<?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[Systematic Sports - Medium]]></title>
        <description><![CDATA[Data not Luck - Medium]]></description>
        <link>https://medium.com/systematic-sports?source=rss----b6325b546080---4</link>
        <image>
            <url>https://cdn-images-1.medium.com/proxy/1*TGH72Nnw24QL3iV9IOm4VA.png</url>
            <title>Systematic Sports - Medium</title>
            <link>https://medium.com/systematic-sports?source=rss----b6325b546080---4</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sat, 21 Nov 2026 21:50:26 GMT</lastBuildDate>
        <atom:link href="https://medium.com/feed/systematic-sports" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[The MLS Betting Paradox: Beating the League That Makes No Sense]]></title>
            <link>https://medium.com/systematic-sports/why-betting-on-mls-feels-impossible-and-how-you-can-profit-from-it-f87dd41b255c?source=rss----b6325b546080---4</link>
            <guid isPermaLink="false">https://medium.com/p/f87dd41b255c</guid>
            <category><![CDATA[football]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[python]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[programming]]></category>
            <dc:creator><![CDATA[Daniel Guerrero]]></dc:creator>
            <pubDate>Wed, 02 Apr 2025 20:57:55 GMT</pubDate>
            <atom:updated>2025-04-02T21:01:06.488Z</atom:updated>
            <content:encoded><![CDATA[<p>MLS is the hardest league to bet on, but hidden inefficiencies make it beatable — if you know where to look.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/974/1*9WpKfdNhZT68AMzLiWNk4w.png" /></figure><p>If you ask sports bettors whether you can make money betting on Major League Soccer (MLS), most will tell you it’s <strong>too random</strong> or even <strong>impossible</strong>. But why?</p><p>What makes MLS harder to predict than the big five European leagues? After all, it’s the same sport, right? And if MLS is truly unpredictable, shouldn’t there be a way to exploit that unpredictability and turn it into an edge?</p><p>In this article, I’ll break it down with data and statistical analysis. If you’re short on time, here’s the answer: Yes, you can make money betting on MLS — because bookies struggle to price it correctly.</p><h4>Is MLS Harder to Predict Than the Top 5 Leagues?</h4><p>To answer this, we don’t need to build our own predictive models. Instead, we can analyze the sharpest bookie in the world — <strong>Pinnacle</strong>. Bookmakers have vast statistical models to set their odds, and since they adjust them based on betting volume, we can assume their final odds closely reflect true probabilities.</p><p>A great way to measure how accurate their odds are is the Brier Score, which quantifies the accuracy of probability predictions (lower is better). Here’s how MLS compares to the big five leagues:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*PeNJyNUPrwtc5UvmUh8DAw.png" /></figure><p>👉 <strong>MLS has the worst Brier Score</strong> of all these leagues, meaning bookies struggle the most to assign reliable probabilities to match outcomes.</p><p>What happens when we look at accuracy — how often the most probable outcome actually occurs?</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*4Quf4ZUU9UoPVL4pd88ZaQ.png" /></figure><p>📉 <strong>Bookies correctly predict MLS results only 47% of the time</strong> — a full 13% lower than the Premier League. That’s a huge difference.<strong> If the sharpest bookie in the world struggles to predict MLS, there must be inefficiencies we can take advantage of</strong>.</p><p>— -</p><h4>Home Advantage in MLS: A Key Factor?</h4><p>Home advantage is often a critical factor in soccer betting models. Let’s compare how often the home team wins across leagues:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*zSeQGthB0nkFjwWDMbB_8g.png" /></figure><p>🔍 <strong>MLS has one of the strongest home advantages</strong>, comparable to the Premier League. This suggests that any profitable MLS betting model should strongly consider home advantage.</p><p>But are bookies pricing it correctly? Looking at the Brier Scores for specific match outcomes:</p><ul><li>Home Win: 0.23</li><li>Draw: 0.22</li><li>Away Win: 0.21</li></ul><p>📊 <strong>Bookies struggle the most with home win probabilities</strong>. This means that home teams frequently fail to win when expected, creating potential inefficiencies.</p><p><strong>Does xG Performance Predict Future Goals in MLS?</strong></p><p>Expected Goals (xG) is one of the most widely used metrics for predicting soccer outcomes. But how well does past xG performance correlate with actual goals scored?</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Uly58gVFKovMbczrTO-67g.png" /></figure><p>📉 <strong>MLS has the weakest correlation between past xG performance and goals scored</strong>. This suggests that teams in MLS are less consistent at converting their xG into actual goals, making prediction models based purely on xG less reliable.