Return to site

How do we decide who debates who? - 辩论比赛中对阵是如何产生的

John Harper

· Debates

Fifteen minutes before each debate round you find out who you will be debating.  It may look like a random process or that we pick your team names from a hat.  That is far from the truth. In this post I’ll explain some of the complex art behind making sure that you are debating an opponent who will both stretch you as a debater, and lead to a fair competition.

在每轮比赛开始前的十五分钟,你会得知自己将要和谁辩论。你可能会觉得匹配对阵是一个随机的过程,或者是我们从一个放满了队伍名字的帽子里抽出了其中一张纸条。这当然不是真的。在这篇博客里,我会向你解释对阵匹配背后复杂而精巧的“艺术”,这个环节使得你能够在辩论中遇到合适的对手,充分锻炼你作为辩手的能力,同时也确保了比赛的公平。

We call the process of deciding who you debate and with which judge “pairing.”  This is a very complex process because of the number of possible opponents and the data necessary to do it accurately, so we use specialised database software to generate the results. One of the most important elements of this system is “Power Pairing”, this process makes sure that the debates are as fair as possible.  

我们把决定你和谁辩论、被谁裁决比赛的过程叫做“对阵匹配”。因为可能出现的对阵情况的数量繁多,以及数据一定要准确无误,这使积分的过程变得非常复杂。因此我们采用了特殊的数据库软件来产生匹配结果。在软件系统中最重要的部分之一就是“按实力匹配”了,正是这个过程确保了每场辩论都是尽可能公平的。

Generating debate pairings

队伍匹配的产生过程

There are two ways to pair a round: randomly or power paired.  In a random round we use a random number generator and any team could debate any other team. At NHSDLC tournaments this happens during the first two rounds.  It’s not totally random though as there are a few constraints: that teams cannot face the same opponent twice, and cannot debate against their own school, unless that school makes up more than ⅓ of the whole competition.  These randomly paired rounds means the competition is fully ‘fair’ as anyone can debate anyone.  So if this is your first competition you have just as much opportunity to debate against the regional champion or against someone else who just starting out.

对阵匹配有两种方式:随机匹配或按实力匹配。在随机轮次的比赛中,我们会使用一个随机数字产生器来决定对阵,并且辩手们可以对阵任何一支队伍。在NHSDLC的比赛中,预赛的前两轮采用随机匹配。其实随机匹配并不是完全随机的,在匹配的过程中有几个限制:一支队伍不能和相同的辩手辩论两次,不能和来自同一学校的队伍比赛除非该学校的参赛选手人数占总人数的三分之一以上。这些随机匹配的轮次意味着比赛是完全公平的,因为任何人都可以互相辩论。因此,如果你是第一次参加比赛,你在比赛中遇到地区赛冠军和遇到同你一样的新手的几率是相同的。

However this randomness wouldnt always tell us who was the best team in the competition. Let's look at two situations, both teams have one been to one tournament before:

然而,这的随机性并不能告诉我们哪些是比赛中最强的队伍。让我们来看看下面这两种情况,这两支队伍都参加过一场赛事。

Debaters Louise and Clark Over 4 rounds they won 3 and lost 1 round.
在辩手Louise和Clark参加的四场比赛中,他们赢了三场,输了一场。

Debater Tom and Jerry, Over 4 rounds they won 3 and lost 1 round.
在辩手Tom和Jerry参加的四场比赛中,他们赢了三场,输了一场。

Considering these two teams, they have had very different tournaments.  Team 1, Louise and Clark have had a much easier tournament than Tom and Jerry, who have faced harder opponents. But if you were to look only at their scores they would seem to have scored the same. So in order to make sure that everyone has a ‘fair’ tournament, and that the best scoring teams are also the best debaters, we instead make sure that people face opponents who are equally as strong as them.


这两支队伍而言,他们经历的比赛是完全不同的。第一支队伍Louise和Clark的比赛比Tom和Jerry的比赛容易多了,因为后者所面对的对手更加强劲。但如果你只看到了他们的胜负结果,他们取得的成绩是一样的。所以,为了确保所有人都能有一场公平的比赛,并且成绩最好的队伍也确实是最强的辩手,我们会保证选手们会遇到和他们实力相当的对手。

So how do we power pair?

所以,我们怎么进行实力匹配呢?

In order to get the best of both systems, at a normal 4 round regional we will randomly pair round 1 and round 2, however round 3 and 4 will be based on the results of your previous debates. So if you win your first two debates, you will debate another team who has also done that, the same for those who win one and lose one and for those teams that lose both.  

为了达到最佳的对阵匹配效果,我们在一般的地区赛(4轮预选赛)中会对第一轮和第二轮进行随机匹配,然后根据辩手在之前轮次中的比赛结果在第三轮和第四轮中进行实力匹配。所以如果你在前两轮的比赛中都获胜的话,在之后的轮次中你的对手也会是有两胜记录的队伍。对其他那些胜一负一或两轮皆负的的队伍来说,他们也会遇到与自己战绩相同的对手。

Considering these two teams, they have had very different tournaments.  Team 1, Louise and Clark have had a much easier tournament than Tom and Jerry, who have faced harder opponents. But if you were to look only at their scores they would seem to have scored the same. So in order to make sure that everyone has a ‘fair’ tournament, and that the best scoring teams are also the best debaters, we instead make sure that people face opponents who are equally as strong as them.

