求:暴雪解释魔兽世界掉落机制的那篇蓝贴原文和翻译版

就黑手这个问题，我们已经争论了很久～

有人说，黑手是他们的怨念……

有人说，黑手是阻碍团队前进的最大障碍……

很多负责开尸体的人都是背负着黑手称号的人。

于是，人们开始研究，掉落究竟是在怪物诞生的时候生成的，还是在他死亡的时候生成的，还是在拾取的那一瞬间生成的，各有一方说辞～

今天，暴雪终于把这个怨念终结了。

魔兽世界美服论坛版主-Drysc，让我们记住这个名字。

以下是他的原话：

所有掉落都是在怪物生成的时候就已经决定了。

当你进入一个副本的时候，或者一个怪物在野外刷新，他的掉落就已经注定了。

所以，黑手是不存在的。

进入某个RAID副本，是有进度的，那么这个进度是由组建团队的人来决定呢，还是第一个进入副本的人来决定呢？看起来后者的可能性更大一些 ：）

无论如何，都不要怪那些去开尸体的人们了，想要继续担当黑手的朋友们，请勇敢地第一个迈入副本吧！

呵呵，不继续开玩笑了，其实这次我想告诉大家的是，暴雪这次带来了更为细致的解释，请看;

Q: What are the variables in the random generator''s kernel, is it based on some attribute of the players in the raid, or the raid leader

A: Heh, no

I believe it uses a time stamp to seed the generator So, there''s nothing you can influence within the game that would alter or coerce the generator

Drysc说事实上服务器内部会按照某种时间概率给掉落生成器发送命令，游戏中的行为并不会对这段命令产生影响。

No, the time it uses to seed would have no real implication into the final outcome of what loot its given On a long enough generation table, you may see some type of pattern based purely on a random generators inability to truly be random But we''re talking about astronomical numbers, something ridiculous like hundreds of billions of generations Nothing you would or could see a pattern in during your time playing the game

What you''re seeing is pure and basic laws of probability Have a few thousand people roll a pair of dice and you''re bound to see hundreds of people able to roll snake-eyes 10 times in a row (not that dice are truly random either)

于是有人问那命令发送是按照时间段，那岂不是杀死某时间段刷出来的怪物或者在某个特定时间进入副本。。。 橙色武器唾手可得？

版主继续解释：发送命令的时间和你们取得的掉落没有直接联系，服务器随机把可能掉落的物品随机排列成为一个内容非常巨大的掉落表，可以解释为有数千亿的天文级的数据，而这个是根据概率论来计算的，因为这个列表过于庞大，所以其实你没有任何办法找到其中的规律。

这可是一种蛋白质啊，蛋白质具有四级结构，想写出酶的结构简式应该不是人类手工完成的事情。

还是给你些相关材料吧：

淀粉酶结构预测

min=17 | max=33 | html | amy | plain_text | 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

TMpred output for amy [ISREC-Server] Date: Sun May 21 1:44:34 Europe/Zurich 2006

Sequence: MHAWRR, length: 475

Prediction parameters: TM-helix length between 17 and 33

1) Possible transmembrane helices The sequence positions in brackets denominate the core region

Only scores above 500 are considered significant

Inside to outside helices : 4 found

from to score center

16 ( 16) 34 ( 34) 1388 25

155 ( 155) 178 ( 173) 387 165

242 ( 245) 266 ( 264) 187 255

437 ( 437) 459 ( 459) 549 447

Outside to inside helices : 2 found

from to score center

16 ( 16) 35 ( 33) 1017 25

438 ( 444) 466 ( 463) 390 455

2) Table of correspondences Here is shown, which of the inside->outside helices correspond to which of the outside->inside helices

Helices shown in brackets are considered insignificant

A "+"-symbol indicates a preference of this orientation

A "++"-symbol indicates a strong preference of this orientation

inside->outside | outside->inside

16- 34 (19) 1388 ++ | 16- 35 (20) 1017

( 155- 178 (24) 387 ++ ) |

( 242- 266 (25) 187 ++ ) |

437- 459 (23) 549 + |( 438- 466 (29) 390 )

3) Suggested models for transmembrane topologyThese suggestions are purely speculative and should be used with extreme caution since they are based on the assumption that all transmembrane helices have been found

In most cases, the Correspondence Table shown above or the prediction plot that is also created should be used for the topology assignment of unknown proteins

2 possible models considered, only significant TM-segments used

the models differ in the number of TM-helices !

-----> STRONGLY prefered model: N-terminus outside

2 strong transmembrane helices, total score : 1566

# from to length score orientation

1 16 35 (20) 1017 o-i

2 437 459 (23) 549 i-o

------> alternative model

1 strong transmembrane helices, total score : 1388

# from to length score orientation

1 16 34 (19) 1388 i-o

由美国加州大学开发的基于PSI-BLAST搜索程序和神经网络算法为用户提供蛋白质空间结构预测的服务器。预测内容包括：二级结构特性预测(包括二级结构分类预测、相对亲水性预测、残基间接触数预测）、肽链拓扑学特性预测（包括残基接触图预测、二级结构单元接触图预测）。