Actually there are two elements that make this interesting. One is the fact that they caught it so early, when the supernova was still brightening. This is relatively rare. Second, this is the nearest supernova that they've found before the gamma ray burst has occurred.
Of course, as has been pointed out, it's also significantly brighter then most supernovas.
Re:This Must Be More Complex Than It Sounds . . .
on
Rock-Paper-Scissors
·
· Score: 1
My point was that a pure random (or pseudorandom?) Roshambot is never guaranteed to finish in the middle of the pack. It almost definitely will, but it's not absolutely certain. When dealing with randomness, there's always the possibility of a fluke, even if that possibility is mathematically insignificant. It still could happen, just as you could get smacked on the head by a stray meteorite. Not bloody likely, but nevertheless possible.
Obviously, to beat a pure "good ole rock" bot, it'd have to randomly pick thousands of "paper" choices, which in and of itself makes it damned unlikely, but there is no reason it can't happen. It's incredibly improbable, but not impossible.
And I could, by picking moves randomly, beat Kasparov at Chess.
Without a doubt this is a valid point, just not one that determines the virtue of the competition.
Which rewards more sophisticated and devious programs.
To a certain extent the statement is guaranteed to be true, due to "dummy" bots that will be exploitable. Several programs in the first Competition attempting to rely on Random (Optimal) too much were rarely if ever defeated but failed to score well against the weaker programs.
In contrast all of the best programs used Random (Optimal) as a fallback strategy when it was obvious that the opponent was exploiting their own moves. Their first choice, however, was to rely on pattern detection.
Slashdot Mirror
Actually there are two elements that make this interesting. One is the fact that they caught it so early, when the supernova was still brightening. This is relatively rare. Second, this is the nearest supernova that they've found before the gamma ray burst has occurred.
Of course, as has been pointed out, it's also significantly brighter then most supernovas.
Obviously, to beat a pure "good ole rock" bot, it'd have to randomly pick thousands of "paper" choices, which in and of itself makes it damned unlikely, but there is no reason it can't happen. It's incredibly improbable, but not impossible.
And I could, by picking moves randomly, beat Kasparov at Chess.
Without a doubt this is a valid point, just not one that determines the virtue of the competition.
To a certain extent the statement is guaranteed to be true, due to "dummy" bots that will be exploitable. Several programs in the first Competition attempting to rely on Random (Optimal) too much were rarely if ever defeated but failed to score well against the weaker programs.
In contrast all of the best programs used Random (Optimal) as a fallback strategy when it was obvious that the opponent was exploiting their own moves. Their first choice, however, was to rely on pattern detection.