<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by David Joerg on Medium]]></title>
        <description><![CDATA[Stories by David Joerg on Medium]]></description>
        <link>https://medium.com/@dsjoerg?source=rss-60b4d12417b6------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*DwInT5btlyf7gYJm.jpeg</url>
            <title>Stories by David Joerg on Medium</title>
            <link>https://medium.com/@dsjoerg?source=rss-60b4d12417b6------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sun, 22 Nov 2026 07:27:10 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@dsjoerg/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Ninthpawn — Opening Thoughts]]></title>
            <link>https://medium.com/@dsjoerg/ninthpawn-opening-thoughts-4b666526439d?source=rss-60b4d12417b6------2</link>
            <guid isPermaLink="false">https://medium.com/p/4b666526439d</guid>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[chess]]></category>
            <dc:creator><![CDATA[David Joerg]]></dc:creator>
            <pubDate>Fri, 17 Mar 2017 15:15:44 GMT</pubDate>
            <atom:updated>2017-03-17T15:16:35.545Z</atom:updated>
            <content:encoded><![CDATA[<p>Chess education is ready for a computer-driven revolution.</p><p>For several years, computers have been the undisputed champions of <strong>playing</strong> chess. However, for all their skill in playing chess, computers have very limited ability to <strong>analyze</strong> chess, and no ability whatsoever to <strong>teach</strong> or <strong>explain</strong>.</p><p>Now that we are living in the future, with Artificial Intelligence and Machine Learning around every corner, and surrounded by Big Data, we should have an “autocoach” that can participate in our chess lives.</p><p>Your autocoach should be able to analyze your games and rate the <strong>quality</strong> of your play. Looking at a set of your games, it should be able to evaluate your skills and suggest a training plan.</p><p>Your autocoach should be able to look at a position and tell you not only what is the best move, but <strong>why</strong> it’s the best move.</p><p>Today’s computers have no problem recognizing blunders. Thanks, computer, but why so negative? Your autocoach should also be able to point out your <strong>best</strong> moves!</p><p>That reminds me. Yesterday I was reviewing <a href="http://en.chessbase.com/post/the-game-that-shook-the-world">The Game of the Century</a>, where Donald Byrne, one of the best players in the United States, faced a 13-year-old Brooklyn schoolboy.</p><p>Here’s how the game looks to a chess engine:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/764/1*tGPWFHaSBfUT12hFeDgO5A.png" /><figcaption>Lichess analysis of the Game of the Century. X-axis is the move #, y-axis is the advantage for white or black, as judged by the Stockfish engine. <a href="https://en.lichess.org/uWrwLdMD">https://en.lichess.org/uWrwLdMD</a></figcaption></figure><p>Donald Byrne’s missteps are called “mistakes” and “blunders”. The engine has no idea whether they were truly blunders in the common sense of the term, or instead subtle but none the less damaging. Far worse, Bobby Fischer’s brilliant moves are entirely unacknowledged. They are simply rated zero; not a mistake. Merely playing the correct move.</p><p>Can we teach the computer that some good moves really are better than others? Can a machine learn that not all 100-centipawn mistakes do equal damage to the player’s winning chances? Yes!</p><p>Can the computer distinguish between a tactical error and a strategic one? I have already built the code!</p><p>Can big data guess a player’s rating from the moves of a single game? Yes, <a href="https://www.kaggle.com/c/finding-elo">I already did that too</a>!</p><p>Automated coaching will be great, but it will not be a magic bullet. There is no substitute for the hard work of training. All we can hope for from any coach, automated or otherwise, is guidance, feedback and motivation.</p><p>I understand fully and deeply how difficult it will be to achieve any of the things described here — future essays will address how.</p><p>Today I released a simple tool, a <a href="http://www.ninthpawn.com/position-eval/">Position Evaluator</a>. It simply shows the details of the single-position evaluation function of Stockfish. It is the first step on a 30-step path that leads to a general explanation capability for chess positions and mistakes.</p><p>When I first encountered Stockfish’s evaluation function, I was pleasantly surprised to see how many chess concepts were embedded within it. Perhaps you’ll have fun playing with it too.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=4b666526439d" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>