n*******s
发帖数: 17267 | 1 是李十番棋3：0 还是古力最近对李的棋3：0？
哪个概率更大？ | a*****g
发帖数: 19398 | 2 这个无所谓的问题
【在 n*******s 的大作中提到】
: 是李十番棋3：0 还是古力最近对李的棋3：0？
: 哪个概率更大？
| n*******s
发帖数: 17267 | | n*********3
发帖数: 534 | 4 Yes, it should, I think.
There is this entropy effect. So, Li Li Gu Gu Li, is of higher probability
than Li Li Gu Gu Gu
In general, high entropy always has higher likelihood than low entropy.
so, if n games, then,
the case of Li, Gu, Li, Gu, Li, Gu .... (n games)
is more likely than
the case of Li Li Li ... (n/2) and Gu, Gu, Gu (n/2 games)
just like throwing coins n times, ...
but one may argue that they play with different level of efforts in
different type of games. That would change the probability a bit... | n*******s
发帖数: 17267 | 5 And, when you win, you tend to keep winning until that stops, lol.
This game, Gu is totally in control, black had so many bombs waiting to be
bang, bang.
小李太用强了， 不输才怪。 | x**h
发帖数: 173 | 6 很有意思的观点. 我可以理解两种分布Entropy的不同. 可是likelihood不一定有差别.
李古棋逢对手,40盘打成20比20. 看来,每人赢一盘的实际概率都是50%. 这样,总盘数
为奇数,不同的分布Likelihood会有不同;否则,likelihood应为相同. 当然,我没有考虑
连续两盘的条件概率.
这样看来,likelihood和entropy的关联可能需要一定限定.
probability
【在 n*********3 的大作中提到】
: Yes, it should, I think.
: There is this entropy effect. So, Li Li Gu Gu Li, is of higher probability
: than Li Li Gu Gu Gu
: In general, high entropy always has higher likelihood than low entropy.
: so, if n games, then,
: the case of Li, Gu, Li, Gu, Li, Gu .... (n games)
: is more likely than
: the case of Li Li Li ... (n/2) and Gu, Gu, Gu (n/2 games)
: just like throwing coins n times, ...
: but one may argue that they play with different level of efforts in
| n*********3
发帖数: 534 | 7 One throws 40 coins on the floor, high chances are 20 heads and 20 tails.
But the chance of all the heads located on side (or one area) and all the
tails on another side (or another area) is extremely low.
The chance of all 20 heads and all 20 tails are randomly mixed is very high.
Now, if one throws the coins one by one, the above logic should be the same. | w**********r
发帖数: 986 | 8 这个看要考虑多少因素了,
比如我7岁的时候和我爸跑100米，跑了两次，都输了，现在又跑两次，都赢了，假如再
跑一次，结果呢？
又比如我昨天扔了两回钢镚，都是正面，今天又扔两回，都是反面，假如再扔一次，会
怎么样呢?
【在 n*******s 的大作中提到】
: 是李十番棋3：0 还是古力最近对李的棋3：0？
: 哪个概率更大？
| n*********3
发帖数: 534 | 9 Oh, yeah, the sample size has to be large enough.
But it is quite certain that, the probability is Zero for
Li Li Li Li Li, and the Gu Gu Gu Gu Gu.
Were There rare cases in the history for such an organized events? maybe
there were a few? | r****y
发帖数: 26819 | 10 我觉得没有这个区别。所谓有区别的印象，都是有限的样本空间留下的。换句话说，来
上无数次，任何一个小片段和任何一个其它小片段出现的概率应该都是一样均匀的。
probability
【在 n*********3 的大作中提到】
: Yes, it should, I think.
: There is this entropy effect. So, Li Li Gu Gu Li, is of higher probability
: than Li Li Gu Gu Gu
: In general, high entropy always has higher likelihood than low entropy.
: so, if n games, then,
: the case of Li, Gu, Li, Gu, Li, Gu .... (n games)
: is more likely than
: the case of Li Li Li ... (n/2) and Gu, Gu, Gu (n/2 games)
: just like throwing coins n times, ...
: but one may argue that they play with different level of efforts in
| | | A******C
发帖数: 1808 | 11 那我们来玩扔硬币好了
如果正反正反正。。。。算你赢
正正正。。。反反反。。。算我赢
其他结果算平
既然你觉得你赢的可能性更大，你给我个uneven odd，咱们赌一把？
probability
【在 n*********3 的大作中提到】
: Yes, it should, I think.
: There is this entropy effect. So, Li Li Gu Gu Li, is of higher probability
: than Li Li Gu Gu Gu
: In general, high entropy always has higher likelihood than low entropy.
