By August Bembel | Pounding The Rock (PtR), 2025-03-17 06:00:00
由生成式人工智能翻译，译文内容可能不准确或不完整，以原文为准。
一项由ESPN发布的指标，在其创建者加入NBA球队后消失，如今又回来了。
你还记得ESPN的综合性指标——真实正负值（RPM）吗？ 它从他们的网站上消失了，而且在消失之前，ESPN已经对其进行了修改。在Reddit上，你可以找到一些喜欢原始RPM、不喜欢更新版本的RPM，并想知道它为什么完全消失的球迷。
好消息是，RPM的创建者Jeremias Engelmann——他曾担任菲尼克斯太阳队和达拉斯独行侠队的分析师，现在经常出现在德国播客“Jeden Tag NBA”中——已经将该指标带回来了。它被称为xRAPM，并且可以在他的网站上公开获取。马刺球迷会发现很多值得喜欢的地方。
Jeremias在2024年12月19日播出的“Jeden Tag NBA”节目中向主持人Jonathan Walker描述了他对xRAPM的方法。对话是用德语进行的，但我会尽力翻译关键点。以下内容并非他们的确切对话——远非如此，因为他们的讨论持续了90分钟！以下与马刺相关的见解也不是他们谈话的一部分；它们是我自己基于Jeremias网站的观察，我强烈建议大家去看看。
为什么RPM现在被称为xRAPM，它又为什么会消失？
RPM最初被称为xRAPM。当Jeremias将其授权给ESPN时，他们将其重命名为RPM，据报道是因为他们不喜欢xRAPM中的“A”，因为它可能与不需要的术语相关联。ESPN仍然拥有RPM这个名称，但当Jeremias加入独行侠队时，他们将其从网站上删除。现在，Jeremias为Royce Webb的Substack自由撰稿，他以其原始名称xRAPM恢复了该指标。
xRAPM到底是什么？
该名称有三个部分：
- APM （调整后的正负值）是基础。该指标仅依赖于阵容数据，而不是个人统计数据，来计算球员对其球队的影响。虽然广为人知的on-off正负值（可在bk-ref上找到）描述了特定球员在场时球队得分和失分之间的差异（根据Jeremias的说法，这是一种“粗略”的计算），但APM更为复杂：你和谁一起打球？你和谁对抗？
- RAPM （正则化调整后的正负值）由华盛顿奇才队的分析师Joe Sill改进，通过将极端APM值推向零来缓和它们。 Jeremias表示，RAPM的“问题”在于，每个球员都有一个零的起始值，无论你是勒布朗·詹姆斯(LeBron James)还是更差的人。
- xRAPM 为每个球员增加了一个基于个人统计数据的起始值：“标准的盒式统计数据，从逐回合播放中提取的数据——比如盖帽和抢断类型——以及其他信息，如投篮防守、干扰等等。” 然后将这些信息与阵容算法混合以创建xRAPM。
为什么要添加个人球员数据？
除了个人数据的明显价值之外，Jeremias还提供了一个在NBA中常见的例子：“假设鲁迪·戈贝尔(Rudy Gobert)和贾登·麦克丹尼尔斯(Jaden McDaniels)总是同时在场。当他们被换下时，他们也会同时被换下。APM[以及RAPM]无法告诉你到底是戈贝尔还是麦克丹尼尔斯让防守变得更好。” 通过纳入个人球员数据，xRAPM解决了这个局限性。
xRAPM中“x”里面是什么？
由于显而易见的原因，Jeremias没有透露他的确切公式，但他解释说，他会选择将哪些统计数据纳入算法，而算法本身会决定每个统计数据的权重。例如，他说：“如果一个统计数据似乎贡献的价值很小，算法可能会将其权重设定为接近于零。”
马刺传奇在xRAPM中的表现如何？
简短的回答是：比你想象的还要好。 xRAPM将蒂姆·邓肯(Tim Duncan)评为2001年至2003年联盟中最好的球员，在2004年、2005年和2007年排名第二，在2000年排名第三。与此同时，大卫·罗宾逊(David Robinson)在1999年被评为头号球员（主要因为他的防守影响力），并一直保持在前10名直到2002年。
然而，从马刺的角度来看，xRAPM最美妙的事情是，它看到了我们很多人可能认为是我们观看NBA篮球时代中最被低估的球员。那就是马努·吉诺比利(Manu Ginobili)。 马努早在2004年就在xRAPM中排名第11位——并且在随后的几年中八次进入前10名。 尽管你可能不喜欢你在29年的xRAPM（因为逐回合播放数据已被跟踪）中看到的一切，但你会喜欢马努甚至进入了前10名（与其他两名前马刺球员）。
点击查看原文：The catch-all metric that loves Spurs legends
The catch-all metric that loves Spurs legends
An ESPN-published metric that disappeared after its creator joined an NBA franchise is now back.
