So I learned something (academically) new on Facebook: Tapered Scoring

To members of administration and others: there are actually interesting academic discussions that take place on Facebook. A few days ago, I learned about one of them: tapered scoring in debate tournaments — particularly British Parliamentary Debate tournaments.

In a nutshell, here’s the basic argument: all rounds don’t count equally when it comes to deciding who makes it to elimination rounds at a debate tournament. And certain types of resolutions privilege certain kinds of debaters.^[1]I remember judging a round about nuclear power at one NPDA Nationals where the debater happened to have a summer job in a nuclear power plant. I learned more about pebble-bed nuclear reactors than I … Continue reading So to make things more fair, in a tapered tabulation system, earlier rounds of the tournament ought to be weighted more strongly than later rounds of the tournament.

To find out more, here’s the link to the Facebook group: https://www.facebook.com/groups/1122952398139491/

I’ll admit, I haven’t read all the guides that are posted on the Facebook group, but the arguments are interesting and do bring some quantitative data to what debaters and coaches have wondered for years: is it better to start off at the top of the bracket and lose, or is it better to start off poorly and work your way back up the bracket? ^[2]There is an old Contemporary Argumentation and Debate article, if memory serves, that also addresses this question in light of trying to decide whether 6 or 8 round tournaments are better at sorting … Continue reading At tournaments where each round has a winner and loser, the question is relevant, but it’s even more relevant in British parli, where each team is ranked from 1-4, and not all teams with a certain score will necessarily clear.

It should be noted that even in British Parli, there are concerns about how determinate speaker points are to determining who breaks into elimination rounds. That’s one reason why in policy-type parli tournaments, and policy debate in general, there’s been more of a desire to have “partial” elimination rounds, where all teams with a winning record are guaranteed to advance to the elimination round. So if there are 23 teams that have a winning record, 23 are advanced, even though that’s an odd number.^[3]There are ways of determining how many byes and debates happen in those circumstances, but thankfully it’s pretty straightforward. If you have 23 teams, the first round of elims needs to get back … Continue reading

The first of the papers arguing for tapered scoring, which can be found at Researchgate, has multiple suggestions for tapering, including tapering by day (at World’s, there are 9 rounds and 3 days for prelims). By tapering, the paper suggests that early rounds should count double or even triple, while the last rounds of the tournament are weighted normally. The paper’s contention is that the last round only really contributes 3% of the variance as to whether a team makes the elimination rounds or not, instead of what we would expect of one round out of nine: 11-12%.

One of the assumptions made by the previously mentioned paper is that the team with the highest speaker points in a lower bracket is almost always better than the team with the lowest speaker points one bracket up.^[4]To give a practical example, a team with a 6-2 record with 400 points is likely lower-skilled than a team with a 5-3 record and 425 points, even though the 6-2 team won one more debate. That assumption in and of itself could use some more research, but let’s say that it’s true. Critics who say that speaker points are arbitrary don’t really have a good response to that assumption. Not all 6-2 teams are equal, and not all undefeated teams win tournaments. Just look at the NCAA basketball tournament. However, the authors do show that at the World University Debating Championships, that in the lowest bracket that typically advances into elimination rounds, the team with the most speaker points typically advances fairly deep into the elimination rounds (often semis or finals). Speaker points serve a function, albeit with some variance that we don’t often admit.^[5]In other words, some team seeded 7th could really be seen as a team that could be anywhere from 5th to 10th, because of the randomness of the judges that they’ve received. Even that might be … Continue reading

I think the “so what” here is that it’s worth examining mathematically whether these assumptions hold true in sufficiently large^[6]I’m not sure what this number is — it’s certainly greater than 30, but how much so? My initial thought is at least 2 to the power of the number of rounds. So, for a 6 round tournament, 64 … Continue reading win/loss debate tournaments. Is there a particular round that holds undue influence when it comes to traditionally judged debate tournaments?

One other assumption made in the paper is that judges who give substantially similar points are more likely to be considered as “strong” judges. In other words, there’s an inherent belief in the British Parli community that given the commentary period after the round, judges will give speakers similar amounts of speaker points. And indeed, I have seen some tournaments even have tests where a potential judge watches a round, and then has to give a critique and points that are reviewed before the judge can be seen as certified. But to a certain extent, there’s also a sense of reinforcement of certain judging paradigms involved: a judge who is already favored can then compel other judges to follow their norms. And in a way, that’s already a problem when it comes to several types of competitive speaking.^[7]This has been well documented in the policy debate community, where white men are preferred far more than other types of judges. There’s also evidence of this in NPDA-style debate, where judging … Continue reading ^[8]Indeed, we have a name for judges who differ from the norm: squirrel judges. I may have played that role a time or two… or a hundred during the course of my judging career. But then again, who’s … Continue reading

While I’m not sure that tapered scoring solves for the problem of some students doing well with certain kinds of resolutions unless all resolutions follow a common research area^[9]Which in and of itself leads to some critiques on the part of “traditional” parli debate coaches because announcing a topic area takes away a portion of the extemporaneous aspects of parli debate. , this is a theory worth some further analysis, and will likely be the subject of further posts on this website.