: so, if n games, then,
: the case of Li, Gu, Li, Gu, Li, Gu .... (n games)
: is more likely than
: the case of Li Li Li ... (n/2) and Gu, Gu, Gu (n/2 games)
: just like throwing coins n times, ...
: but one may argue that they play with different level of efforts in
| n*********3
发帖数: 534 | 12 two uniquely different outcomes will of course have the same 1/n probability
.
there are less outcomes that give more structure, such as HHHH.... TTTT...
than the outcomes (not a specific single one outcome) that give less
structure, such as HTHTHT... or THHTTHTHT... or ...
I know, the initial question is tricky to involve gambler's fallacy,...
But then, those arguments often do not consider and group the outcomes
according to their structures. | A******C
发帖数: 1808 | 13 所以你最初的意思是HHHHHTTTTT和所有其他可能比？
那你何必还扯什么entropy。。。
probability
【在 n*********3 的大作中提到】
: two uniquely different outcomes will of course have the same 1/n probability
: .
: there are less outcomes that give more structure, such as HHHH.... TTTT...
: than the outcomes (not a specific single one outcome) that give less
: structure, such as HTHTHT... or THHTTHTHT... or ...
: I know, the initial question is tricky to involve gambler's fallacy,...
: But then, those arguments often do not consider and group the outcomes
: according to their structures.
| n*********3
发帖数: 534 | 14 Entropy always consider more than one or two events/possibilities. | A******C
发帖数: 1808 | 15 1. you clearly don't understand entropy
2. if you're trying to prove multiple events having a higher probability
than a single event when each is equally likely, you don't need entropy
【在 n*********3 的大作中提到】
: Entropy always consider more than one or two events/possibilities.
| r****y
发帖数: 26819 | 16 哈哈，2用加法就可以了。。。
【在 A******C 的大作中提到】
: 1. you clearly don't understand entropy
: 2. if you're trying to prove multiple events having a higher probability
: than a single event when each is equally likely, you don't need entropy
| r****y
发帖数: 26819 | 17 完全相互独立的事件之间的entropy没啥意义。就好比俩人先扔一次硬币，再下一盘棋。
硬币的正反，和棋的输赢，都可以随便映射，你可以定义硬币的正面为赢棋，也可以
定义输棋为硬币正面。怎么定义都成立，只是一种名称而已。
所以一个十番棋，跟俩人扔三次硬币、下两盘棋、再玩五次剪刀石头布，没啥区别。
你可以等硬币下棋剪刀石头布比完了，重新改规则推翻原来的输赢结果，也是一样的。
【在 n*********3 的大作中提到】
: Entropy always consider more than one or two events/possibilities.
| A******C
发帖数: 1808 | 18 他大概想说输赢全连着比输赢全错开熵小。。。问题是这两熵一样
棋。
【在 r****y 的大作中提到】
: 完全相互独立的事件之间的entropy没啥意义。就好比俩人先扔一次硬币，再下一盘棋。
: 硬币的正反，和棋的输赢，都可以随便映射，你可以定义硬币的正面为赢棋，也可以
: 定义输棋为硬币正面。怎么定义都成立，只是一种名称而已。
: 所以一个十番棋，跟俩人扔三次硬币、下两盘棋、再玩五次剪刀石头布，没啥区别。
: 你可以等硬币下棋剪刀石头布比完了，重新改规则推翻原来的输赢结果，也是一样的。
| r****y
发帖数: 26819 | 19 是啊，关键是每次输赢都是独立事件。可以比赛完全不同的东西。所以第一个赢与第二
个赢之间的距离，和第一个赢与第二个输之间的距离是完全一样的。
【在 A******C 的大作中提到】
: 他大概想说输赢全连着比输赢全错开熵小。。。问题是这两熵一样
:
: 棋。
| M****e
发帖数: 104 | 20 连续两盘输赢有相关，不是独立的
别.
【在 x**h 的大作中提到】
: 很有意思的观点. 我可以理解两种分布Entropy的不同. 可是likelihood不一定有差别.
: 李古棋逢对手,40盘打成20比20. 看来,每人赢一盘的实际概率都是50%. 这样,总盘数
: 为奇数,不同的分布Likelihood会有不同;否则,likelihood应为相同. 当然,我没有考虑
: 连续两盘的条件概率.
: 这样看来,likelihood和entropy的关联可能需要一定限定.
:
: probability
| h*****a
发帖数: 1718 | 21 Why?
【在 M****e 的大作中提到】
: 连续两盘输赢有相关，不是独立的
:
: 别.
| n*******s
发帖数: 17267 | 22 心态跟前一盘不一样？
但对结果的影响还是不可知，Lol。
【在 h*****a 的大作中提到】
: Why?
|
|
