Is it a solenoid like what I saw the Mythbusters do in their Indiana Jones temple escape? How come we never see ties? Ties must happen. How’d it work with Watson? Surely Watson couldn’t press the buzzer. And when someone presses the buzzer, what happens?
A circuit that accepts only the first buzzer to be pressed, and locks out the others, is quite straightforward. I literally wired one up when I was in grade school (following a schematic someone else designed).
Yeahbut. . . ties? No ties ever? And how’d they do Watson?
It’s unlikely that two people would press at exactly the same time, which it would have to be considering how fast the circuit must respond.
You are also locked out until the host reads the question and it is shown, which makes ties even less likely.
I sucked at pressing the buzzer when I was on. Not just during the show, but in practice too.
If the buzzers continuously sample, the probability of a tie is zero. And if you do get one, randomly breaking it is pretty easy.
Watson had some “reaction speed” programmed in so that it wouldn’t just always be first.
I think the effect of ties might depend on the nature of the circuit in question. A simple circuit based on logic gates might do something like lock out all the buzzers if two are pressed at exactly the same time.
A typical (simple) buzzer circuit would have a built-in bias, which would mean that if two buttons were pressed at the same time, the one that is favored would always win. If one was worried about this case (which would mean caring about button presses nanoseconds apart), then it is possible to design the circuit so that ties are decided randomly.
IIRC, the circuit I wired up would always result in one light lighting up and the other one not. It might have slightly favored one over the other, such that A could be a few nanoseconds later but still win, but then you’re getting into things like the lengths of the wires (or, more realistically, physical differences in the construction of the buzzers). There would never be a case where one was measurably first but still lost. I’m not sure it’s even meaningful to speak of the two being pressed “at the same time”.
The guy who wrote that book about Jeopardy (Lost In Trebekistan IIRC) said he spent a LOT of time practicing but buzzer.
When you’re getting down to nanoseconds like that, you could also have relativistic problems with “simultaneity”. Different observers could disagree on who buzzed first, and whether the buzzer picked the right one.
Realistically, the errors created by the pushbutton itself would make determining who pushed first impossible if the signals came in with even millisecond separation. This is starting to get to the point where it actually could make a difference. I guess the contestants just have to trust the system.
If it were possible for signals to come in truly simultaneously, then I think we all agree that a fair way of handling it would be to determine randomly, yes? And when the timing of the signals is so close that it can’t be determined, this is effectively what any real system will do. So in practice, there’s no problem.
That’s not true in practice.
This is all about Metrology. In general, you need a more sophisticated and expensive measurement device to determine the precision of the device you are building. That device will show biases in your circuit. For example, the switching speed of the transistor in the input of one channel might be slightly faster than the other channels, leading to a bias. Your very expensive oscilloscope might be able to show that difference, even though your buzzer circuit appears to be fair.
And any bias, of course, can be adjusted out up to a point. After that you’re left with random uncertainty, which is determined by the thermal noise in a circuit. This means that, no matter how carefully you design your circuit, it’s impossible to to determine the winner for two signals that come in less than X nanoseconds apart.
Came out after my time, and I didn’t know it would be a problem. I did practice with the computer game, which was absurdly easy. But it did help in practice for computing Final Jeopardy bets.
I recommend the book. An easy read, funny, and informative.
Back in the 80s, I was told that a tie would result in both buzzers being locked out for some tiny fraction of a second (I don’t remember what). Of course, it’s possible that electronics have improved since then.
Voyager, I’m with you. I lost because I couldn’t master the buzzer, not because I didn’t know the answers. I mean questions. Responses.
Oh, and Watson was programmed to assess the probability that its answer was correct. The higher the possibility, the faster it buzzed in.
I’ve always thought the buzzer took away from game play. The game should be a test of knowledge, not reflexes. I’d prefer that the players take turns picking the answers, on you turn you must provide a question and if incorrect, the next player in turn would have to respond and if incorrect, the third player must respond. As is, the buzzer reflex seems like it’s 95% of the game.
But why? I’d think the likelihood of Watson knowing an answer would be high enough that the best strategy would be hitting the buzzer immediately on every answer.