<?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 Imeth De Silva on Medium]]></title>
        <description><![CDATA[Stories by Imeth De Silva on Medium]]></description>
        <link>https://medium.com/@imethdesilva?source=rss-1a39e89b963e------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*JUKh6wZsnGLqvjVbYf6-4Q.jpeg</url>
            <title>Stories by Imeth De Silva on Medium</title>
            <link>https://medium.com/@imethdesilva?source=rss-1a39e89b963e------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sun, 22 Nov 2026 05:32:11 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@imethdesilva/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[Technical Documentation — NSS WAR Calculator]]></title>
            <link>https://medium.com/@imethdesilva/technical-documentation-nss-war-calculator-4c7641c9875d?source=rss-1a39e89b963e------2</link>
            <guid isPermaLink="false">https://medium.com/p/4c7641c9875d</guid>
            <category><![CDATA[calculations]]></category>
            <category><![CDATA[coding]]></category>
            <category><![CDATA[math]]></category>
            <dc:creator><![CDATA[Imeth De Silva]]></dc:creator>
            <pubDate>Sat, 25 Apr 2026 21:26:39 GMT</pubDate>
            <atom:updated>2026-09-22T17:22:41.143Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1000/1*pgk1z7nhIhyh6vdSphhIxQ.jpeg" /></figure><h3>Introduction &amp; The WAR Concept</h3><p>Selecting a national team to represent a country on the international stage is a responsibility that requires absolute precision and transparency. In the context of Sri Lankan Scrabble, this is achieved through the Weighted Average Rating (WAR).</p><p>The WAR system is designed to reward not just high skill, but consistent and recent performance. Rather than taking a snapshot of a player’s current rating, WAR evaluates performance over a 20-month window, divided into five 4-month “Quadrimesters,” where more recent games carry higher mathematical weight.</p><p>For a deep dive into the logic behind this system and the official selection rules, please refer to the detailed article:<br><a href="https://medium.com/@imethdesilva/weighted-ratings-and-the-national-scrabble-selections-process-567231d9c486">https://medium.com/@imethdesilva/weighted-ratings-and-the-national-scrabble-selections-process-567231d9c486</a></p><h3>The Problem Statement</h3><p>Before the development of this portal, the national selections process faced two primary “pain points”:</p><p><strong>For the Committee:</strong> Calculating WAR was a gruelling manual process. Administrators had to handle multiple spreadsheets, manually categorize tournaments into quadrimesters, and calculate weights by hand. This was prone to human error and delayed the announcement of official squads.</p><p><strong>For the Players:</strong> There was no “live” portal. Players often had to wait weeks for updated standings, making it difficult to understand exactly what they needed to achieve in upcoming tournaments to qualify for the World Youth Scrabble Championship (WYSC) or the World Scrabble Championship (WSC).</p><h3>The Solution: Automated Selections Engine</h3><blockquote>The National Scrabble Selections WAR Calculator, or NSS WAR Calculator, is a professional-grade administrative dashboard that streamlines the entire workflow from ingesting raw tournament results all the way to an audit-ready, explainable selection report.</blockquote><blockquote>The tool has since evolved from a one-off calculator you fed files into each time, into a persistent, GitHub-backed Tournament Archive that the committee builds up over time, with password-protected admin access, a per-admin audit trail, and full transparency built into every exported report.</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*9tkS-gecpWUV7kRsemfyuw.png" /><figcaption>National Scrabble Selections — WAR Calculator</figcaption></figure><h3>How it Works:</h3><p><strong>Unlock:</strong> The admin opens the Tournament Archive tab, enters the shared admin password, and their name or initials, which is now recorded against every change they make, so a shared password doesn’t mean anonymous edits.</p><p><strong>Archive:</strong> Tournament result files (exported as .txt from tournament software like TSH) live permanently in a year-by-year archive, synced to GitHub. The admin can browse, edit, rename, or delete any file, or add a brand-new one via the “+ Add Tournament File” button which parses the file, shows the detected tournament name/date/player count for review, and only pushes it to GitHub once explicitly confirmed.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*5CJJ8dxTdRr75UAxZH-JUw.png" /><figcaption>Tournaments Archive</figcaption></figure><p><strong>Configure &amp; Run:</strong> Rather than configuring and running one classification at a time, the admin enters both the WSC and WYSC international event dates in a single form and runs both calculations together in one click. Each classification correctly applies its own cut-off offset: six months before the event for WSC, three months for WYSC and both results are cached, so switching between them afterwards is instant.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ntHdGLSAV1pJavTGaJwVhA.png" /><figcaption>Running WARs</figcaption></figure><p><strong>Automation: </strong>The app parses names, dates, ratings, and game counts; determines eligibility against the full official criteria; calculates WAR; and generates a live, ranked leaderboard — complete with a Data Quality summary flagging any unparseable lines up front, and a [PROVISIONAL] badge on any player whose most recent rating hasn’t yet been fully established.</p><p><strong>Explain:</strong> Every result can be exported as a fully self-contained, audit-ready PDF not just a list of names and numbers, but a report that opens with the exact calculation details and quadrimester schedule used, and shows the complete arithmetic behind every player’s WAR: Weight × Rating = Weighted Value for each tournament, the totals, and the final division spelled out.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/509/1*PxEIWwxV_XGp57iONXmNZw.png" /><figcaption>Individual Player Audit</figcaption></figure><p><strong>Integrate:</strong> Standardized CSV exports remain available and can still be pushed live to the Federation’s website with one click, so players can check their standings in real time. The full archive can also be downloaded as a single zip for backup or offline reference.</p><h3>Feature Breakdown &amp; Mathematical Logic</h3><p>The app dynamically calculates the 20-month selection window based on a cut-off “Buffer” period (six months for WSC, three months for WYSC) measured back from the international event date. It then works out which five 4-month Quadrimesters fall inside that window.</p><p>Unlike a simple rolling four-month lookback, quadrimesters are anchored to fixed calendar seasons: January to April, May to August, and September to December, matching the official criteria exactly. The most recent quadrimester (Quadrimester 5) is whichever fixed season the cut-off date falls into, with one important exception straight from the selection criteria: if that season has no tournaments yet, or exactly one “stray” tournament in its first month, it gets folded into the previous quadrimester instead of standing alone with too little data.