</p><p>However, what happens when we add home advantage into the equation?</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*wOhrdL1mdGVE6N3s99_yww.png" /></figure><p>📊 <strong>Adding home advantage significantly improves the correlation</strong> for MLS, more so than for any other league. This reinforces our earlier finding: home advantage plays an even bigger role in MLS than elsewhere.</p><h4><strong>How Can You Profit From Betting on MLS?</strong></h4><p>Let’s summarize our findings:</p><p>✅ MLS is the hardest league for bookies to predict <br>✅ Bookies struggle the most with pricing home odds <br>✅ Past xG performance is weakly correlated with goals, but home advantage strengthens the correlation <br>✅ Home teams often lose unexpectedly — creating mispriced betting opportunities</p><p>By combining these insights, we built a <strong>profitable MLS betting model at Systematic Sports</strong>, which:</p><p>📈 Generated <strong>over $500 in profit</strong> in past seasons <br>📊 Delivered a <strong>58% cumulative return</strong> <br>🚀 Uses home advantage and xG inefficiencies to find mispriced odds</p><p>Want to see how it works? Check out our fact sheet <a href="https://www.systematicsports.co.uk/betting-tips/value-model"><strong>here</strong></a> and sign up for a <a href="https://whop.com/value-betting-strategy/"><strong>free trial</strong></a> to access the model today!</p><h4>Final Thoughts</h4><p>The data is clear: <strong>MLS is chaotic, but that chaos creates opportunity.</strong> While most bettors struggle with MLS randomness, the inefficiencies in bookie pricing mean there’s a way to profit — if you know what to look for.</p><p>If you’re serious about making money from soccer betting, don’t ignore MLS. Instead, use its unpredictability to your advantage.</p><p>📊 Want to get started? Check out our free trial and start betting smarter today!</p><h4><strong>The Importance of Bet Sizing</strong></h4><p>Understanding how much to bet is just as crucial as knowing what to bet on. Poor bet sizing can lead to unnecessary risk, while optimal bet sizing can maximize long-term profits. If you’re interested in learning more about how to determine the right bet size using mathematical principles like the Kelly Criterion, check out this article:</p><p><a href="https://medium.com/systematic-sports/bet-sizing-and-the-kelly-criterion-4939fad24d45">Bet Sizing and the Kelly Criterion</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f87dd41b255c" width="1" height="1" alt=""><hr><p><a href="https://medium.com/systematic-sports/why-betting-on-mls-feels-impossible-and-how-you-can-profit-from-it-f87dd41b255c">The MLS Betting Paradox: Beating the League That Makes No Sense</a> was originally published in <a href="https://medium.com/systematic-sports">Systematic Sports</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Systematic Betting Strategy]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://medium.com/systematic-sports/systematic-betting-strategy-8581e90d2ed8?source=rss----b6325b546080---4"><img src="https://cdn-images-1.medium.com/max/1179/1*Pe8Pbfiv30tc5QD9NhWV0w.jpeg" width="1179"></a></p><p class="medium-feed-snippet">Betting &#xA3;1,000 on the algorithm and &#xA3;1,000 on my ball knowledge for Season 4</p><p class="medium-feed-link"><a href="https://medium.com/systematic-sports/systematic-betting-strategy-8581e90d2ed8?source=rss----b6325b546080---4">Continue reading on Systematic Sports »</a></p></div>]]></description>
            <link>https://medium.com/systematic-sports/systematic-betting-strategy-8581e90d2ed8?source=rss----b6325b546080---4</link>
            <guid isPermaLink="false">https://medium.com/p/8581e90d2ed8</guid>
            <category><![CDATA[football]]></category>
            <category><![CDATA[statistics]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[football-betting]]></category>
            <category><![CDATA[technology]]></category>
            <dc:creator><![CDATA[Liam Hartley]]></dc:creator>
            <pubDate>Mon, 23 Sep 2024 07:39:30 GMT</pubDate>
            <atom:updated>2024-10-04T08:00:06.770Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[Five secrets of our Football Betting Strategy]]></title>
            <link>https://medium.com/systematic-sports/five-secrets-of-our-football-betting-strategy-e3c188d5c335?source=rss----b6325b546080---4</link>
            <guid isPermaLink="false">https://medium.com/p/e3c188d5c335</guid>
            <category><![CDATA[football]]></category>
            <category><![CDATA[sports-betting]]></category>
            <category><![CDATA[entrepreneurship]]></category>
            <category><![CDATA[quantitative-finance]]></category>
            <category><![CDATA[betting]]></category>
            <dc:creator><![CDATA[Estèphe]]></dc:creator>
            <pubDate>Mon, 16 Sep 2024 08:31:58 GMT</pubDate>
            <atom:updated>2024-10-04T08:00:15.786Z</atom:updated>