这两支队伍而言,他们经历的比赛是完全不同的。第一支队伍Louise和Clark的比赛比Tom和Jerry的比赛容易多了,因为后者所面对的对手更加强劲。但如果你只看到了他们的胜负结果,他们取得的成绩是一样的。所以,为了确保所有人都能有一场公平的比赛,并且成绩最好的队伍也确实是最强的辩手,我们会保证选手们会遇到和他们实力相当的对手。

So how do we power pair?

所以,我们怎么进行实力匹配呢?

In order to get the best of both systems, at a normal 4 round regional we will randomly pair round 1 and round 2, however round 3 and 4 will be based on the results of your previous debates. So if you win your first two debates, you will debate another team who has also done that, the same for those who win one and lose one and for those teams that lose both.  

为了达到最佳的对阵匹配效果,我们在一般的地区赛(4轮预选赛)中会对第一轮和第二轮进行随机匹配,然后根据辩手在之前轮次中的比赛结果在第三轮和第四轮中进行实力匹配。所以如果你在前两轮的比赛中都获胜的话,在之后的轮次中你的对手也会是有两胜记录的队伍。对其他那些胜一负一或两轮皆负的的队伍来说,他们也会遇到与自己战绩相同的对手。

This means that for example: If you have won two debates already, then in order to to win a debate in round 3 you have to beat another team who have won two debates. So over the course of the tournament the teams will be sorted by ability, as getting each extra win is progressively harder.

这就意味着:假设你前两轮都赢了,如果你想赢得第三轮你就必须打败在前两局同样获得全胜的队伍。在整个比赛过程中,队伍会根据实力进行匹配,也就意味着越到后面赢得比赛的难度就越大。

So for example if Louise and Clark won three debates to get a fourth win they would have to beat a team who had also won three debates. But if Tom and Jerry lost all 3 of their debates, then to win round 4 they would only have to beat a team to had also lost all of their debates. This means we can be confident that teams with a higher number of wins are better than those with a lower number, and didn’t just get lucky with their opponents

拿Louise和Clark举例来说,如果他们已经赢得了前三场比赛,要想获得第四场胜利他们就必须打败另一支有同样战绩的队伍。如果另一支队伍Tom和Jerry输掉了前三场所有的比赛,那么他们在第四轮面对的对手也会是三轮皆败的队伍。这样我们就能确保获胜次数多的队伍一定比获胜次数少的队伍水平高,而不会是偶然的结果。

Now you might notice that this only partially sorts the tournament, as there will often be more than two teams with a certain number of wins. So in order to sort within the group of teams who have the same win/loss record (called a “bracket”) we use the combined speaker points of the team.  Speaker points are also used when the number of teams in each bracket cannot be evenly divided into rooms, e.g. there are 11 teams on a 2/0 record, and 9 teams on a 1/1 record, so the leftover teams in each bracket would be paired against each other.

读到这里你可能意识到这种方法只能匹配部分队伍,因为有同样获胜记录的队伍不可能只有两支。所以为了能够在有相同获胜/落败次数的这些队伍(我们叫“一组”)中进行更合理的队伍匹配,我们会结合两位辩手的辩手得分来进行队伍匹配。如果当这一组中的队伍不能按照两两对阵的方式形成对阵匹配,我们就会使用辩手得分,比如:2胜0负组有11支队伍,1胜1负组有9支队伍,那每组落单的队伍就会根据辩手得分来进行匹配。

As you might guess, collecting all of the necessary information to generate these pairing between each round is quite time consuming. So much of our staff’s time at tournaments is taken up with making sure all results and ballots get in to our tabulation room on time, that results are inputted and double checked. Fortunately we can manage it through hard work, dedication and huge amounts of caffeine.

你可能也猜到了,要想在每轮都收集这些关键信息并形成对阵匹配需要非常多的时间。其实这也是为什么我们的学术导师们在比赛的大多数时间要确保所有的结果和评分表能准确的到达计分室,然后这些结果会被输入电脑并进行在此确认和检查。幸运的是我们靠着紧锣密鼓的工作节奏和大量的咖啡能够顺利地完成这项工作。

Elimination rounds work on a slightly different system, which we will discuss in a later blog post. Hope you found that an interesting insight into how things work behind the scenes. If you have any other questions feel free to post them in the comments.

而对于淘汰赛来说有略有不同,我们会在下一篇的博客中为大家进行介绍。希望你能通过我们的文章进一步了解比赛幕后的故事和原理。如果你有任何问题,可以在文章下方进行评论。

All Posts
×

Almost done…

We just sent you an email. Please click the link in the email to confirm your subscription!

OKSubscriptions powered by Strikingly