</p><p>Each Quadrimester is assigned a weight starting from 1.00 and increasing by 0.25 per period:</p><p>Weight = 1.00 + (i − 1) × 0.25</p><pre># Logic for weight assignment<br>for i in range(5, 0, -1):<br>    q_start, q_end = get_season_bounds(curr_end)   # fixed Jan-Apr / May-Aug / Sep-Dec<br>    weight = 1.00 + (i - 1) * 0.25<br>    quads.append({&quot;quad&quot;: i, &quot;start&quot;: q_start, &quot;end&quot;: q_end, &quot;weight&quot;: weight})<br>    curr_end = get_prev_season_end(q_start)</pre><h3>The WAR Calculation</h3><p>This is the core brain of the app. It calculates the weighted sum of every qualifying tournament performance and divides it by the sum of the weights applied.</p><p><strong>The Math:</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/356/1*HYutPyfSwDi2jNa3NtZ-kA.png" /></figure><pre># Calculating the Weighted Average Rating<br>tw = sum(h[&#39;Weight&#39;] for h in data[&#39;history&#39;])<br>twr = sum(h[&#39;WeightedVal&#39;] for h in data[&#39;history&#39;])<br>war_precise = twr / tw if tw &gt; 0 else 0<br>war = round_half_up(war_precise)   </pre><h3>Eligibility</h3><p>The system automatically audits every player against the full official criteria; not just games and tournaments played:</p><p><em>WSC: 80 rated games, 5 tournaments (2 of them in the two most recent quadrimesters), at least 3 distinct quadrimesters, 1 “major” (18-round) tournament, WAR ≥ 800.</em></p><p><em>WYSC: 50 rated games, 3 tournaments (1 in the two most recent quadrimesters), at least 3 distinct quadrimesters, 1 “major” (18-round) tournament, WAR ≥ 800.</em></p><p><strong>Inactivity:</strong> a player who hasn’t played a rated tournament in over a year is automatically marked ineligible; if they resume, they’re re-considered only after 50 rated games played since resumption, exactly as specified in the criteria.</p><p><strong>Tie-Break:</strong> if two players land on the exact same WAR, the official tie-break chain applies automatically; higher current rating first, then WAR rounded to two decimal places.</p><h3>System Validation (Test Cases)</h3><p>The following test cases were conducted to ensure the system meets required standards.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*n1NaTYFkSI_pV1cuhZaqQw.png" /><figcaption>Configuration Test Cases</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*-mr9bxc6hcArQKTzG3RlxQ.png" /><figcaption>Individual Results Test Cases</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*tQeo5lFwJUrCkHYuHl3HYg.png" /><figcaption>Policy and leader board Test Cases</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*TM89vFw7jd0LzdLFhX7mVw.png" /><figcaption>Security and Reports related Test Cases</figcaption></figure><h3>Conclusion</h3><p>The NSS WAR Calculator modernizes complex calculations and replaces a tedious, error-prone manual process with a system that is fast, accurate, and completely transparent from a persistent, admin-managed tournament archive through to a fully explainable, audit-ready result for every player.</p><p>This tool ensures that every player is evaluated fairly against the official 2024 criteria, providing the committee with reliable, defensible data, and giving players a clear, real-time view of their national standing.</p><h3>Important Links</h3><ul><li><strong>View the Code:</strong> <a href="https://github.com/imethdesilva/WAR_Calculator">GitHub Repository — WAR_Calculator</a></li><li><strong>Process Overview:</strong> <a href="https://medium.com/@imethdesilva/weighted-ratings-and-the-national-scrabble-selections-process-567231d9c486">Read about the Selection Process on Medium</a></li><li><strong>SLSF Website:</strong> <a href="https://www.scrabblefederation.lk/">https://www.scrabblefederation.lk/</a></li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=4c7641c9875d" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Weighted Ratings and the National Scrabble Selections Process]]></title>
            <link>https://medium.com/@imethdesilva/weighted-ratings-and-the-national-scrabble-selections-process-567231d9c486?source=rss-1a39e89b963e------2</link>
            <guid isPermaLink="false">https://medium.com/p/567231d9c486</guid>
            <category><![CDATA[formula]]></category>
            <category><![CDATA[calculations]]></category>
            <category><![CDATA[math]]></category>
            <category><![CDATA[scrabble]]></category>
            <category><![CDATA[ratings]]></category>
            <dc:creator><![CDATA[Imeth De Silva]]></dc:creator>
            <pubDate>Wed, 07 May 2025 19:36:20 GMT</pubDate>
            <atom:updated>2025-07-21T14:51:39.761Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*vIwDNl9ehgpBisbiQ8Zekg.png" /></figure><p><strong>This article will cover everything from the equations and mathematics behind weighted ratings to how players are selected to represent Sri Lanka in international competitions such as the World Scrabble Championship and the World Youth Scrabble Championship.</strong><br>It is written solely with the aim of providing Scrabble players with concise and clear information on how players are transparently selected for national teams based on their tournament performances.</p><iframe src="https://drive.google.com/viewerng/viewer?url=https%3A//www.scrabblefederation.lk/static/docs/Selections%2520Criteria%25202024.pdf&amp;embedded=true" width="600" height="780" frameborder="0" scrolling="no"><a href="https://medium.com/media/8ce3338a0f7b9fe8471018db913b53b4/href">https://medium.com/media/8ce3338a0f7b9fe8471018db913b53b4/href</a></iframe><p>The full selection criteria document shown above can be found on the Scrabble Federation of Sri Lanka website.</p><h3>What Are Weighted Ratings?</h3><p>The first thing to understand is the meaning of <em>weighted ratings</em>. A weighted rating is essentially an average of your rating over a period of time, with more recent performances given higher importance. Typically, your weighted rating will be lower than your current rating, but there can be exceptions. A player’s weighted rating is the primary metric used to determine selection for international tournaments.</p><p>In this article, I will break down the requirements for players to get selected for international competitions such as the World Scrabble Championship (WSC), the World Youth Scrabble Championship (WYSC), and the recently introduced Asian Youth Scrabble Championship (AYSC). The requirements for WYSC and AYSC are identical, while WSC has a few additional criteria.</p><h4>Requirements for the WSC and the WYSC</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ZVlAoEypMHmiNVjXy4nzYA.jpeg" /><figcaption>Requirements for the World Scrabble Championship</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Otf9XKvvx2Mqopd-XyzRcA.jpeg" /><figcaption>Requirements for the World Youth Scrabble Championship</figcaption></figure><p>As shown above, the two competitions differ in the cut-off date of announcement, the number of required games and tournaments, and the rating threshold.</p><h3>What Is a Quadrimester?</h3><p>A <em>quadrimester</em>, in simple terms, is a period of four months. For selection purposes, a five-quadrimester period is considered — i.e., the 20 months leading up to the cut-off date. Based on your tournament participation and rating changes during this 20-month period, your <strong>Weighted Average Rating (WAR)</strong> is calculated.