            <content:encoded><![CDATA[<h3>Our Five Football Betting Strategy Secrets</h3><h4>Systematic Sport’s Value Betting Strategy for Season 4</h4><p>Last season, our <a href="https://www.systematicsports.co.uk/betting-tips/value-model">Football Value Betting Strategy</a> delivered a 23% return.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/499/1*FJU0SEbX6WfI1zY3mvvu9A@2x.jpeg" /><figcaption>Season 3 Performance of our Value Betting Strategy</figcaption></figure><p>This season, we’re setting our sights even higher. Here’s how we’ve taken our betting strategy to the next level for our fourth Season (<a href="https://whop.com/systematicsports/?pass=prod_qojRlElwICSZ8&amp;plan=plan_liVTQglWhouBL">Season 4</a>).</p><h4><strong>Evolving our Strategies</strong></h4><p>Staying ahead of the curve in sports betting means continuously refining your strategy. Over the last five years, Systematic Sports has significantly enhanced its betting models for each season.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/640/1*V3syw_UBRZ7GVT2UDCF1_w@2x.jpeg" /><figcaption>Backtested performance of the <a href="https://www.systematicsports.co.uk/betting-tips/value-model">Season 4 Football Value Betting Strategy</a></figcaption></figure><p>For Season 4, we continued to improve three key metrics:</p><ol><li>Annualised return: Average yearly profit of an investment over time.</li><li>Sharpe Ratio: The higher, the better. Sharpe measures how much return you get for the risk you take. Top fund managers in finance aim for a Sharpe ratio above 1.5.</li><li>Maximum drawdown: The lower, the better. The maximum drawdown shows the most significant drop from a portfolio’s peak to its lowest point. Top managers aim to keep the maximum drawdown below 20%.</li></ol><p>We’ve added five new signals to leverage a more granular data set to improve these metrics.</p><p><a href="https://medium.com/systematic-sports/backtesting-a-sports-betting-strategy-283833a5eca3">Backtesting a Sports Betting Strategy</a></p><h4><strong>1. Factoring in Player Skill: Adjusting for xG Outliers</strong></h4><p>Expected Goals (xG) is a powerful metric, but it doesn’t always tell the whole story. Some players consistently outperform their xG due to superior finishing skills. In Season 4, we’re adjusting xG figures for these players.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/717/1*xSlDEkmtf10BOv1FYRYGKw.png" /><figcaption>Top and bottom finishers since 2020</figcaption></figure><p>By recognising and rewarding consistent over-performers, our model now accounts for individual player skill levels, providing a more accurate reflection of likely outcomes.</p><h4>2. Home-Away Advantage: Going Granular</h4><p>Home-field advantage isn’t a one-size-fits-all metric. Some teams thrive at home, while others benefit less from the home crowd.</p><p>This season, we’ve introduced a more granular approach to factoring home-away advantage into our predictions. Instead of applying a generic adjustment, we now analyse team-specific home and away performance since 2015 (excluding 2020 for COVID), allowing for a more tailored prediction model.</p><h4>3. Factoring Team ELO: Betting Smarter Against Favorites</h4><p>In Season 3, we often bet too heavily against favourites, targeting the value in the odds offered for the underdog.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*6fluhKPgr26R7QkhFCsqxg.jpeg" /><figcaption>Four big teams in the Premier League</figcaption></figure><p>We’ve integrated a team ELO rating into our model to address this. ELO is a powerful indicator of a team’s strength relative to others. By incorporating it, we now have a more balanced view of matchups, particularly when assessing underdogs against top-tier teams.</p><h4>4. Decay Factor: Adjusting for Time</h4><p>Teams evolve throughout the season and dip in and out of form. A game played three months ago doesn’t have the same relevance as one played last week. That’s why we’ve introduced a decay factor to our model.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*NYhgKKsj4Z4LQGYDePl2fQ.png" /><figcaption>Example of an exponentially decaying weight applied to fixture-level metrics.</figcaption></figure><p>This means the older a match, the less weight its statistics have in our current predictions. Doing so ensures that our models stay sharp, focusing on the most relevant and timely information.</p><p>We previously weighed the last six gameweeks equally in our calculations. This new approach prevents outdated data from skewing predictions, ensuring our strategy is always grounded in the most current and relevant performance metrics.</p><h4>5. Recalibrating Value and Probability Thresholds with Backtesting</h4><p>Finally, we’ve recalibrated our value and probability thresholds using our backtesting engine. This has fine-tuned our “value bet” criteria, ensuring we’re not just picking winners but doing so with calculated profitable precision.