Sophie and I are on the same wavelength here (see the comments).
We talked a lot about things like changing the speaking times on TVFT, but that scares me. Even when I hear a proposal that makes sense, I don’t trust myself. I’m not so sure my instincts are necessarily correct. It’s not that I believe that the present system is perfect, but I don’t believe that the bias that exists for neg is necessarily linked to the timing of the speeches, or more to the point, that changing the timing is the entirety of the problem, or that changing the timing will actually fix the problem. The thing is, changing speech times is a deep sea change, a change of monumental proportions. Taking such a step would be, at the very least, slow and plodding. Enormous work would have to go into making it happen, and if it did happen, even more enormous work would have to go into any subsequent changes. Such a step must be taken in total confidence. I lack that confidence. Changes in the tab room, on the other hand, don’t have the gravity of changing a speech. We can experiment and measure the results. The damage is minimal, if any. Look at how we approached MJP. We tried it at a small event, and everyone was pleased. We eased into it. Now we do it more often. We continue to study it in detail and to keep an eye on it, but it has proven a pretty good change in the way we do things. As I’ve said many times, short of totally random judging (which we do at MHLs, for instance, so I have nothing against it per se, I just think it needs to be applied to the appropriate situations), someone has to rank the judges. It might as well be the paying customers.
Anyhow, the next question I have, which I don’t know about, is the whole high-low high-high pairing thing. As a rule, we pair every bracketed round high-low, meaning that the top of the bracket meets the bottom of the bracket, then second from the top meets the second from the bottom, and so forth, meeting in the middle. But I’m not entirely sure of the philosophy behind that in prelims, or why we should never pair high-high. When I first started going to Big Bronx, the 4th and 6th rounds, if I remember correctly, we paired high-high. I think the point was to cut out any riffraff that might have snuck through. High-low seems intuitively to support the better debater. Why do we do that? High-high seems intuitively seems to pit the best debaters against one another. Why don’t we do that? As I say, I really don’t know.
We expect to TVFT tomorrow. We should address these questions.
2 comments:
I think the justification for high-low is simply that you want the better debaters to do well. It's much the same reason that in brackets, the top seed hits the lowest seed. Pairing high-high basically magnifies the problem of the strongest debaters not clearing, and in general the structure of a tournament should be set up so that the best debaters in the tournament will be most likely to clear based on natural skill alone.
The trade-off to doubling the brackets to address side bias is you compromise the ability of the power pairing to order the debaters best to worst at the end of the prelims. If you have six prelims and double (3,4) and (5,6), then you only power pair twice versus four rounds in a row normally. The lag pairing will significantly increase the variation of the ordering after six prelims (spend an hour with Excel and you can show this).
So how bad is the side bias? If you think one side wins 75/25 in evenly matched rounds, then it probably makes sense to double. But if the bias is only 55/45, then you probably hurt yourself removing the power pairing and introducing the bias of random draws to easy/hard opponents. Someone would need to do a lot of math to figure out the correct break point.
Another compromise approach would be to make sure each debater is on opposite sides rounds 3 and 5.
But if it's believed that side bias is significant then perhaps elims should not have flips or side-locks. The higher seed should get to choose sides and thus not have to compromise their earned bracket position by the bias.
Post a Comment