Mutual Judge Puppies

So here’s the deal. On the last TVFT we talked about MJP and raised a few questions. Here’s a few answers.

We agreed that we would simply go with best possible judges, rather than some hoo-ha about using less preferred judges in the randoms, which had been the intention (although apparently not the reality) of Greenhill. We would stick to even rankings at a lower level rather than using a 1-2. We would start assignments with the brackets on the brink with all 1-1s and then let the devil take the hindmost. Given the depth of the pool (I think the Pups did a fantastic job of providing hires, and I had a funny conversation with Marc Wallach, chief Pup of the event, who just didn’t seem to believe that this was the case—what was he going to do, bring John Marshall, John Jay, Thurgood Marshall, etc., back from the dead to make it even better?), we had the resources at hand to do what we needed to do.

So what happened?

First of all, because there are a few schools that don’t do preferences, for whatever reason, and because the field was big enough that one debater’s preference is another debater’s judge from hell, I noticed that a lot of judges who TRPC showed as the least preferred got bunches of rounds. Granted, their 1-1 status was illusory, but at least it solved the PJ Conundrum and gave rounds to those judges. Of course, they got fewer rounds than the highly preferreds, but they didn’t sit on their hands all weekend, and that’s good enough. By the same token, we did our best to give the highly preferreds all at least one round off, and we mostly did, unless they were hired by the tournament, which meant that we tried to give them a round off, but when push came to shove, they were the shovees. [Note to Microsoft: when I type shovees, I mean shovees. Curse you, autocorrect!] When it came to a situation of giving debaters 2-5 prefs or waking up a half-dead judge, the latter was the choice we were forced to make.

Secondly, in the entire set of prelims, as far as I can remember we had exactly one 3-3 pairing for a round out of contention, and no 2-2s at all. We learned early on that TRPC mostly gave us 1-1s but occasionally gave us 1-2s, and that you could wrestle your way out of some of the 1-2s but not all of them. The alternative to these slight imbalances, i.e., mutuality, always kicked in at 4-4 or 5-5. JV and I couldn’t imagine anyone preferring a 4-4 pairing to a 1-2 pairing, so after doing our best, we always had to leave a few 1-2s in the round. I would say, of the 80 or so pairings, 3 or 4 per round were 1-2s (or as TRPC generously puts it, A+ A). So, when you’re ranking next time, think of 1 and 2 as A+ and A and we’ll all be on the same page.

In other words, throughout prelims, you probably got the judge you wanted. I know that some coaches still think that MJP is the end of LD life as we know it, but I did notice that some of these coaches’s teams nevertheless do in fact do prefs. My feeling is that if you are against MJP, then going in and doing prefs is a bit of, shall we say, cognitive dissonance. (And, of course, as I’ve discussed here and on TVFT ad nauseum, the judges 90% of the time are exactly the same in my tab rooms without MJP because I always try to put experienced judges in the most meaningful rounds at invitationals. Short of totally random assignments, as at MHLs, it’s never going to be all that different, so at worst MJP is merely the beginning of the end of LD life as we know it.)

Things got a little dicier in the break rounds. First of all, we tried to release as many folks as we could: prelims ended Saturday night, and who wants to get up early to judge Sunday when their teams are out? Granted, everyone was obligated, but JV and I, in an uncharacteristic demonstration of bonhomie, decided to let some folks off the hook. Still, we had plenty of pickin’s for the doubles double-flighted round, very 1-1-ish throughout. As the day developed, however, and the pool thinned out, we had to settle for 1-2 2-1 1-1 type panels, with the numbers totally up equally for both sides. A 3-3 or two snuck in. Even a 5-5 at one point (and only one point); this was a curious ranking, as it turned out, for both debaters—the 5-5 was the squirrel.

Overall, I would say that everyone got judges they preferred, over and over again. On the technical side, it was not easy, although we get better as the tournament progressed. The thing was not so much that the software was not amenable to our optimizations, manual or automatic, insofar as the assignments were concerned, but that the number of single flights and splitting of rooms leads to a lot of tab busy work. At least we figured out early on not to assign rooms until after all the judges were optimized. But still, tabbing was much more time-consuming than when we don’t have MJP. Between the extra time to pair, and a couple of judges who were in a different time-space continuum than the rest of the tournament (all it takes is one judge with a blank ballot to stop a tournament dead in its tracks), we probably slipped an hour off the schedule we wanted to run by the time the sun set. In the end we decided to forego the seventh round in favor of a run-off for all 2-2s. Since the whole point of the seventh round was the breaking of all 2-2s, the original goal was still achieved, if not exactly in the same way. I gather that the spirit of the tournament was very much in favor of foregoing that seventh round: the cafeteria looked hell on a bad day, and felt worse temperature-wise, and we were asking, even if we were exactly on schedule, that the debaters put in a 14-hour workday.

I think next time out with MJP—the Bronx, in my case—I’ll be as fast as is humanly possible now that we’ve worked out the kinks, but that’s not as fast as not doing MJP. We might have to ask O’C to eliminate four or five of his award ceremonies to squeeze in the tournament, however, which never sits well with him. I hear that he might just hold one random prelim and then immediately break to sexydeadchemicaldecimals. Works for me!