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/596/1*jlM4LP4iShSouHNv4neAYA.png" /><figcaption>A backtest run of £1,000 for multiple Probability and Value Thresholds for our Season 4 Strategy</figcaption></figure><p>This recalibration ensures that every bet we take meets stringent criteria, reducing the noise and focusing on bets with the highest expected value.</p><h3>Ready to Upgrade Your Strategy?</h3><p>Our model now consumes more metrics and information than ever for Season 4, and the backtests are the best they have ever been.</p><p><a href="https://whop.com/systematicsports/">Systematic Sports - Get Access</a></p><p>Are you interested in seeing these strategies in action?</p><p>Join us at Systematic Sports and elevate your betting game with data not luck.</p><p><a href="https://t.me/systematicsports">Systematic Sports</a></p><p><a href="https://www.youtube.com/watch?v=Iz9WtenP9U0">Video Version of Article</a></p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FIz9WtenP9U0&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DIz9WtenP9U0&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=google" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/ed17fd83d360592b1cfbbb217dba628f/href">https://medium.com/media/ed17fd83d360592b1cfbbb217dba628f/href</a></iframe><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e3c188d5c335" width="1" height="1" alt=""><hr><p><a href="https://medium.com/systematic-sports/five-secrets-of-our-football-betting-strategy-e3c188d5c335">Five secrets of our Football Betting Strategy</a> was originally published in <a href="https://medium.com/systematic-sports">Systematic Sports</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Winning at Sports Betting: Scraping and Analyzing Odds Data with Python]]></title>
            <link>https://medium.com/systematic-sports/winning-at-sports-betting-scraping-and-analyzing-odds-data-with-python-b77896286a4?source=rss----b6325b546080---4</link>
            <guid isPermaLink="false">https://medium.com/p/b77896286a4</guid>
            <category><![CDATA[beautifulsoup]]></category>
            <category><![CDATA[sports-betting]]></category>
            <category><![CDATA[scraping]]></category>
            <category><![CDATA[selenium]]></category>
            <category><![CDATA[ml-so-good]]></category>
            <dc:creator><![CDATA[Daniel Guerrero]]></dc:creator>
            <pubDate>Tue, 27 Aug 2024 09:41:25 GMT</pubDate>
            <atom:updated>2024-08-27T09:41:25.764Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*3vRTvzuDJc5vxJEh" /><figcaption>Photo by <a href="https://unsplash.com/pt-br/@cgower?utm_source=medium&amp;utm_medium=referral">Christopher Gower</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>Are you looking for an edge in sports betting? Sports betting can be lucrative, but it requires careful analysis and accurate predictions to generate profits. One way to gain an edge is by analyzing odds data, which can help you identify value bets and find the best value for your bets. However, analyzing odds data comes with challenges, such as the large volume of data and the need for accurate predictions. This post will show you how to scrape and analyze odds data using Python and BetExplorer.com.</p><p>You can find this code and other projects on my <a href="https://github.com/danielguerreros">GitHub</a>. You can also follow me on <a href="https://twitter.com/danielguerrer0s">Twitter</a>.</p><h4>Average Odds</h4><p>We’ll first show you how to scrape average odds data from BetExplorer.com using Python’s requests and BeautifulSoup libraries. Average odds represent the market consensus on the likely outcome of a match. They can help you identify matches where there is a discrepancy between the market and your own analysis.</p><p>Firstly, we will define the URL that will be used. Then, we will use the Selenium package with the Google Chrome driver to retrieve the HTML data from the page and BeautifulSoup to parse it. Finally, we will use find to locate the table that contains the odds data and extract it into a Pandas DataFrame.</p><pre>from selenium import webdriver<br>from selenium.webdriver.chrome.service import Service<br>from webdriver_manager.chrome import ChromeDriverManager<br><br>URL = &quot;https://www.betexplorer.com/football/england/premier-league-2023-2024/results/&quot;<br>driver = webdriver.Chrome(service=Service(ChromeDriverManager().install()))<br>driver.get(URL)<br>soup = BeautifulSoup(driver.page_source, &#39;html.parser&#39;)<br>driver.quit()</pre><p>Using inspect element, we can see that the table we want to extract has the class “table-main js-tablebanner-t js-tablebanner-ntb”, and the primary columns of the table are Match, Result, 1, X,2, and date.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ftTUQAkgtRD1Fn4b0pBLNA.png" /><figcaption>Premier League page using inspect element</figcaption></figure><p>We will need to extract this table, create a DataFrame with those columns, and read the rows of the table to store them in the DataFrame. The easiest way to do this is to iterate over all the table&#39;s elements and store the element&#39;s text.