| |
BGonline.org Forums
random vs. non-random rolls (long)
Posted By: Chuck Bower In Response To: random vs. non-random rolls (Chuck Bower)
Date: Saturday, 16 March 2013, at 3:03 p.m.
I should have linked Stein's post initially. Here is the link:
http://www.bgonline.org/forums/webbbs_config.pl?noframes;read=138985
Actually he started out talking about pseudo-random diceroll generation and even mentioned non-precision dice. I was most interested in his general statement which was prefaced ...any dice throw that is not intentionally rigged.... I now enumerate several cases:
0) pseudo-random vs. true random dicerolls (not covered here); 1) Intentionally throwing non-randomly (not covered here); 2) Taking advantage of dicerolls known by one or both players to be non-random (not covered here); 3) both players unknowingly using non-randomly generated dicerolls from improper equipment (emphasized by UBK), 4) one player, through his poor randomization methods, 'unknowingly' generating non-random dicerolls from otherwise proper equipment (emphasized by Colin).
3) both players unknowingly using non-randomly generated dicerolls from improper equipment: Consider two opponents, one who understands (and uses) the fact that getting hit in a certain position occurs next shake with probability P, or that opponent entering from the bar occurs with probability Q, or that opponent escaping a partial prime occurs with probability R. His opponent is one of those "I just play for fun" kind of people who doesn't like to tax his brain with details; just play the dice as they arrive and don't be concerned about what might happen in the future. Now take two cases: a) the dicerolls are truly random (so 1/36 probability of each dice combo occurring) and b) biased dice which favor some of the dice combos more heavily than others. Will the long term results of matches between these two players, but split between the two dice generation methods, lead to similar results? It is clear that when non-random dice generation is in effect that the value of the absolute skill of the calculating player will be compromised. Is that compensated for by a reduction in the performance of the non-calculating player?
4) One player, through his poor randomization methods, 'unknowingly' generating non-random dicerolls from otherwise proper equipment: I see two cases worth considering here. A) there is no subconscious effort to take advantage of the non-randomness; B) there is a subconscious effort to take advantage. A) Presumably if a player is exercising such poor technique that he is getting non-random dicerolls with good equipment, then the most obvious potential bias is that the dice thrown have a memory of the previous diceroll. (This is relevant in subcase B as well.) Now I point out the distinction between shared dice (e.g. in clocked matches) and unshared dice. Is there favoritism if a diceroll is influenced by the previous roll, but in a non-selective way? ("Selective" covered in case B). I don't know. But on initial consideration it seems as though the argument above (calculating player's absolute performance is damaged) still holds. I'm less confident in the lack of compensation by opp here, though.
4B) there is a subconscious effort to take advantage of a poor randomization: In the past (but it's been a while) I've perceived a tendency (rarely) for a player to shake vigorously after the immediately preceding roll with those dice was (or would have been) a poor one for him and to shake weakly (at best) when it was or will be a good one. Of course maybe I was just using selective, observational bias. Further, if it was true, maybe the act was intentional. But it certainly is plausible that the situation Colin describes was in effect then -- through no conscious, premeditated intent, a player was choosing to randomize (or choosing not to randomize) based upon the most recent roll of the pair of dice in use.
At least some of these questions could be answered with simulations of readily available bots (with software adjustments). And possibly some naysayers feel that, even if the effects are real, they are negligible. But as Phil and others have expressed, in the least it's a distraction no one should be subjected to.
| |
BGonline.org Forums is maintained by Stick with WebBBS 5.12.