: "Quizzing is more about fun than mere regurgitation of facts."

However, fun should not be at the expense of complete unfairness. Most will agree that in essence unfairness has less to do with the 'system' than the kind of questions, and how they are framed. Quizzing should be about the questions and not about the system.

Quiz Axiom 3a: "Quizzers are cribbers"

My chief grouse about quizzes that I lost was there wasn't a balance of topics and some questions were really vague and had no 'Ahaa' value. (Quiz Axiom 3b: You don't complain when you win)

I hope you all agree that we need to talk about content rather than method.

I also believe, that given a reasonably balanced quiz, a good team, regardless of the system, will more often than not finish in the top 3.

Quiz Axiom 4: If you are good, you are good.
OR
You can't fool all the people all the time.

:: Hirak