Page 2 of 2
Posted: Thu May 17, 2007 8:43 am
by lank
the statistical independence of each die roll is why you can multiply probabilities. otherwise it's much more complicated...
Posted: Thu May 17, 2007 7:54 pm
by Frumius
Archer wrote: a number of trials that isn't infinite, the average occurrence [throwing 3 duds in a row] will probably not be 1/1000
Yeah, tell me about it. Seems like my loser duffs throw duds all the time!
Archer wrote: (though it will approach it as trials increase).
~J
I must not have thrown quite enough....
Posted: Thu May 17, 2007 9:27 pm
by AblitERateOR
it might seem like it but if you were to sit there throwing bottles and count how many throws it took to throw 3 duds in a row would probably be a lot longer than you expected.. stats doesnt lie
...presuming the virtual die throw isn't ruined by pseudo randomness
Posted: Thu May 17, 2007 10:58 pm
by Frumius
I was kidding
Posted: Fri May 18, 2007 12:30 am
by iron
One thing that adds to the complexity of duds is bounces. When a bottle bounces, it hasn't dudded yet (if the smoke trail's still happening that is) and yet it hasn't exploded either. And when it hits the ground a second time, why that's a second 1/10 chance of dudding ... with the same bottle.
I do remember though during the early days of vTFL development, we sat down and threw a huge number of bottles in TFL and then in myth 2/vTFL. The vTFL dudding ratio was significantly higher, even though the tag settings and the random number generation code etc were identical.
Also there were of course the storms of protest that 1.4 introduced more duds (along with dozens of other gameplay issues, some genuine and others not)...
Posted: Fri May 18, 2007 1:06 am
by Frumius
iron wrote:Also there were of course the storms of protest that 1.4 introduced more duds (along with dozens of other gameplay issues, some genuine and others not)...
Bah to that. You guys have helped it all along!
It's better than ever.
Of course there have been complaints, though. Some guy gets used to using a tool one way to wank up some points and it changes, he's bummed. Too bad! I appreciate what you guys have done, and I bet that even most people with gripes (lost their favorite tool/tactic/whatever) would agree that overall it's been vastly improved since the beginning. Thanks!
Posted: Fri May 18, 2007 6:45 am
by Enculator
So as a conclusion, lets say : When in doubt take 6 dorfs on killing grounds (0.000001 chances to dud when they all throw at the same time) and 9 dorfs on desert (0.000000001)