</p><pre>table_matches = soup.find(&#39;table&#39;, attrs={&#39;class&#39;:&#39;table-main js-tablebanner-t js-tablebanner-ntb&#39;})<br>data = []<br>rows = table_matches.find_all(&#39;tr&#39;)<br>for row in rows:<br>    utils = []<br>    cols = row.find_all(&#39;td&#39;)<br>    for element in cols:<br>        utils.append(element.text)<br>    data.append(utils)<br><br>df = pd.DataFrame(data,columns=[&quot;Match&quot;,&quot;Result&quot;,&quot;1&quot;,&quot;X&quot;,&quot;2&quot;,&quot;Date&quot;])</pre><p>However, there are some issues with doing this. Some rows are empty, and the code cannot locate the odds as shown in the output.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/878/1*kXhLTHX6a0Fb1dBgoaoPiQ.png" /><figcaption>Output of plain code</figcaption></figure><p>Upon exploring the HTML code, we found that the odds are stored in an attribute called data-odd. The winning odd has a box, so we must access it as a span text. To achieve this, we will use a try-except block only in the cases of the odds. Finally, to avoid empty rows, we will only add them if they are not empty. The final code and output will appear as shown below.</p><pre>data=[]<br>for row in rows:<br>    utils = []<br>    cols = row.find_all(&#39;td&#39;)<br>    for element in cols:<br>        try:<br>            # Store the odds that win and didnt win<br>            if &#39;data-odd&#39; in element.attrs:<br>                utils.append(element[&#39;data-odd&#39;])<br>            else:<br>                utils.append(element.span.span.span[&#39;data-odd&#39;])<br>        except:<br>            # Store the text<br>            utils.append(element.text)<br>    if utils:<br>        data.append(utils)<br>df = pd.DataFrame(data,columns=[&quot;Match&quot;,&quot;Result&quot;,&quot;1&quot;,&quot;X&quot;,&quot;2&quot;,&quot;Date&quot;])<br>df.head()</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/848/1*dnz8QRAsZrHN_S492nXnfg.png" /><figcaption>Final Output</figcaption></figure><h4>Odds from different Bookmakers</h4><p>Next, we will demonstrate how to scrape odds data from all bookmakers for a specific match. This will enable you to compare the odds from different bookmakers and identify the best value for your bets. Let’s use the example of Brighton — Manchester Utd. In this case, the HTML content is dynamically generated using JavaScript, which implies that when we request the URL using requests, we are not getting the fully rendered page. As a result, we will use Selenium to obtain the full HTML code.</p><pre>URL = &quot;https://www.betexplorer.com/soccer/england/premier-league/brighton-manchester-united/6DkGCi44/&quot;<br><br>driver = webdriver.Chrome()<br>driver.get(URL)<br>html = driver.page_source<br>soup = BeautifulSoup(html, &#39;html.parser&#39;)<br>driver.quit()</pre><p>After acquiring the HTML code, the process is almost the same as before. The class of the table in this instance is “table-main h-mb15 sortable”. To shorten the code, we will use list comprehension and only add the row if it has four elements (Bookie, 1, X, 2).</p><pre>table = soup.find(&#39;table&#39;, attrs={&#39;class&#39;:&quot;table-main h-mb15 sortable&quot;})<br>rows = table.find_all(&#39;tr&#39;)<br><br>data = []<br>for row in rows:<br>    cols = row.find_all(&#39;td&#39;)<br>    bet = [element.text for element in cols if element.text!=&#39;&#39;]<br>    if len(bet)==4:<br>        data.append(bet)<br>bet_odds = pd.DataFrame(data,columns= [&#39;Bookmaker&#39;,&#39;1&#39;,&#39;X&#39;,&#39;2&#39;])<br>bet_odds</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/474/1*1h6IfwKadbDoyUkswzo6cQ.png" /><figcaption>First 5 rows of odds from bookies</figcaption></figure><p>Now that we have scraped and filtered the odds data, we can begin analyzing it. In future posts, we will demonstrate how to utilize the data to identify value bets and create a model for predicting match outcomes.</p><p>We have implemented this in soccer, but you may utilize this code in the sport of your preference. In our next post, we will show you how to build a machine-learning model to try to beat the bookmakers.</p><p>We will guide you through the process step-by-step, using Python libraries like Scikit-Learn. We will also offer tips and tricks for optimizing your model and achieving the highest possible accuracy.</p><p>Try out the code for yourself and see how it can help you gain an edge in sports betting.</p><p>Check out more articles about odds data and quantitative sports analysis:</p><p><a href="https://medium.com/systematic-sports">Systematic Sports</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=b77896286a4" width="1" height="1" alt=""><hr><p><a href="https://medium.com/systematic-sports/winning-at-sports-betting-scraping-and-analyzing-odds-data-with-python-b77896286a4">Winning at Sports Betting: Scraping and Analyzing Odds Data with Python</a> was originally published in <a href="https://medium.com/systematic-sports">Systematic Sports</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Bet Sizing and the Kelly Criterion]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://medium.com/systematic-sports/bet-sizing-and-the-kelly-criterion-4939fad24d45?source=rss----b6325b546080---4"><img src="https://cdn-images-1.medium.com/max/1478/1*RotvjokxwolvAenSuif1gQ.jpeg" width="1478"></a></p><p class="medium-feed-snippet">Bet Smarter, Not Bigger</p><p class="medium-feed-link"><a href="https://medium.com/systematic-sports/bet-sizing-and-the-kelly-criterion-4939fad24d45?source=rss----b6325b546080---4">Continue reading on Systematic Sports »</a></p></div>]]></description>