</p><p>The most recent quadrimester is given a weight of 2, while each previous quadrimester’s weight decreases by 0.25, as shown below:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/635/1*JAIQAvhnO6S_JwVdD_liUQ.jpeg" /></figure><p>This system ensures that more recent performances carry more influence. Ratings from each tournament are multiplied by their respective weight, and the total is divided by the sum of weights to calculate the WAR.</p><p>For any given year, the quadrimesters are fixed as:</p><ul><li><strong>1st January — 30th April</strong> — QUAD 1</li><li><strong>1st May — 31st August</strong> — QUAD 2</li><li><strong>1st September — 31st December</strong> — QUAD 3</li></ul><p>Let’s go through an example to understand this better.</p><h3>Example Scenario</h3><p>Let’s say the 2026 World Youth Scrabble Championship is scheduled for <strong>1st September 2026</strong>. The cut-off date to announce selected players would be three months prior: <strong>1st June 2026</strong>. Therefore, the five quadrimesters considered would be:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*kDoSBG4jzGaRyhJ_yIs3_g.png" /></figure><p>Note: Although QUAD 5 only includes one month, it is still considered the most recent quadrimester because quadrimesters are fixed and any activity during this period counts.</p><h3>WAR Calculation Example</h3><p>Let’s say Gregory played in 5 rated tournaments between 1st January 2025 and 1st June 2026. His rating changes after each tournament are noted.</p><p>The WAR is calculated by applying the appropriate weight to each rating and dividing the total by the sum of the weights.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*GUAjQS3ke3HY6OFuHnb7bQ.png" /></figure><p>The following is how Gregory’s Weighted Average Rating (WAR) will be calculated:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*tDYxBA94Ne2SxPQeVxi_Rw.jpeg" /></figure><p>Similarly, the WARs of all eligible players are calculated by the Selections Committee. Players with the highest WARs are selected to represent Sri Lanka in international tournaments.</p><p>As of now, the quotas are:</p><ul><li><strong>World Scrabble Championship</strong>: 4 players</li><li><strong>World Youth Scrabble Championship</strong>: 10 players</li><li><strong>Asian Youth Scrabble Championship</strong>: 5 players</li></ul><p>Players who have not participated in tournaments for over a year are considered inactive. To be eligible again, they must play a <strong>minimum of 50 rated games</strong>.</p><p>According to the official selection criteria, this WAR-based system represents the most common and fair method for selection. It ensures that players are judged based on consistent and recent performance, with transparency in how ratings are weighted and selections are made.</p><p>(Proofread by Ramith Siriwardana — Chairman | Selections Committee SFSL 2024/2025)</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=567231d9c486" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[PLINRAT System]]></title>
            <link>https://medium.com/@imethdesilva/plinrat-system-566976311741?source=rss-1a39e89b963e------2</link>
            <guid isPermaLink="false">https://medium.com/p/566976311741</guid>
            <dc:creator><![CDATA[Imeth De Silva]]></dc:creator>
            <pubDate>Wed, 30 Apr 2025 00:47:00 GMT</pubDate>
            <atom:updated>2025-05-24T18:49:38.829Z</atom:updated>
            <content:encoded><![CDATA[<p>PLINRAT is the current rating system that is being used in Sri Lankan Scrabble to calculate the National Ratings.</p><p><a href="https://www.scrabblefederation.lk/ratings">https://www.scrabblefederation.lk/ratings</a></p><p>This article will extensively cover how the PLINRAT calculations work.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*pN8Tq6AIGFYe27jaxl0lHQ.jpeg" /></figure><p>First thing to understand is that there are two types of rated players; provisionally rated players and non-provisionally rated players.</p><p>Provisionally rated players are players who have played below 50 rated games. Those who have partaken in more than 50 rated games are non-provisionally rated.</p><p>The rating fluctuation of provisionally rated players is extremely high compared to non-provisionally rated players. In simple terms, this means that your ratings are not stable or most accurate until you have played more than 50 rated games. The formula for calculating the rating change or difference of a provisionally rated player is complex; therefore this article will only focus on the calculations of non-provisionally rated players.</p><p>In the PLINRAT system the following factors are taken into consideration when calculating one’s rating difference at the conclusion of a tournament.</p><ol><li>Average Opponent Rating</li><li>Expected Win Percentage</li></ol><p>For a concise understanding, let’s consider a round robin tournament between 4 players. Their ratings are as follows;</p><p>John — 1200</p><p>Steve — 1100</p><p>Dora — 900</p><p>Barbara — 700</p><p>(names are hypothetical btw)</p><p>A round robin tournament means all the players meet and play against each other. In a tournament like this, the average opponent rating can be calculated prior to the start of a tournament.</p><p>Considering from John’s perspective; his opponents’ total average rating will be 900 (Calculation below)</p><p>(1100 + 900 + 700) / 3 = 900</p><p>The next step to consider is the expected win percentage. How you achieve this is by subtracting John’s personal rating from the average opponent rating and dividing it by 12, which is a constant in the equation. Afterwards the value is added to +50 to get the expected win percentage. (See the example calculation below)</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/963/1*04NkDUeQKlO9TTVc252uew.jpeg" /></figure><p>John plays a total of 3 games. One each against Steve, Dora and Barbara.</p><p>Next, you calculated the number of expected wins. This is easily done by multiplying the expected win percentage with the number of games. Afterwards, you deduct the actual number of wins you have with the expected number of wins. Finally that is multiplied by a factor of 20, which is again a constant in this equation like the number 12 previously. The answer you get is the rating difference from the tournament. (See the example calculation below)</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/981/1*Z7trdZEq7GL8wAospm0rPA.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*0ReXpnoDCMXviYlGHgeDYg.jpeg" /><figcaption>Formula for Rating Difference</figcaption></figure><p>This simple example is how the PLINRAT system revolves around ratings. For tournaments which are not round robin, it is difficult to estimate your rating change as you do not know the exact opponents you may be facing. And this number may fluctuate to a certain extent if you are paired up with provisionally rated players who have not completed 50+ games.