吃饭睡觉做卷子 吃饭睡觉做卷子
关注数: 49 粉丝数: 38 发帖数: 4,267 关注贴吧数: 92
《理论水》一点点关于先手价值的有趣的研究 最近闲来无事(打dota的机油都回国了),点开awbw(其实是想找人打dota),然后发现了自己drive上以前的草稿,顺着其思路做了点东西。聊胜于无发出来给大家看看。 下面切入正题:aw这样一个回合制游戏,同样的兵种在没有任何攻防修正的前提下进行战斗,先手方的优势有多大?(排除运气因素,在本帖中luck暂时被锁定为0) 首先我们假设双方是aw最基础的兵种 infantry,互相攻击力为55%,那么让我们来模拟一下战斗结果,以下数字为血量%我们可以看到先手方很轻易的打倒了对手,并且留下了45%生命值。 那么,我们逐渐减少先手方的血量%,使双方尽可能的接近平手我们可以看到在先手一方血量到达90%的时候,出现了临界点,先手不再打得过后手了。 不过由于aw伤害计算公式为了方便玩家计算,将单位显示的数字血量向上取整,因此不能求得精确的临界点 注: 伤害公式: 实际的aw伤害公式 ROUNDUP(攻击方血量/10)*10*0.55 实际的aw血量精确到整数 理想aw伤害公式 攻击方血量*0.55 下面我们采用不取整的伤害公式进行再次进行模拟(血量在计算伤害时不再取整,血量本身不再取整)我们可以看到临界点不再出现在90血量附近,下面我们继续细化血量逼近临界点这次临界点出现在87-88之间,我们继续无限细分得出当先手方血量在87.62xxxxxxx的时候,可以与后手方战平 另外此临界点受基础攻击力等因素影响,另附基础攻击力在60%时候的数据可见临界点在85%血量左右 实用价值: 这个故事告诉我们,不要没事拿7血的坦克去撞人家10血的坦克,根本打不过的。 思考题: 1:我们能否用数学的方法求得该临界值 提示:所用知识不超过初中数学范围 2:我们在前面看到先手满血inf互殴会存活下来一个45(50.5)不取整的inf 我们又知道87.62 HP inf 先手战斗力=一个满血inf 那么50.5血 inf 是否= 满血inf - 87.62hp inf呢? 附录:先手的价值.xls 已经上传到群文件,有兴趣的可以来看
【理论&技巧】贴身占位 不说别的先上图现在该蓝色人族行动.试问如何压制红色 注:本图为虚拟情景,非实战解1: 两tank上去打等离子,工程师火箭跟上神族反击: 由于奔袭攻击后移动不占打击面的特性,在等离子先手7血坦克之后通过两个奔袭上去可以直接杀掉坦克,而且自身反伤很低,然后最后一个奔袭杀工程师,人族陷入被动解2: 两个坦克贴身卡在等离子前面不攻击,下轮如果神族想用奔袭舰攻击坦克则必须要把等离子移走,于是浪费了神族等离子/奔袭舰的火力 而无论是两个等离子还是等离子+两个奔袭舰都是无法合死满血坦克的(注:等离子+三奔袭可打死坦克),于是也就保住了后面的工程师。于是可以下轮使用emp利用剩下的坦克卡位做肉盾并且杀等离子,而神族如果要杀坦克则前线基地不能造兵(见杀工程师那张图),神族不好走 总结: 贴身占位主要运用于对抗攻击后移动的兵种,(最典型的就是奔袭),通过与普通单位(等离子)的贴身来浪费对面输出,从而打到保护自己前排肉盾的目的。 另: 本帖改编自aw错位站位法理论,作者jenni 原帖地址: http://tieba.baidu.com/p/884298870 另附几个数据: 解1中杀死坦克的概率: 0.729212 解2中两个等离子杀死坦克的概率: 0.1768 解2中等离子+两个奔袭杀死坦克的概率: 0.057 解2中等离子+三个奔袭杀死坦克的概率: 0.748739 注:以上概率中的基础伤害概率均取自伤害计算器,另计算概率时不考虑伤害更高的暴击以及伤害更低的弱击情形(一般小于1%) uniwar伤害计算器: http://tieba.baidu.com/mo/q/checkurl?url=http%3A%2F%2Ftool.slpo.net%2Funiwarcalc%2F%3FAU%3D8%26DU%3D23%26AT%3D0%26DT%3D0%26B%3D0%26U%3D0%26AUH%3D11%26DUH%3D9&urlrefer=0f5d092159f7242a43a27acd33d17fe7
冒出来转个blog Newest Post (7/16/12) Game Review - japanj vs. Hellraider This post will be something I have wanted to do for a long time, but never found the time or motivation for. It is a game review, where I'll go through a AWBW game, analyze the most important turns, explain the thought processes behind them and point out mistakes. The game I chose to review is one I played myself in the sixth season of the colosseum, a cross-division match against japanj. Not only do I think that it is one of the closer and more exciting games that have happened in the last few months, it also makes it easier for me to analyze as I have already studied it carefully during the course of the game. For the start, I will talk a bit about the CO choice. I don't actually remember the banlist now, but I recall that the only COs I even considered were Drake and Javier. While mass damage is incredibly powerful in High Funds settings, I decided to pick Javier because I was afraid of losing contested properties. The irony. Day 4Nothing of interest happens until day 4, where I as CI have the option to build a tank with 8k funds. Generally, it is very important to capture all your properties as fast as possible, especially in High Funds settings. That makes a tank in the east completely unviable as there are several neutral properties. A tank in the west was a possibility, but would still slow down the capture of one city and the tower. Having the first tank can sometimes be a minor advantage depending on several factors. It will never win you the game and also nearly never slow down the opponent's capture phase, but it allows you to choose the initial area of conflict. On maps where your tanks can switch fronts quickly to threaten the opponent, you can sometimes capitalize on even small positional errors. However, I didn't deem the first tank as important enough to build it on this turn. What seemed like a rather unimportant decision to be at that time turned out rather badly for me, leading the game towards the course it took. Thinking only about the slowdown of my own capture, I failed to realize that the response the tank forced of him would have hampered his captures as well, most likely more than my own. This is what I'd call a clear mistake. When talking about mistakes, I usually differentiate between mistakes in execution and mistakes in judgement. Mistakes in execution, which you could also call tactical mistakes, are moves like miscalculating your meat shield or having your infantry harrased by a copter. Mistakes in judgements, often related to decisions about which units to build or whether to attack, are often times a lot harder to spot, but can have far bigger consequences. In this case, I made a mistake in judgement, a strategical error that bore consequences until the end of the game.
1 下一页