本文共 7467 字，大约阅读时间需要 24 分钟。
mac玩炉石传说体验怎样
by Alan Wilson
通过艾伦·威尔逊(Alan Wilson)
I’ve been playing Hearthstone since the beta. I’m a fairly casual player, but I have aspirations of eventually reaching the prestigious “Legend” ranking.
自测试版以来，我一直在玩《炉石传说》。 我是一个相当休闲的玩家，但我渴望最终达到享有声望的“传奇”排名。
Back in March I started to track my games to see if I had a chance of making it. I play most of my games on an iPad, so this was a manual process in Google Sheets.
早在三月份，我就开始跟踪自己的游戏，看我是否有机会参加比赛。 我在iPad上玩大多数游戏，因此这是Google表格中的手动操作。
I used some simple conditional formatting to make the table easier to read, but most of my analysis takes place in a free tool called . It starts out very basic—just a line chart of my rank—but even this is useful. Now I can see how well my attempts at Legend are going.
我使用了一些简单的条件格式来使表更易于阅读，但是我的大部分分析都是在名为的免费工具中进行的。 它起初非常基础-只是我的等级折线图-但这还是很有用的。 现在，我可以看到我对Legend的尝试进行得如何。
As you can see, my rank was fairly volatile, so I evolved the visualization using moving averages. The raw numbers are still there (in the dots), but the lines now represent a 25-game moving average. This made it easier to see real trends in my rank and reduce tendencies to tilt on losing streaks.
如您所见，我的排名相当不稳定，因此我使用移动平均值来发展可视化。 原始数字仍然存在(点中)，但是这些线现在代表25场移动平均线。 这样可以更轻松地查看我的排名中的真实趋势，并减少因失去条纹而倾向于的趋势。
Next I plotted the seasons on top of one another so I could compare them.
接下来，我将各个季节相互叠加，以便可以进行比较。
Now I can see how efficient my climb each season is and how it compares to the current season. But I’m still blind to how much time there is left in the season.
现在，我可以看到每个季节的攀登效率以及与当前季节相比的效率。 但是我仍然看不到赛季剩下多少时间。
Do I have a day or a week left? Am I ranked higher usual for the 5th day of the season?
我还有一天还是一周？ 在本赛季的第5天，我的排名会比往常更高吗？
To answer these questions I create a chart that tracks progress by day of the season. Here I have strip-plots for each day and a line connecting the maximum rank achieved each day. This is a better way to evaluate my chances of reaching Legend in a given season.
为了回答这些问题，我创建了一个图表，可以按季节跟踪进度。 在这里，我每天都有条形图，并且有一条线连接每天达到的最高排名。 这是评估我在给定季节中获得传奇的机会的更好方法。
But what about decks? Obviously they play a huge role. When it comes to decks, it’s all about the win-rate, so I visualize each deck as a line. It’s a bit cluttered, but I can see the win-rate, when it stabilizes, and how one deck compares to another. Tableau’s highlight feature is really helpful here.
但是甲板呢？ 显然，他们发挥了巨大作用。 当涉及到牌组时，一切都取决于获胜率，因此我将每个牌组可视化为一条线。 有点杂乱无章，但是我可以看到赢率，何时稳定下来以及一个套与另一个套的比较。 Tableau的突出显示功能在这里确实很有帮助。
Next I create a deck-specific view of this chart to see how it performs against different classes. I have to play a lot of games before trends emerge, but once they do I get a sense for how common each class is and how I’m performing against it. Depending on the meta, I also get an understanding of deck match-ups.
接下来，我将创建此图表的特定于甲板的视图，以查看其在不同类别上的性能。 在趋势出现之前，我必须玩很多游戏，但是一旦有了趋势，我就会感觉到每个班级的普遍程度以及我对此的表现如何。 根据元数据，我还了解了甲板对战。
Now, it goes without saying that my win-rate is going to suffer as I get into higher ranks, so I created a bubble chart to help me better understand the performance of my decks at different ranks and seasons.
现在，不用说，随着我进入更高的级别，我的获胜率将受到影响，因此我创建了一个气泡图，以帮助我更好地了解不同级别和季节的牌组表现。
All this got me thinking about difficulty, seasons, and the ranking system (bear with me if you’re well-versed in Hearthstone’s ranking system). First off, the ranking system isn’t linear as one might assume. The ranks (starting at 25 and going up to 1) aren’t spaced evenly because the number of stars in-between each rank changes as you climb up the ladder. All told there are 120 possible placements.