</p><p>The PLINRAT rating system is a simple enough rating system to understand. But I believe it fails to do justice in certain situations. For example, there have been instances where provisional players have gained 900+ or even 1000+ ratings after playing just one tournament or less than 10 games. While this number may adjust into their correct ratings after a period of time, it gives a bit of an unfair advantage particularly on the context when selections are considered through weighted ratings. (covered in another article) The system does not do full justice to non-provisionally rated players who started out on a weak rating. For them, the rating fluctuation happens on a small scale unless they do something incredible for 2–3 tournaments straight but the progress on ratings happens relatively slowly and may take a lot of time. In a simple example, there can be a 700 rated non-provisionally rated player who is leagues ahead of a 900 provisionally rated player at a certain point.</p><p>Until recently WESPA ratings also followed the PLINRAT system before adjusting to a new system called the RJR rating system, which appeared to be a better rating system maintaining fairness to many countries alike. But unlike the WESPA ratings base, which has 1000+ players; the Sri Lankan system at best may have around 200 players currently. So the feasibility of the new system should be carefully studied and tested based on the numbers. For your reference, I will link the full report of the RJR system provided by WESPA below.</p><iframe src="https://drive.google.com/viewerng/viewer?url=https%3A//wespa.org/wp-content/uploads/2024/10/WESPA-Ratings-V4-Report.pdf&amp;embedded=true" width="600" height="780" frameborder="0" scrolling="no"><a href="https://medium.com/media/8921556bc667a57be47e0453769b7262/href">https://medium.com/media/8921556bc667a57be47e0453769b7262/href</a></iframe><p>I am not an expert in the topic of ratings, so I am not sure if I made any inaccurate conclusions particularly about my opinions on the PLINRAT system. This article is just to give an outline on how the rating difference is calculated as I got this question asked multiple times.</p><p>If you are interested in discussing more about ratings or have any interesting suggestions about a new system feel free to reach out to me!</p><p><em>Proofread by Kavindu Malawaraarachchi (He knows a lot of stuff)</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=566976311741" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Machine Learning | Classification vs Regression Models]]></title>
            <link>https://medium.com/@imethdesilva/machine-learning-classification-vs-regression-models-d9c3b43c70e9?source=rss-1a39e89b963e------2</link>
            <guid isPermaLink="false">https://medium.com/p/d9c3b43c70e9</guid>
            <category><![CDATA[model-regression]]></category>
            <category><![CDATA[classification-models]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[machine-learning-models]]></category>
            <dc:creator><![CDATA[Imeth De Silva]]></dc:creator>
            <pubDate>Fri, 14 Mar 2025 00:27:01 GMT</pubDate>
            <atom:updated>2025-05-07T19:07:16.106Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/556/1*obW9bzULH2sdFJ7cpdvNaw.jpeg" /></figure><p><strong>Machine Learning Models</strong> are like different kinds of fortune-telling techniques — some predict categories, while others forecast exact numbers. But when should you use which? And how do they really work?</p><p>When we discuss about the basics of Machine Learning, the first thing we learn is Supervised Learning and Unsupervised Learning. In Supervised Learning, the algorithm is based off labeled data while in Unsupervised Learning it is based on unlabeled data. To simplify this a bit further, labeled data has both the input and output values whether it is categorical or numerical. The idea of labeled data or supervised learning is to map the inputs to the outputs and learn through the model in order to predict or label new and unseen data.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/700/1*0KxP69zsLRpz-sjfErFOdw.png" /></figure><p>Unlabeled data may contain the input values but with missing output values. The goal of unsupervised learning is to find patterns, structures and relationships between data.</p><p>The models in questions; Classification model and the Regression model are both models which fall under the category of Supervised Learning. The main characteristic or the difference between a Classification and Regression model is that the classification model will output a categorical value as the target variable, while a Regression model will output a numerical value.</p><p>To provide an example for this, classification models are used as algorithms that can predict true/false values, gender descriptions, spam detections etc. Regression models predict prices, salaries, data which has numbers.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/640/1*HFgNTD1wU7eRu81rdG_rbw.gif" /></figure><p>When you train a machine learning model, the main step is to preprocess the data properly and ensure that you have a fully cleaned dataset for the training process. When you train a model, the next step you do is to break the dataset into two parts; the training dataset and the testing dataset. Typically the ratio for this is 70–30 or 80–20.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*LoCLpvk29E5j4lgwUGMvJQ.png" /></figure><p>The main goal of a machine learning model is to generalize well. This means to predict an output value based on some input values. One of the main problems associated with Supervised Learning in particular is Overfitting and Underfitting.</p><p>Overfitting means, the model predict high accuracy on the training dataset and low accuracy on the test/unseen dataset. This generally means that the model has high variance and low bias.</p><p>Underfitting is when the model’s predictions are low in accuracy on the training dataset as well as weak predictions in the test dataset. This indicates that the Machine Learning model has low variance and high bias.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*UvuEB2Yq7d_4-6mE0JLw2A.png" /></figure><p>Examples for classification algorithms include K-Nearest Neighbors (KNN), Naive Bayes, Logistic Regression (LR) and Decision Tress (DT) classifications.</p><p>Examples for regression algorithms include Regression Trees, Linear Regression and Random Forest Regressor to name a few.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d9c3b43c70e9" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[A User-Friendly Guide to TSH (Tournament SHell)]]></title>
            <link>https://medium.com/@imethdesilva/a-user-friendly-guide-to-tsh-21a93a3dc6f3?source=rss-1a39e89b963e------2</link>
            <guid isPermaLink="false">https://medium.com/p/21a93a3dc6f3</guid>
            <dc:creator><![CDATA[Imeth De Silva]]></dc:creator>
            <pubDate>Tue, 24 Dec 2024 20:17:28 GMT</pubDate>