Do you remember ESPN’s catch-all the metric, Real Plus Minus (RPM)? It vanished from their website, and before it did, ESPN had modified it. On Reddit, you can find fans who appreciated the original RPM, disliked the updated version, and wondered why it disappeared altogether.
The good news is that the creator of RPM, Jeremias Engelmann — who previously worked as an analyst for the Phoenix Suns and later the Dallas Mavericks, and now appears regularly on the German podcast “Jeden Tag NBA” — has brought the metric back. It’s called xRAPM, and it’s publicly available on his website. Spurs fans will find plenty to like about it.
Jeremias described his approach to xRAPM in a December 19, 2024 episode of “Jeden Tag NBA.” to host Jonathan Walker. The conversation is in German, but I’ll translate the crucial points as best I can. What follows isn’t their exact conversation — far from it, as their discussion lasted 90 minutes! The Spurs-related insights that follow weren’t part of their talk either; they’re my own observations based on Jeremias’ website, which I highly recommend checking out.
Why Is RPM Now Called xRAPM, and Why Did It Disappear?
RPM was originally called xRAPM. When Jeremias licensed it to ESPN, they renamed it RPM, reportedly because they disliked the “A” in xRAPM due to possible associations with unwanted terms. ESPN still owns the RPM name but removed it from their website when Jeremias joined the Mavericks. Now freelancing for Royce Webb’s Substack, Jeremias has revived the metric under its original name, xRAPM.
What Exactly Is xRAPM?
The name has three parts:
- APM (Adjusted Plus Minus) was the foundation. This metric relies solely on lineup data, not individual stats, to calculate a player’s impact on their team. While the widely known on-off plusminus (available on bk-ref) describes the difference between team points scored and team points allowed while a specific player is on the court (a “rough” calculation, according to Jeremias), APM is more complex: Who did you play with? Who did you play against?
- RAPM (Regularized Adjusted Plus Minus) was refined by Washington Wizards analyst Joe Sill to temper extreme APM values by pushing them closer to zero. The “problem” with the RAPM was, according to Jeremias, that a starting value of zero was attached to each player, regardless if you’re LeBron James or someone way wors
- xRAPM adds a starting value for each player, based on individual stats: “ “the standard box score, data that is extracted from play-by-play – like block and steal type – and other information like shot defense, deflections etc.” This information is then blended with the lineup algorithm to create xRAPM.
Why Add Individual Player Data?
Beyond the obvious value of individual numbers, Jeremias provided an example common in the NBA: “Suppose Rudy Gobert and Jaden McDaniels are always on the court together. And when they’re subbed out, they’re both subbed out. APM [as well as RAPM] is incapable of telling you whether it’s Gobert or McDaniels who makes the defense better.” By incorporating individual player data, xRAPM addresses this limitation.
What’s Inside the “x” of xRAPM?
Jeremias doesn’t reveal his exact formula for obvious reasons, but he explained that he selects which stats to include in the algorithm, while the algorithm itself determines how heavily each stat is weighted. For instance, “If a stat appears to contribute little value, the algorithm may factor it close to zero,” he said.
How Do Spurs Legends Fare in xRAPM?
The short answer: Even better than you might expect. xRAPM ranks Tim Duncan as the best player in the league from 2001 to 2003, second in 2004, 2005, and 2007, and third in 2000. Meanwhile, David Robinson is ranked the number one player in 1999 (largely for his defensive impact) and remains in the Top 10 until 2002.
However, the most beautiful thing about xRAPM, from a Spurs perspective, is how it sees whom many among us probably believe is the most underrated player in our time of watching NBA basketball. And that is Manu Ginobili. Manu came in eleventh in xRAPM as early as 2004 – and he cracked the Top 10 eight times in the years that followed. And though you’ll not like everything that you’ll see in the 29 years xRAPM (since play-by-play data has been tracked), you’ll love it that Manu has made even that Top 10 (amongst two other former Spurs).
By August Bembel, via Pounding The Rock