Revising 3rd/4th/5th Person Bonuses

[josiah-leinbach]

We reward a 3PB because it is difficult to achieve, but that reward is the same as for the even more difficult feats of getting a 4PB and 5PB. Therefore, I propose that we amend 6.1.2.1. Points Earned to read:

6.1.2.1. Points Earned +10 points for the 1st correct answer given by the 3rd quizzer on the team +20 points for the 1st correct answer given by the 4th quizzer on the team +30 points for the 1st correct answer given by the 5th quizzer on the team

The advantages of these changes are similar to those which I have outlined in Issue #199, except that the additional points would not count toward individual scores, so the status quo would remain the same for that.

Under this revision, a question could be worth (in theory) 50 points if a 5PB is achieved. Is 50 points a lot of points? Certainly. But how difficult is it to get 5 quizzers to all get questions correct in the same quiz? Incredibly. Difficulty should be rewarded. If a team is down 50 points on Question 20, has already had four quizzers answer questions correctly that quiz, and their 5th quizzer gets Question 20, that is beyond clutch and should receive a higher compensation than when the third quizzer answered correctly.

This sort of non-linear scoring system would also make late-quiz decisions far more exciting and give coaches additional strategic options, as outlined in Issue #199 as well.

[24RMiller]

After seeing a rule very similar to this being used at 2 major tournaments this year, I am in favor of this rule change. I would not be in favor for the 2pbs that have been tested, but I think the suggested point values above are the right balance. 4pbs and 5pbs are extremely rare, so the effects would be very minor, but it would give opportunities for larger comebacks and provide a slight boost to deep teams. Admittedly, it does mean it is far more difficult to "put a quiz away" when a team could get 90 points on the last two questions, but it would make the ends of quizzes more exciting, rather than one team just sitting the final question or two knowing they can't lose their 50 or 60 point lead.

[ShaileneCCD]

I like this rule idea. I think it encourages a focus on team based quizzing rather than a team carried by one or two individuals. It also would add a very exciting element to the competition level. Like mentioned above, there is a risk of a team accumulating many points early in a quiz (depending on the depth of the team) that it could make other teams discouraged because they'd never be able to catch up. However, I think it could also push individual quizzers to "pull up their socks" and contribute to their team because 1 question could mean significant team points.

[josiah-leinbach]

There are two reasons I favor adding the 2PB, as we tested it at Heartland:

1. On the whole, it actually levels the competition a bit for small but strong teams. In the simulations we did beforehand, and from the tournament data, the 2PB offered strong teams of 2 or 3 (especially 2) a marginally better chance to compete with larger but less strong teams, who could get 3PBs and more. A larger but less strong team will benefit more from the "Progressive Bonuses", since a 3PB is worth more than 2PB. However, a smaller but stronger team of 2 will earn more 2PBs over a tournament than the larger, less strong will earn 3PBs, so the probability-payouts are fairly equal.
2. It adds consistency. Why reward team contributions only after 2 quizzers have answered a question? Plus, statistically, starting the 2PB with +10 and incrementally raising the PBs by 10 follows the underlying probabilities of achieving those outcomes. 2 quizzers on a team getting correct questions is less likely than 1 getting a question correct; 3 quizzers getting correct questions is less likely than 4, etc.

Empirically, the 2PBs at Heartland made quizzes slightly more competitive when one team with a strong quizzer but a weak team quizzed against another team with 2-3 moderately good quizzers. The first team's top quizzer would usually quiz out, and some combination of 2 quizzers from the second team would get questions. In those cases, the 2PB gave such teams a marginally better chance to win. On the other hand, all that first team had to do was have another quizzer get one question and they could all but ward off that second kind of team.

[JoshJetto]

I am not in favor of changing the current xPB rule with progressively higher values in the Internationals rulebook primarily for one reason (with a couple other reasons that follow). 1) By incrementally increasing the point value of additional quizzers answering questions, you devalue the competitive advantage and reward of the study time put in by the top quizzers by decreasing the relative value of their contribution to the team score. For instance, if a single quizzer on a team quizzes out perfectly, they raise their team score by 90 points. If a team has 5 quizzers each answer one question correctly under the proposed progressive xPB rule (from 2PB-5PB), then as a team, they will raise their score by 220 points. Not only that, but it makes the contribution of the fifth person equal in value to a quiz out! In other words, it takes the value of studying enough to achieve a quiz-out at Internationals and makes it equivalent to the value of studying enough to answer one question in a quiz. This is kind of like Marxist equity (without the class-hatred and atheism), and the effect of this rule would be ultimately demotivating for the top quizzers at Internationals - and I don't think it's wise to create rules that reduce the reward and motivation for studying by rewarding less study and less skill at the International level. Ultimately, I don't think it's fair to have such a drastic reward system for additional team members scoring. 2) Under our current rules, a team that achieves a 5PB is already well-rewarded, but not wholly untouchable in terms of another team being able to catch them. It's much more doable to catch a team that has scored 150 points vs. 240 points. And again, I think it's a good value in quizzing to reward the top quizzers for their study efforts by enabling their per-question contribution to the team to be roughly equal to that of any other quizzer on the platform. 3) It greatly offsets the current risk-reward framework for erring vs. answering questions correctly. A -10/+30 risk-reward is, I feel, a good balance, but a -10/+70 one is much too high, in my opinion, and it will lead to one of two things (or both of these things in this situation). Reckless jumps by the team that stands to benefit 7x (or 6x or 5x) more from a right answer than they stand to lose - making it worth it to take three risky jumps to get a unique word on the first word of the question - or reckless jumping and question burning on the part of the other teams trying to defend against such a devastating single question blow to their chances of winning. In either case, it makes the quizzing a bit nonsensical, and, I think, does a disservice to the questions by making the reward so great that teams are strategically "forced" to throw caution to the wind. And especially at the International level when the quizzers have spent so much time preparing, they should not have their hard work and the gains of their answering questions correctly be able to be swallowed up by a single question or even two.

From a rules change standpoint, I would be okay with making the 5PB +20 instead of +10, but I would want to leave all the other PBs the same as they currently are.

[l-koral]

I am personally in favor of this revision since it allows for more incentive for quizzers who do not normally jump in a given quiz. It is very rare that a team is more "stacked" than other teams and if that is why some are worrying, I would advise that teams be chosen carefully to avoid one team running away with a fifty point question late in the quiz.