            <atom:updated>2025-07-08T07:06:28.441Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/894/1*uI0K-YDSFoz_GBz08J_n6w.png" /><figcaption>Generated standings from a round in TSH</figcaption></figure><p><strong>TSH (Tournament SHell)</strong> is a Scrabble tournament management software developed by John Chew. It is the most widely used and popular software among tournament directors and organizers for smoothly running tournaments. The software is highly convenient and supports multiple pairing systems such as Swiss Pairings, Round Robins, and King of the Hill (KOTH), among others.</p><p>The complete TSH guide is available at: <a href="https://www.poslarchive.com/tsh/doc/">https://www.poslarchive.com/tsh/doc/</a></p><p>I will break down the most important steps to help you get up and running with a tournament in no time.</p><ol><li>Setting up TSH</li><li>a.t</li><li>config.tsh</li><li>dryrun</li><li>Main commands</li><li>Accessing HTML files</li><li>Pros and Cons of TSH</li></ol><h3>1. Setting up TSH</h3><p>Installing TSH and its prerequisites is a simple process. TSH is a program written in Perl. For Windows users, you will need an IDE for Perl, Notepad, and the TSH zip archive.</p><p>TSH recommends two versions of Perl: <strong>ActivePerl</strong> and <strong>Strawberry Perl</strong>. Of these, Strawberry Perl is considered the most user-friendly. However, you can download either version to your device.</p><ul><li><strong>Strawberry Perl (recommended)</strong>: <a href="https://strawberryperl.com/">https://strawberryperl.com/</a></li><li><strong>ActivePerl</strong>: <a href="https://www.activestate.com/">https://www.activestate.com/</a></li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*s_JEt3CRDJfyuQPl3X45hQ.jpeg" /><figcaption>Strawberry Perl for Windows</figcaption></figure><p>Windows already comes with <strong>Notepad</strong> as its default text editor, so you can use that.</p><p>You will also need to download the TSH zip folder: <a href="https://www.poslarchive.com/tsh/tsh.zip">https://www.poslarchive.com/tsh/tsh.zip</a></p><p>Once downloaded, extract the folder and place it in a preferred location on your device.</p><h3>2. a.t</h3><p>To run a tournament with your preferred settings, two key files must be prepared: the a.t file and the configuration file.</p><p>The a.t file contains the list of players and information about their rating status.</p><p><strong>Step 1</strong> — Within the extracted TSH folder, create a new folder and rename it to your tournament’s name. Inside this newly created folder, create a new text file and name it a.t.<br> It is extremely important to use the correct filename and extension so that the compiler can recognize the file.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/947/1*x6C3pVhgZruAZ-3K-ihpaw.jpeg" /><figcaption>Creating an a.t file</figcaption></figure><p>Click <strong>“Yes”</strong> when prompted. Then, open the a.t file and enter the player information using the following format:</p><p><strong>Last Name, First Name | Rating</strong><br><em>If a player does not have a rating, you may assign a default value (e.g., 0).</em></p><p>Refer to the image below for an example.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/825/1*i9zdANhjWCRE7iQUI22D9w.jpeg" /><figcaption>Setting Up the a.t File</figcaption></figure><p>For <strong>team events</strong>, an additional extension is required after the player’s rating:<br>;; team &quot;Team Name&quot;</p><p><strong>Example:</strong><br>Doe, John 1000 ;; team TheWinners</p><p>There is no need to enter the names in alphabetical order or sort them by rating — player details can be entered in any order.<br> Once you have double-checked for spelling and formatting accuracy, save the a.t file.</p><h3>3. config.tsh</h3><p>The config.tsh file is where you define the core tournament settings, such as the number of rounds, tournament format, margin caps, and other controller options.</p><p>Similar to the a.t file, you will need to create a new text file and rename it to config.tsh, making sure to use the correct file extension so that TSH can recognize it.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/734/1*jZUdKAogZZ3II-op16kv4g.png" /><figcaption>Creating a config.tsh file</figcaption></figure><p>Click <strong>“Yes”</strong> if prompted to confirm the filename change.</p><p>The config.tsh file includes both<strong> settings</strong> and optional <strong>HTML blocks</strong> for styling — such as the tournament name, logo, date, and other display details.</p><pre>division a a.t<br><br>config event_name = &quot;Test Tournament 1&quot;<br>config event_date = &quot;May 20th, 2025&quot;<br><br><br>config html_top = &#39;&lt;body style=&quot;background-color:#ffffff&quot;&gt;&lt;/body&gt;&lt;p&gt;&lt;table align=center border=0 cellpadding=3 cellspacing=0&gt;&lt;tr&gt;&lt;td rowspan=2&gt;&lt;a href=&quot;https://www.testtournamentlogo/&quot;&gt;&lt;img src=&quot;C:\Users\DELL\Desktop\logo&quot; height=175 width=175 border=1px grey solid&gt;&lt;/a&gt;&lt;/td&gt;&lt;h1&gt;&lt;td&gt;&lt;font size=&quot;+2&quot;&gt;&lt;b&gt;Test Tournament 1&lt;br&gt;&lt;br&gt;2024&lt;/b&gt;&lt;/td&gt;&lt;/h1&gt;&lt;tr&gt;&lt;h1&gt;&lt;td&gt;&lt;font size=&quot;+1&quot;&gt;&lt;b&gt;May 20th, 2025&lt;/b&gt;&lt;/font&gt;&lt;/td&gt;&lt;/h1&gt;&lt;/a&gt;&lt;/tr&gt;&lt;/table&gt;&lt;/p&gt;&#39;<br><br>config assign_firsts = 1 <br>config bye_spread = 1<br>config final_draw = 1<br>config check_by_winner = 1<br>config pairing_system = &#39;auto&#39;<br>config board_stability = 1<br>config table_method = &#39;consecutive&#39;<br>config max_rounds = 18<br><br>autopair a 0 1 ns 0 0 a<br>autopair a 1 2 ns 0 1 a<br>autopair a 2 3 ns 0 2 a<br>autopair a 3 4 ns 0 3 a<br>autopair a 4 5 ns 0 4 a<br>autopair a 5 6 ns 0 5 a<br>autopair a 6 7 ns 0 6 a<br>autopair a 7 8 ns 0 7 a<br><br>autopair a 8 9 ns 0 8 a<br>autopair a 9 10 ns 1 9 a<br>autopair a 10 11 ns 1 10 a<br>autopair a 11 12 ns 1 11 a<br>autopair a 12 13 ns 2 12 a<br>autopair a 13 14 koth 2 13 a<br>autopair a 14 15 koth 2 14 a<br>autopair a 15 16 koth 3 15 a<br><br>autopair a 16 17 koth 3 16 a<br>autopair a 17 18 koth 3 17 a<br><br><br>config spread_cap = [200, 700, 700, 700, 700, 700, 700, 700, 700, 700, 700, 700, 700, 700, 700, 700, 700, 700]<br>config rating_system = &#39;aus&#39;<br>prize rank 1 a 25pts<br>prize rank 2 a 20pts <br>prize rank 3 a 17pts <br>prize rank 4 a 14pts <br>prize rank 5 a 12pts <br>prize rank 6 a 10pts <br>prize rank 7 a 8pts <br>prize rank 8 a 6pts <br>prize rank 9 a 4pts <br>prize rank 10 a 3pts <br>prize highwin 1 a &#39;Potential Trophy &amp; Certificate&#39;<br><br>config no_index_tally_slips = 1<br>config no_text_files = 1<br>config no_boards = 1<br>config gibson = 1<br>config gibson = &#39;strict&#39;<br>config prize_bands{&#39;A&#39;} = [1, 10]<br>config html_index_recent_first = 1<br>config show_inactive = 1<br>config session_breaks = [4,8,12,16,20]</pre><p>Above is a sample configuration block for an 18-round Scrabble tournament. We’ll break it down line by line to help you understand each part.</p><pre>division a a.t<br><br>config event_name = &quot;Test Tournament 1&quot;<br>config event_date = &quot;May 20th, 2025&quot;</pre><p>TSH allows you to run multiple divisions simultaneously.</p><ul><li>By default, the first division is labeled as division a.</li><li>If your tournament includes a second category (e.g., a Plate division), you can configure it by adding “division b b.t” beneath the line “division a a.t”</li><li><strong>Line 2</strong>: Specifies the <strong>name</strong> of the tournament.</li><li><strong>Line 3</strong>: Specifies the <strong>date</strong> of the tournament.</li></ul><pre>config html_top = &#39;&lt;body style=&quot;background-color:#ffffff&quot;&gt;&lt;/body&gt;&lt;p&gt;&lt;table align=center border=0 cellpadding=3 cellspacing=0&gt;&lt;tr&gt;&lt;td rowspan=2&gt;&lt;a href=&quot;https://www.testtournamentlogo/&quot;&gt;&lt;img src=&quot;C:\Users\DELL\Desktop\logo&quot; height=175 width=175 border=1px grey solid&gt;&lt;/a&gt;&lt;/td&gt;&lt;h1&gt;&lt;td&gt;&lt;font size=&quot;+2&quot;&gt;&lt;b&gt;Test Tournament 1&lt;br&gt;&lt;br&gt;2024&lt;/b&gt;&lt;/td&gt;&lt;/h1&gt;&lt;tr&gt;&lt;h1&gt;&lt;td&gt;&lt;font size=&quot;+1&quot;&gt;&lt;b&gt;May 20th, 2025&lt;/b&gt;&lt;/font&gt;&lt;/td&gt;&lt;/h1&gt;&lt;/a&gt;&lt;/tr&gt;&lt;/table&gt;&lt;/p&gt;&#39;</pre><p>The HTML code within config.tsh is used purely for <strong>styling the output pages</strong>.