At some point in the book, I have to discuss splitting your team's strength, taking some members of your A team and placing them temporarily on separate teams. There's some good reasons to do this, and some bad reasons as well. This discussion is separate from doing it in practice, which you should be doing anyway to figure out the holes in your team. It demands you have multiple events ahead of you, and enough players for events for multiple teams to enter. You do this in a game situation because you want to test your team against real opponents, and you do this if the team can benefit from trying this while actual statistics are being kept. [Since I would advocate that at least for the first week, you keep copious statistics in practice, this technique applies in practice well after the first week.]
Why to split strength:
The positive cases for splitting strength revolve around either gathering more information about how your team interacts, or what your teams' needs will be in future.
If you're passing on particular non-subject skills, like captaining, that depend on game pressure.
If you have a second player on your A team who wants to learn how to be the team captain, it's not a bad idea to split strength and give them an opportunity with a lineup they aren't familiar with as their teammates. The new captain has to learn on the fly important skills, like listening to their teammates, waiting to offer their answer and generally keeping their teammates involved on all bonus decisions.
If you're trying to give a bunch of new players guidance on how they work as a team, putting a complementary piece to their skills in the captain's chair can help them become a better team.
This is the flip of the previous. If you're trying to encourage a new set of players, giving them a captain whose goal is not to pad their own stats but to encourage the team to contribute to the team is a pretty good plan. I'll give the guidance that I got from watching players better than me captain: They always pause before answering to see if anyone has suggestions, then offer their possibility to the team, but offering full intonation in to their confidence in their own answer. They never give certainty without asking for confirmation, and genuinely encourage their teammates to offer answers up.
My final year in college, we did this by sending three teams to Harvard's early December tournament, putting a grad student on each team to captain, with the express goal of getting each team member to contribute and join in the decision making. When we ended up placing all three teams on the same side of the elimination bracket, we ended up with intramural atmosphere, and knowing all twelve of the players were improving from the experience.
If you want to highlight the holes that a team has without a certain player.
This demands a lot of knowledge of your team and their limitations. This is also a variation of going solo. It's a little mean if the player isn't aware of the situation and willing to do this. The team knows they're weak in a category, and every time the points slip past them, it's a recognition of that hole biting at them. The lesson they're learning in this is the importance of the hole and how damaging it is. And if you do this, the now motivated team should be given lots of study materials to patch the hole.
If you want to establish "who is the replacement" for a particular player.
This is something to consider when you get near the end of the year and you've already qualified for everything you're going to qualify for. Subbing in underclassmen for your graduating players can tell you something of what categories your team will struggle with next year. It also becomes a way of figuring out what important pieces of knowledge haven't been transferred from your graduating players to the rest of the team, and gives you a list of things for knowledge transfer in the near term and for you to teach the team in the long term.
To see if there is a subject shadow effect where two players have the same knowledge, but only one answers the questions.
A shadow effect occurs when two players on a team know a particular subject, but one has more experience with earlier clues, and buzzes in correctly. This means that second player's knowledge may not be accounted for, because it is overshadowed by the first player. Moving the second player to a team without the first player allows the team to know that the second player has knowledge in a particular field.
This is of limited value in game situations, because the coach will not observe all the matches directly.
Situationally to offer the team overall the best chance to succeed.
This was a situation which occurred my last year in college, the tournament at Swarthmore was late in the calendar, and interfered with spring break, so most of the team was off-campus, and though we had written two packets and had two teams, we only had five players available as we drove out Friday. In this case, because we couldn't give the hosts notice in time to cancel the team, I played solo, and the other four players, all in their first year, played as a team. This offered them the best chance to compete as a team. I snuck into the playoffs as the 4th seed in my pool on a tiebreak, and they just missed as the 5th seed of their pool.
This specific example is highly unlikely to ever happen again, but if you have to split strength because of scheduling, you do want to give ALL your players the best chance to succeed, not just your A team.
Why not to split strength:
The general case of these cases when you should not split strength can be summarized "don't be a jerk to your opponents."
To try for additional qualifiers at an event.
This is the flip side of our Swarthmore example. And it's putting the cart before the horse.
To confuse the host's bracketing.
This is just annoying to the host. If you change your lineups between the time the brackets are set up and the first match, you may secure a small advantage for one team, but place your other team or teams at a disadvantage.
The occasional advantages of the "Gordian Knot"
[I wrote this while I was at the county fair this week, out of wifi range, and pasting it in I realized I had forgotten the most important thing here, the real name of the system, and a picture. This passage will not make it into the book without that research. But we’re late already for this…]
The "Gordian Knot" is a nickname of a particular buzzer system originally used for the Latin competition Certamen, but used by a few quiz bowl teams (most notably Maryland) in the 1980s and 1990s. The buzzer supports four teams of four, and has sixteen buzzers labeled "A1" to "D4" connected via very long cables to a central unit with a digital readout. The chief merit of this system, besides being able to support a larger quorum at practice, is that it allows the moderator to note the first THREE people to buzz.
So what can you do with a system that accounts for the first three buzzes? Well you can enforce a rule that anyone who buzzes is checked in order, so that people don't wait for their teammates. This allows you to account for shadow effect in a particular category. If the same pattern of buzzes appears for a subject, you may be able to detect a shadow effect in practice. It alos allows you to play a variation of buzzing, where tossups are read until the third buzz occurs, at which point the third, second, and first people to buzz give their answers in that order. As long as they are honest about when they change their answer from the first or second buzz, this gives the team the ability to spot common wrong answers, and to hear more clues in a game situation.
Because online buzzer systems are designed to support multiple competitions including Certamen, they have been programmed with a setting to record buzzes after the first, which you can exploit in practice without the possibility of accidental strangulation.
So why don't I like the Gordian Knot for normal situations? First, the "Gordian Knot" nickname is well placed, if you are disposed to the mild compulsion that one team gets the green lights, and another the red, you're going to be frustrated by the Gordian Knot. Most knots stay in service for years, and as they age, they get increasingly tangled by practices. The buzzers don't light by themselves. The readout requires the moderator to interpret the buzz, relay who buzzed first to the player, who then has to remember their letter-number combination, and react to it. It is designed for exactitude, and not for efficiency. As such you can run a practice with it quite well, but using it in an actual game situation slows down the mechanics of the game. I once moderated a podcast recording room with a Gordian Knot in it, while that allowed the audience to follow who individually buzzed in on each question, it also took up time that could have been two or three tossup-bonus cycles.