            <link>https://medium.com/systematic-sports/bet-sizing-and-the-kelly-criterion-4939fad24d45?source=rss----b6325b546080---4</link>
            <guid isPermaLink="false">https://medium.com/p/4939fad24d45</guid>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[sports-betting]]></category>
            <category><![CDATA[money]]></category>
            <category><![CDATA[science]]></category>
            <category><![CDATA[data-science]]></category>
            <dc:creator><![CDATA[Liam Hartley]]></dc:creator>
            <pubDate>Sun, 05 Nov 2023 16:45:44 GMT</pubDate>
            <atom:updated>2024-09-03T09:06:34.567Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[Backtesting a Sports Betting Strategy]]></title>
            <link>https://medium.com/systematic-sports/backtesting-a-sports-betting-strategy-283833a5eca3?source=rss----b6325b546080---4</link>
            <guid isPermaLink="false">https://medium.com/p/283833a5eca3</guid>
            <category><![CDATA[sports-betting]]></category>
            <category><![CDATA[programming]]></category>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[machine-learning]]></category>
            <dc:creator><![CDATA[Estèphe]]></dc:creator>
            <pubDate>Fri, 27 Oct 2023 13:17:08 GMT</pubDate>
            <atom:updated>2023-10-27T13:17:08.286Z</atom:updated>
            <content:encoded><![CDATA[<h4>Assessing the past to predict future profits</h4><p><strong>Hedge funds employ backtesting to trade billions of dollars. We’ve done it for the sports markets.</strong></p><p>So, how do you backtest a strategy? And how should you interpret the results?</p><h3>What is Backtesting?</h3><p>Backtesting is the process of testing a trading strategy using historical data. It helps <a href="https://www.investopedia.com/articles/active-trading/111214/quants-what-they-do-and-how-theyve-evolved.asp">quants</a> understand how a strategy would have performed in the past, allowing them to validate and refine their strategies before live trading.</p><figure><img alt="Ted Jiang in the big short, the quant" src="https://cdn-images-1.medium.com/max/1024/1*86JJREwIBk1NL20ELZzBFg.png" /><figcaption>Ted Jiang — the quant from The Big Short (<a href="https://youtu.be/QpsI_Gvn7C8?si=wZyoVdIa4bYExfaE&amp;t=104">source</a>)</figcaption></figure><p>We have built the data, infrastructure and software for backtesting at <a href="https://www.systematicsports.co.uk/index">Systematic Sports</a> to enable our quant research and improve our <a href="https://www.systematicsports.co.uk/betting-tips">betting strategies</a>.</p><h3>Historical Data</h3><p>First, ensure you have the data. At Systematic Sports, we have onboarded a decade of historical football data into our data lake, starting from the 2014/2015 season.</p><h3><strong>Building your Engine</strong></h3><p>A backtesting engine is a software that combines your historical data with trading logic to simulate how your strategy would have performed.</p><p>Systematic Sports’ backtesting engine works as follows:</p><h4>1. Historic strategy bets are passed to the engine</h4><figure><img alt="Fixture, Team, Market, Stake" src="https://cdn-images-1.medium.com/max/746/1*__wV6d1A8AeMFJlNUHzS_Q.png" /><figcaption>Schema of data passed into the engine.</figcaption></figure><p>These bets are calculated by passing our historical data to the betting algorithm to simulate the bets that would have been taken in the past. Additionally, a starting bankroll is specified (default $1000).</p><h4>2. Retrieve historical odds</h4><p>The engine maps these bets to the historical odds that were available then. Our engine simulates these odds by taking each game&#39;s best odds available at kick-off.</p><h4>3. Calculates each bet’s PnL</h4><p>The engine evaluates each bet’s outcome by referring to the fixture’s result. It then calculates the bet’s “Profit and Loss” or “PnL”, accounting for exchange trading costs if the betting venue is Betfair Exchange. Betfair Exchange charges a 5% commission on winnings; this significantly impacts your betting returns and can’t be ignored in a reliable backtest.</p><blockquote>Bet won: PnL = (Bookie Odd -1) * Stake * (1 - exchange comission rate)</blockquote><blockquote>Bet lost: PnL = — Stake</blockquote><blockquote>Bet void: PnL = 0</blockquote><h4>4. Generate a continuous time series of the strategy’s returns</h4><p>On each historic day of our backtest, each bet’s PnL is cumulatively summed to our current bankroll. The return can be plotted to view the strategy’s performance.