</p><pre>config html_top = &#39;&#39;&#39;<br>&lt;body style=&quot;background-color:#ffffff;&quot;&gt;<br>  &lt;p&gt;<br>    &lt;table align=&quot;center&quot; border=&quot;0&quot; cellpadding=&quot;3&quot; cellspacing=&quot;0&quot;&gt;<br>      &lt;tr&gt;<br>        &lt;td rowspan=&quot;2&quot;&gt;<br>          &lt;a href=&quot;https://www.testtournamentlogo/&quot;&gt;<br>            &lt;img src=&quot;C:/Users/DELL/Desktop/logo&quot; height=&quot;175&quot; width=&quot;175&quot; style=&quot;border:1px solid grey;&quot;&gt;<br>          &lt;/a&gt;<br>        &lt;/td&gt;<br>        &lt;td&gt;<br>          &lt;h1 style=&quot;font-size: 1.5em; margin: 0;&quot;&gt;&lt;b&gt;Test Tournament 1&lt;br&gt;2024&lt;/b&gt;&lt;/h1&gt;<br>        &lt;/td&gt;<br>      &lt;/tr&gt;<br>      &lt;tr&gt;<br>        &lt;td&gt;<br>          &lt;h2 style=&quot;font-size: 1.2em; margin: 0;&quot;&gt;&lt;b&gt;May 20th, 2025&lt;/b&gt;&lt;/h2&gt;<br>        &lt;/td&gt;<br>      &lt;/tr&gt;<br>    &lt;/table&gt;<br>  &lt;/p&gt;<br>&lt;/body&gt;<br>&#39;&#39;&#39;</pre><p>If you’re familiar with HTML, the structure will be easy to understand. If not, there’s no need to worry — you mainly need to update only a few key parts:</p><p>Replace the tournament name and year in the following line:</p><p>&lt;b&gt;Test Tournament 1&lt;br&gt;2024&lt;/b&gt;</p><p>You can include a logo stored locally on your computer by setting the src attribute to the image’s relative path:</p><p>&lt;img src=”C:/Users/DELL/Desktop/logo.png” height=”175&quot; width=”175&quot; style=”border:1px solid grey;”&gt;</p><p>The styling includes some basic inline CSS, which you can modify according to your preferences (e.g., adjusting the image size, borders, etc.).</p><pre>autopair a 0 1 ns 0 0 a<br>autopair a 1 2 ns 0 1 a<br>autopair a 2 3 ns 0 2 a<br>autopair a 3 4 ns 0 3 a<br>autopair a 4 5 ns 0 4 a<br>autopair a 5 6 ns 0 5 a<br>autopair a 6 7 ns 0 6 a<br>autopair a 7 8 ns 0 7 a<br><br>autopair a 8 9 ns 0 8 a<br>autopair a 9 10 ns 1 9 a<br>autopair a 10 11 ns 1 10 a<br>autopair a 11 12 ns 1 11 a<br>autopair a 12 13 ns 2 12 a<br>autopair a 13 14 koth 2 13 a<br>autopair a 14 15 koth 2 14 a<br>autopair a 15 16 koth 3 15 a<br><br>autopair a 16 17 koth 3 16 a<br>autopair a 17 18 koth 3 17 a</pre><h3>Configuring Tournament Pairings</h3><p>This section is crucial — it determines how the rounds in your tournament will be conducted. Here, you define the <strong>pairing system</strong>, whether it’s <strong>Round Robin</strong>, <strong>Swiss Pairing</strong>, or <strong>King of the Hill (KOTH)</strong>.</p><p>You can also configure the <strong>maximum number of repeats</strong>, which specifies how many times a player can be paired against the same opponent during the tournament.</p><p>In the example provided, the tournament consists of <strong>18 rounds</strong>, and each pairing instruction begins with the keyword autopair.</p><p>TSH supports two main types of pairings:</p><ul><li><strong>autopair</strong> – Allows the system to automatically generate pairings for each round.</li><li><strong>pm (pair manual)</strong> – Enables manual pairing, where you choose which players face each other.</li></ul><p>Most tournaments are run using <strong>autopair</strong>, and we will focus on this method in this guide.</p><p>Each autopair line in TSH contains a sequence of parameters that define <strong>how a round should be paired</strong>, based on previous results, the pairing method, and repeat constraints.</p><pre>autopair a 0 1 ns 0 0 a</pre><p>Let’s break this down:</p><ul><li><strong>a</strong> – Indicates the <strong>division</strong> (e.g., Division A).</li><li><strong>0</strong> – Refers to the <strong>previous round</strong> whose results are considered. Since this is <strong>Round 1</strong>, there are no previous results, so it starts with 0.</li><li><strong>1</strong> – The <strong>current round</strong> being paired.</li><li><strong>ns</strong> – The <strong>draw type</strong>, in this case <strong>Swiss Pairing</strong>.</li><li><strong>0</strong> – The <strong>maximum number of repeats</strong> allowed <em>up to that point</em>. Since it’s the first round, no repeats are possible.</li><li><strong>0</strong> – This number typically <strong>matches the round number used earlier</strong> (i.e., the 0 referring to the previous round).</li><li><strong>a</strong> – Marks the <strong>division</strong> again to confirm which division this pairing applies to.</li></ul><h3>Common Draw Types in TSH</h3><p>TSH supports several draw types:</p><ul><li><strong>ns</strong> – <strong>Swiss Pairing</strong><br> Pairs players based on similar win records and ratings. In Round 1, higher-rated players are generally matched against lower-rated ones. This system is widely accepted and commonly used.</li><li><strong>rr</strong> – <strong>Round Robin</strong><br> Every player faces <strong>all other players</strong> in the tournament.</li><li><strong>koth</strong> – <strong>King-of-the-Hill</strong><br> Players are paired strictly based on standings. For example, 1st plays 2nd, 3rd plays 4th, and so on.</li><li><strong>random</strong> – <strong>Random Pairing</strong><br> Players are paired <strong>completely randomly</strong>, without consideration of rating or past performance.</li></ul><p>⚠️ Random pairings are <strong>not recommended</strong> for most tournaments, as they may negatively affect draw quality and fairness.</p><h3>Setting Maximum Repeats</h3><ul><li>The <strong>repeat count</strong> controls how many times a player can face the same opponent.</li><li>In <strong>Round 1</strong>, this value is set to 0 (as no repeats can happen yet).</li><li>If you want to <strong>prevent repeats until Round 10</strong>, you would set the repeat count to 0 for the first 9 rounds.</li><li>Starting from <strong>Round 10</strong>, you can increase this value to 1, allowing players to potentially face the same opponent again.</li><li>The tournament can allow a <strong>maximum of 3 repeats</strong> in total, configurable based on your preferences.</li></ul><h3>In this example configuration:</h3><ul><li>The tournament has <strong>18 rounds</strong>.</li><li><strong>Rounds 1–13</strong> use <strong>Swiss Pairing (</strong><strong>ns)</strong>.</li><li><strong>Rounds 14–18</strong> use <strong>King-of-the-Hill (</strong><strong>koth)</strong>.</li><li>The <strong>maximum number of repeats allowed</strong> is <strong>3</strong>.</li></ul><h3>4. dryrun</h3><p>Before officially running your tournament using the config.tsh file, it is highly recommended to perform a <strong>dry run</strong>. This step helps ensure that your configuration and player data are working as intended.</p><p>A dry run allows you to:</p><ul><li>Identify errors such as <strong>exceeding the maximum number of repeats</strong>, <strong>insufficient number of players</strong>, or <strong>invalid formatting</strong>.</li><li>Estimate how many <strong>wins</strong> a player might need to secure the tournament.