所有这些使我开始思考难度，季节和排名系统(如果您精通《炉石传说》的排名系统，请和我在一起)。 首先，排名系统并不像人们想象的那样是线性的。 等级(从25上升到1)并不是均匀分布的，因为随着您爬上梯子，每个等级之间的恒星数会发生变化。 总共有120个可能的展示位置。
But Blizzard didn’t stop there. They created additional resistance starting at rank 20, and then again at rank 5.
但是暴雪并没有就此止步。 他们创造了额外的抵抗力，从等级20开始，然后在等级5再次出现。
Lastly, Hearthstone has very short seasons—just one month. At the end of every month everyone is sent back down to the bottom of the ladder.
最后，炉石传说的季节非常短-仅一个月。 在每个月的月底，每个人都被送回阶梯的底部。
Now, this isn’t news to anyone who plays a lot of ranked games. But it brings me back to my original question: how do I measure the difficulty of a given game?
现在，这对玩很多排名游戏的人来说不是什么新闻。 但这使我回到了最初的问题：如何衡量给定游戏的难度？
I began with a few simple assumptions:
我从几个简单的假设开始：
Using these assumptions as a guide I created a model for difficulty. At first it was very simple (and inaccurate).
使用这些假设作为指导，我创建了难度模型。 最初，它非常简单(而且不准确)。
I refined it until it reflected my perception of difficulty. Note that this is not driven by any data and I doubt it’s completely accurate, but it’s better than rank alone at gauging difficulty.
我对其进行了细化，直到它反映出我对困难的感知 。 请注意，这不是由任何数据驱动的，我怀疑它是完全准确的，但是比衡量难度单独排名更好。
Now let’s revisit some of the earlier charts using this new “difficulty” metric in place of rank. This offers a different perspective — hopefully a more accurate one.
现在，让我们使用此新的“难度”指标代替排名来重新查看一些早期的图表。 这提供了不同的观点-希望是一个更准确的观点。
What about that bubble chart? I’ve put difficulty into categories, much like I did with rank and I’m already feeling better about the perspective this provides. It’s equally valuable in the early and late season.
那气泡图呢？ 我将困难分为几类，就像我对排名所做的那样，并且我已经对提供的观点感到更好了。 在早期和晚期，它同样有价值。
Next I use difficulty to create another metric. This one is pretty straightforward. I call it “quality.” If I win a game I add the difficulty of that game to quality. If I lose a game I subtract the difficulty of that game from quality.
接下来，我将使用难度创建另一个指标。 这很简单。 我称之为“质量”。 如果我赢了一场比赛，那我会把比赛的难度增加到质量上。 如果我输了一场比赛，我会从质量上减去该场比赛的难度。
If “win” then + difficulty
如果“获胜”，那么+困难
Else if “loss” then – difficulty
否则，如果“亏损” –困难
This allows me to reward winning difficult games more than easy ones and yields some very interesting results. Remember that chart showing win-rates for decks? Well, here it is again but with our new quality metric on the y-axis (color is still win-rate).
这使我比简单的游戏更能赢得困难游戏，并获得一些非常有趣的结果。 还记得那个显示牌组胜率的图表吗？ 好吧，这里再次出现，但是我们在y轴上使用了新的质量指标(颜色仍然是胜率)。
I also summarize this into a simple rank.
我还将其概括为一个简单的等级。
There are still a lot of things I want to explore further with my data-set. It would also be interesting to apply these techniques to pro players and compare them to one another. Are they rank 1 Legend because they win difficult games? Is win-rate alone the best predictor for skill? What role does the volume of games played have on rank advancement?
我还想通过数据集来探索很多东西。 将这些技术应用于职业玩家并将它们相互比较也很有趣。 因为赢得了艰难的比赛，他们是否排名1 Legend？ 仅胜率是技能的最佳预测器吗？ 游戏量在排名提升中扮演什么角色？
In case you couldn’t tell from the data I still haven’t made it Legend, but I’m honestly having as much fun analyzing my stats as I am playing the game.
万一您不能从数据中看出来，我仍然没有成为Legend，但老实说，我很高兴在分析我的数据时和玩游戏时一样有趣。
翻译自:
mac玩炉石传说体验怎样
转载地址：http://neewd.baihongyu.com/