</p><figure><img alt="A graph showing the backtesting performance of Systematic Sports betting strategy" src="https://cdn-images-1.medium.com/max/748/0*16dVwh_ytHXcHP8U" /><figcaption>The backtested performance of our flagship <a href="https://www.systematicsports.co.uk/betting-tips/value-model">Value Model</a>.</figcaption></figure><h4>5. Calculate performance metrics for the strategy</h4><p>Performance metrics evaluate a strategy’s key characteristics:</p><ul><li>Return</li><li>Risk</li></ul><p>Returns indicate the profit potential. Risk gauges the potential for losses and the strategy’s volatility. Investors make informed decisions by analysing both, balancing their desire for profit and risk tolerance.</p><figure><img alt="A collection of metrics for the backtesting strategy. Annualised returns, sharpe, volatility and max drawdown alongside returns." src="https://cdn-images-1.medium.com/max/761/0*ggLL3ijK7iPQUDUu" /><figcaption>Backtested performance metrics of our flagship <a href="https://www.systematicsports.co.uk/betting-tips/value-model">Value Model</a>.</figcaption></figure><p>We will explain these metrics in detail in a separate article.</p><h3><strong>Understanding your Backtest</strong></h3><p>A backtest helps you understand the strengths and weaknesses of your strategy.</p><p>Risk management is crucial to protecting your strategy from randomness and large drawdowns. A high volatility backtest indicates that your strategy takes on a lot of risk. Risk can be controlled by reducing the size of your stakes or the odds of bets taken.</p><p><a href="https://medium.com/systematic-sports/football-value-betting-guide-2023-24-382a534a36d2">Football Value Betting Guide 2023/24</a></p><p>The feedback from your backtest informs future improvements for your strategy. Systematic Sports models have been refined through backtests to:</p><ul><li>Control the minimum required probability of a bet.</li><li>Control exposure to single fixtures.</li><li>While still picking high-value bets.</li></ul><p>Our backtesting engine will continue to inform the research we do to improve our strategies.</p><h3><strong>Pitfalls of Backtesting</strong></h3><blockquote><em>“Theory will only take you so far” — Oppenheimer</em></blockquote><h4>1. Overfitting</h4><p>Imagine predicting someone’s outfit based on a week’s observation. On a rainy week, the subject wore a raincoat every day. You may conclude that they wear raincoats every day. If you base future predictions on this, you’d be wrong often.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/387/1*t0o_eqso3J9ZW7rIeEt54g.png" /><figcaption>Overfitting a dataset.</figcaption></figure><p>Overfitting happens when a model is trained too closely on a specific dataset. It’s like memorising the answers for a test instead of understanding the subject.</p><p>This is mitigated by ensuring a large dataset and varying market environments to backtest on (e.g. multiple leagues and seasons). The more parameters your strategy has, the easier it is to overfit. Limit the number of parameters and make sure each has a logical justification.</p><h4>2. Data Quality</h4><p>Garbage in, garbage out. Inaccurate or incomplete datasets can lead to misleading outcomes. Cleanse and validate your data before backtesting.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/287/1*HWEAqA4l6igVa2_IPkzG2w.png" /></figure><h4>3. Simulating Live Trading</h4><p>Remember, backtesting is a simulation. Even accounting for exchange costs, live trading will differ from the results simulated on a backtest, this is known as “slippage”.</p><figure><img alt="Thomas Shelby from the peaky blinders looking happy about backtesting before looking sad about live trading" src="https://cdn-images-1.medium.com/max/474/0*C0r6GQlLGcVFp19d" /></figure><p>Large bets have a tangible market impact, consuming liquidity and moving the price as the bet is placed. Backtests often assume idealised trade execution at the last known price, which can lead to overly optimistic results.</p><h4><strong>Summary</strong></h4><p><a href="https://www.systematicsports.co.uk/">Systematic Sports</a></p><p>Backtesting provides valuable insights about your strategy and should be used as one of the tools in your research and decision-making.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=283833a5eca3" width="1" height="1" alt=""><hr><p><a href="https://medium.com/systematic-sports/backtesting-a-sports-betting-strategy-283833a5eca3">Backtesting a Sports Betting Strategy</a> was originally published in <a href="https://medium.com/systematic-sports">Systematic Sports</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Python Betting Model for Six Football Leagues]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://medium.com/systematic-sports/python-betting-model-for-six-football-leagues-11b6ccd8cd54?source=rss----b6325b546080---4"><img src="https://cdn-images-1.medium.com/max/600/1*dCXnK4w0fcseAcROF5CtNQ.png" width="600"></a></p><p class="medium-feed-snippet">Using statistics, Pandas, BeautifulSoup and AWS to identify value bets</p><p class="medium-feed-link"><a href="https://medium.com/systematic-sports/python-betting-model-for-six-football-leagues-11b6ccd8cd54?source=rss----b6325b546080---4">Continue reading on Systematic Sports »</a></p></div>]]></description>