</li><li>Determine whether <strong>Gibsonization</strong> (early pairing of a clear winner) is likely to occur.</li></ul><ol><li>Ensure that both the a.t and config.tsh files are correctly placed <strong>inside your tournament folder</strong>.</li><li>TSH automatically recognizes the <strong>most recently updated folder</strong>, so double-check that the folder for your test tournament has the correct and up-to-date files.</li></ol><p>Your folder structure should look something like this:</p><pre>TSH/<br>└── TestTournament/<br>    ├── a.t<br>    └── config.tsh</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*L2yjdIEVHwX1lADLS5sPyQ.png" /><figcaption>Test Tournament Folder</figcaption></figure><p>Navigate to the tsh.pl file and double-click it to launch the TSH interface.</p><figure><img alt="Navigate to tsh.pl file and double click it. This will open up TSH" src="https://cdn-images-1.medium.com/max/1024/1*W5nOAdymvdZ9LM2PLYfRHw.png" /><figcaption>TSH Files</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*OZq2g2Tzfz12_1-fbAeR0Q.png" /><figcaption>TSH terminal</figcaption></figure><p>In the terminal (Command Prompt), you can start by typing the command rosters to view the list of registered players. Then, type dryrun and press <strong>Enter</strong> to run a simulated version of the tournament. This will highlight any potential errors, such as invalid pairings or repeat violations.</p><p>Once the dry run is complete (it should only take a few seconds), type reseteverything to reset your a.t and config.tsh files to their initial state.</p><p><em>(Refer to the video below for a step-by-step demonstration.)</em></p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FpC6C9wJJF70%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DpC6C9wJJF70&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FpC6C9wJJF70%2Fhqdefault.jpg&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/aad0967db242cb93b299d74f0769d0cc/href">https://medium.com/media/aad0967db242cb93b299d74f0769d0cc/href</a></iframe><h3>5. Main Commands</h3><h4>There are a few commands you need to familiarize yourself with in order to use TSH effectively and to manage any tournament. Those commands are summarized below;</h4><h3>1. Show Pairings (sp)</h3><p><strong>Command:</strong> sp 1 a<br>Displays pairings for Round 1. Change number to view results from any other specific round. “a” stands for division.</p><h3>2. Enter Scores (a)</h3><p><strong>Command:</strong> a 1 a<br>Starts score entry for Round 1.<br>Format:<br>p1 s1 p2 s2 → player number/name + score.<br>Type “m” to check for missing results.</p><h3>3. Check Round Scores (crs)</h3><p><strong>Command:</strong> crs 1<br>Cross-check entered data for Round 1 with results slips.</p><h3>4. Generate Ratings/Standings (rat)</h3><p><strong>Command:</strong> rat<br>Produces player standings after a round.<br>You can Print and post the ratings.</p><h3>5. Correct Score Entry Errors</h3><p><strong>Command:</strong> es a X Y <br>Where:</p><ul><li>a = Division</li><li>X = Player name/number</li><li>Y = Round number</li></ul><p>(es stands from edit score)</p><p>Then enter:<br>ps os → new player score and opponent score (type the correct figures in order)</p><p>Example: es a 15 5 → enter → 400 300</p><h3>6. No-Shows (without notice)</h3><p><strong>Command:</strong> floss a X R 75<br>Applies a 75-point forfeit loss. (forfeit margin can be customized)</p><h3>7. Withdrawals (with notice)</h3><p>Before pairings:</p><ul><li>es a X 0</li><li>off -75 → player gets -75 for each game missed (withdrawal margin can be customized)</li></ul><p>When they return:</p><ul><li>es a X 0</li><li>on</li></ul><h3>6. Accessing HTML files</h3><p>Once you use TSH to generate even a single round of pairings or standings, the software will automatically create the relevant <strong>HTML output files</strong> for the tournament.</p><p>If you navigate back to your tournament folder, you’ll notice that, in addition to the original a.t and config.tsh files, three new folders have been created:</p><ul><li>html</li><li>old</li><li>tsh.lock</li></ul><p>For practical purposes, you only need to be concerned with the <strong>html</strong> folder, which contains all the updated standings, pairings, and results in a web-friendly format.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ons3D21xXnhTRbyxTNVrQA.png" /><figcaption>HTML Files from TSH</figcaption></figure><p>Once you open the html folder, you will see multiple HTML files generated based on the commands you have run in TSH.</p><p>You can double-click any of these files to view pairings or standings for a specific round in your web browser.</p><p>Additionally, these files can be uploaded to a website and styled to your preference.</p><p>To get started, locate and open the file named index.html by double-clicking on it.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*htHa28_revDaN4Kh6nXuYQ.png" /><figcaption>HTML Files</figcaption></figure><p>This will open the tournament’s main index page, from which you can navigate to all other generated results and reports.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*rVjRaDQ9iWODxSOnAAdUHA.png" /><figcaption>index.html</figcaption></figure><p><strong>Tip:</strong> In the TSH terminal, type the command stats and press <strong>Enter</strong>. This will generate all the default statistics based on the entered results.</p><p>The generated stats will include details such as:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/676/1*VCODWniPDssd-yjU1fdtFQ.png" /><figcaption>stats</figcaption></figure><h3>7. Pros and Cons of TSH</h3><p>Pros:</p><ul><li>Highly configurable and customizable tournament settings to suit your specific needs.</li><li>Supports multiple pairing systems, including Round Robin, Swiss Pairing, and King-of-the-Hill.</li><li>User-friendly commands that require no prior programming knowledge.</li><li>Instantly generates standings, pairings, and statistics.</li><li>Convenient and widely regarded as the most popular Scrabble tournament software worldwide.</li></ul><p>Cons:</p><ul><li>The terminal-based interface can appear more complicated than it actually is.</li><li>Limited flexibility for editing tournament details once the event is underway.</li><li>Less convenient for managing team tournaments.</li></ul><h3>SUMMARY</h3><blockquote>TSH (Tournament SHell) is a powerful and reliable tool for organizing Scrabble tournaments of any size. While its command-line interface might seem intimidating initially, it offers a wide range of pairing formats, score tracking, and results generation options.</blockquote><blockquote>With just two essential setup files — a.t for player registration and config.tsh for tournament structure — you can have your tournament up and running in minutes. Because of its flexibility and efficiency, TSH remains the preferred software for tournament directors around the world.</blockquote><blockquote>Once you become familiar with the basic commands and perform dry run tests, managing your tournament will become smooth and straightforward. TSH provides everything you need to keep your competition fair, well-organized, and efficient.</blockquote><p>If you need any assistance with setting up TSH or have any questions, feel free to reach out to me at <strong>imethdesilva@gmail.com</strong> :)</p><p><em>(Special thanks to Kavindu Malawaraarachchi for proofreading this guide — he knows TSH better than me!)</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=21a93a3dc6f3" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Your first drum kit?]]></title>