            <link>https://medium.com/systematic-sports/python-betting-model-for-six-football-leagues-11b6ccd8cd54?source=rss----b6325b546080---4</link>
            <guid isPermaLink="false">https://medium.com/p/11b6ccd8cd54</guid>
            <category><![CDATA[programming]]></category>
            <category><![CDATA[football]]></category>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[software-engineering]]></category>
            <dc:creator><![CDATA[Liam Hartley]]></dc:creator>
            <pubDate>Fri, 20 Oct 2023 08:54:16 GMT</pubDate>
            <atom:updated>2023-09-04T06:41:32.568Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[Football Value Betting Guide 2023/24]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://medium.com/systematic-sports/football-value-betting-guide-2023-24-382a534a36d2?source=rss----b6325b546080---4"><img src="https://cdn-images-1.medium.com/max/634/1*6s-owxJLad9Ag4WAr3b52w.jpeg" width="634"></a></p><p class="medium-feed-snippet">The Complete Guide to Value Betting (EV) in Football</p><p class="medium-feed-link"><a href="https://medium.com/systematic-sports/football-value-betting-guide-2023-24-382a534a36d2?source=rss----b6325b546080---4">Continue reading on Systematic Sports »</a></p></div>]]></description>
            <link>https://medium.com/systematic-sports/football-value-betting-guide-2023-24-382a534a36d2?source=rss----b6325b546080---4</link>
            <guid isPermaLink="false">https://medium.com/p/382a534a36d2</guid>
            <category><![CDATA[programming]]></category>
            <category><![CDATA[football]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[gambling]]></category>
            <category><![CDATA[statistics]]></category>
            <dc:creator><![CDATA[Liam Hartley]]></dc:creator>
            <pubDate>Mon, 04 Sep 2023 06:50:29 GMT</pubDate>
            <atom:updated>2025-11-26T22:02:06.470Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[Modern Football Statistics]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://medium.com/systematic-sports/football-statistically-mysterious-812f22b3f4a6?source=rss----b6325b546080---4"><img src="https://cdn-images-1.medium.com/max/1500/0*-Xv1ITHUyo_xmh5s.jpg" width="1500"></a></p><p class="medium-feed-snippet">Why does football remain difficult to quantify?</p><p class="medium-feed-link"><a href="https://medium.com/systematic-sports/football-statistically-mysterious-812f22b3f4a6?source=rss----b6325b546080---4">Continue reading on Systematic Sports »</a></p></div>]]></description>
            <link>https://medium.com/systematic-sports/football-statistically-mysterious-812f22b3f4a6?source=rss----b6325b546080---4</link>
            <guid isPermaLink="false">https://medium.com/p/812f22b3f4a6</guid>
            <category><![CDATA[liverpool]]></category>
            <category><![CDATA[mathematics]]></category>
            <category><![CDATA[football]]></category>
            <category><![CDATA[baseball]]></category>
            <category><![CDATA[statistics]]></category>
            <dc:creator><![CDATA[Liam Hartley]]></dc:creator>
            <pubDate>Thu, 16 Mar 2023 21:51:56 GMT</pubDate>
            <atom:updated>2023-03-16T21:54:08.586Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[Betting on Football With Python]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://medium.com/systematic-sports/betting-on-football-with-python-c32dd3c6a5a8?source=rss----b6325b546080---4"><img src="https://cdn-images-1.medium.com/max/630/1*6rN3Tjv47Jnyo6tQ8B_f2A.jpeg" width="630"></a></p><p class="medium-feed-snippet">Turning implied probabilities into betting strategies</p><p class="medium-feed-link"><a href="https://medium.com/systematic-sports/betting-on-football-with-python-c32dd3c6a5a8?source=rss----b6325b546080---4">Continue reading on Systematic Sports »</a></p></div>]]></description>
            <link>https://medium.com/systematic-sports/betting-on-football-with-python-c32dd3c6a5a8?source=rss----b6325b546080---4</link>
            <guid isPermaLink="false">https://medium.com/p/c32dd3c6a5a8</guid>
            <category><![CDATA[data]]></category>
            <category><![CDATA[python]]></category>
            <category><![CDATA[data-engineering]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[football]]></category>
            <dc:creator><![CDATA[Liam Hartley]]></dc:creator>
            <pubDate>Thu, 16 Mar 2023 21:48:50 GMT</pubDate>
            <atom:updated>2023-03-16T21:52:48.076Z</atom:updated>
        </item>
    </channel>
</rss>