            <link>https://medium.com/@imethdesilva/your-first-drum-kit-347c8a17a4d9?source=rss-1a39e89b963e------2</link>
            <guid isPermaLink="false">https://medium.com/p/347c8a17a4d9</guid>
            <category><![CDATA[drum-head]]></category>
            <category><![CDATA[drum-kit]]></category>
            <category><![CDATA[cymbal]]></category>
            <category><![CDATA[drums]]></category>
            <dc:creator><![CDATA[Imeth De Silva]]></dc:creator>
            <pubDate>Wed, 27 Nov 2024 05:10:59 GMT</pubDate>
            <atom:updated>2024-11-27T05:17:01.689Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/728/1*Tf74U4tK3NpOtpB57JuVTw.jpeg" /></figure><p>Finally decided to write something down on Medium...cause why not?</p><p>I remember asking a gazillion opinions before buying my first drum kit. 2019 was 1/4th of my life ago! Truth is you can’t find many great options here in Sri Lanka. If you want one of the top drum kits (and you are rich to the brim) you can get it imported from the US or Japan.</p><p>To name a few of the models; Pearl, Yamaha, Tama, Mapex, Ludwig and Sonor will top the charts. I think one of the main important features is the hardware. You don’t want to be hammering at a cymbal only for your cymbal stand to break down a couple of weeks after spending a fortune on the kit! In terms of hardware, whatever kit you buy from one of these recognized models will last.</p><p>When you are choosing your first drum kit, hardware matters. Make sure you check the quality of the stands, especially the cymbal stands. Mine broke a year or two back, but to be fair it lasted a decent 3–4 years. The model I have is a Pearl RoadShow. But-it’s the Chinese copy; not the original.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/728/1*6ltOd8E1jdNHW1fDa72cnA.jpeg" /></figure><p>If you go to typical music stores in Sri Lanka, what is mostly available is the Chinese copy and hardly the original of the manufacturing brand. For originals, I recommend Yamaha in Colombo. To my knowledge, that’s the best place.</p><p>The Chinese copy is not bad at all. My kit has been surviving for 5+ years now and given the hours and hours of practice I have been going on at it, I would give it some credit for lasting this long. But there have been many upgrades along the way.</p><p>If you are buying a Chinese copy it will be generally way cheaper than buying an original. If I remember correct, back in 2019 I got the Pearl RoadShow for about 60,000 LKR (200–250 USD) Good luck trying to find a kit for that price these days!</p><p>Always, ALWAYS check the hardware. It’s not going to be good as the hardware components that come with original kits. Just double check that it’s strong enough to last at least a couple of years. (You also have the option of purchasing the stands separately)</p><p>Now that wehave established the importance of the hardware components, we come to our next topic. The cymbals and heads. To put it in short — the CYMBALS ARE A NIGHTMARE. Even if you buy an original kit, the cymbals that come with it are factory made and maybe it will sound good when you hit it once or twice. But thrice is not going to make the cut. Statistically (I think) 75% of the cymbals that come originally with the drum kit sound like something that is being cooked in a kitchen or if you are from Sri Lanka, it’s like that noise that comes from the Koththu stations.</p><p>But hey — the Pearl RoadShow crash cymbal surprisingly demonstrated better results (than my grades! — jk) It has been the ugly duckling in my family of original cymbals and to be honest, sounds good for something that came in a Chinese copy!</p><p>Ok back to the topic. So even if you buy an original drum kit, you will eventually have to purchase cymbals and drum heads separately to really elevate the standard of your kit. Just like drum kit brands, there are recognized and recommended brands for this.</p><p>If we are speaking in the context of cymbals; the go to options are Paiste, Sabian and Zildjian. Oh there’s Meinl too. If you do a research on any of the top drummers you know, be it in YouTube or concerts or whatever; there won’t be a drummer who does not use one of these brands. The cymbals are an important component to drumming. If it sounds like as I mentioned before; a Koththu station, trust me, not even your cat is gonna wanna hear what you’re playing.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*hxrn0uaHs-hcx194rfzAHg.jpeg" /></figure><p>If you are just buying your first kit, you can first invest on a good crash cymbal. Then hi-hats and then a ride. Having a good ride cymbal just gives you that extra touch of satisfaction. I currently have a Sabian hi-hat set and also a Zildjian ride. (Not a big fan of the ride cymbal, but the hi-hats sounded decent for a year or two)</p><p>Finally, we come to talk about the drum heads. Try as you may, but the drum heads that originally come with your kit is going to sound bad after a week or two. Shifting to one of the recognized brands like Remo or Evans is advisable. There are many variations on choosing the correct drum heads. For this, a quick YouTube search on comparison of drum heads will help you find the right drum head with the tone and sound you are looking for. There’s no better feeling as a drummer, when you finally discover that perfect snare shot you have been mystically searching on for years. (It’s been 5+ years for me and the search still goes on!)</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2Fj5BpRNNwzi4%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3Dj5BpRNNwzi4&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2Fj5BpRNNwzi4%2Fhqdefault.jpg&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/99be8caaa85df721bdc2dfb7796ec69b/href">https://medium.com/media/99be8caaa85df721bdc2dfb7796ec69b/href</a></iframe><p>Any drum head you buy should be fine for you as a beginner. The next bit is the tuning part. Ahhhh — this is where things can get really frustrating. Tuning the kit is a bit of a tedious process but with experience you will get the hang of it. The easiest way to tune the drum heads is to remove all the tension rods and put the new head and then exert some pressure on to the drum head. Then slowly start to tighten the 6 or 8 tension rods in the drum head; firstly tighten them as much as possible just by using your fingers. Then, when you can tighten it no more; exert more pressure into the drum head and then use your drum tuner to tighten it until the wrinkles around the perimeter start to disappear. That should give you the basic tone you are looking for, and then you can readjust based on your preference.</p><p>So to summarize, if you are buying a new kit; check the hardware components. Check whether it’s a Chinese copy or an original from the manufacturer. Invest on some branded cymbals and drum heads. Learn the basics of tuning. Now you are good to go Jeff Porcaro around the kit!</p><blockquote>Imeth De Silva | 27th Nov, 2024</blockquote><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=347c8a17a4d9" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>