<?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 JakeAllenData on Medium]]></title>
        <description><![CDATA[Stories by JakeAllenData on Medium]]></description>
        <link>https://medium.com/@allenjake440?source=rss-b69642b334cd------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*FRh43u7n5Tu4wfKeM0mUAw.jpeg</url>
            <title>Stories by JakeAllenData on Medium</title>
            <link>https://medium.com/@allenjake440?source=rss-b69642b334cd------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Fri, 20 Nov 2026 08:37:48 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@allenjake440/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[2025 NBA Champion Machine Learning Prediction]]></title>
            <link>https://allenjake440.medium.com/2025-nba-champion-machine-learning-prediction-8d9afb8db533?source=rss-b69642b334cd------2</link>
            <guid isPermaLink="false">https://medium.com/p/8d9afb8db533</guid>
            <dc:creator><![CDATA[JakeAllenData]]></dc:creator>
            <pubDate>Wed, 16 Apr 2025 01:05:06 GMT</pubDate>
            <atom:updated>2025-06-23T23:10:37.082Z</atom:updated>
            <content:encoded><![CDATA[<p>Using <strong>Machine Learning</strong> to predict the <strong>2025 NBA Champion</strong> right before the playoffs <strong>(April 2025).</strong></p><p>Shoutout to <a href="https://thejk.medium.com/"><strong><em>TheJk</em></strong></a><strong><em> </em></strong>he’s the real MVP.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/600/1*sjpxM_JLkJ7Mftj_zbYn0g.png" /><figcaption>Photo by Author and <a href="https://dribbble.com/shots/14301082-NBA-Finals-2020">Michael Meek</a> on the Dribble</figcaption></figure><h3>Introduction</h3><p>Here we are again. As the NBA season winds down and conference seedings lock into place, fans everywhere are buzzing with playoff predictions and bracket talk all trying to figure out who has what it takes to win it all and what teams can be a surprise.</p><p>But this isn’t going to be your typical emotionally driven take on the playoffs. That’s not what I’m entirely about. If you’ve seen my work before, you know I’m all about the numbers. I take massive amounts of data and extract meaningful insights giving us the ability to evaluate every team on an unbiased level. Because let’s be honest: emotions tend to sway people’s predictions. So, without further ado, let’s dive into this years prediction.</p><h3>For Full Details of the Predictions, Check out this Article!</h3><p><a href="https://allenjake440.medium.com/predicting-the-nba-champion-with-machine-learning-25e3a45a82f9">Predicting The NBA Champion With Machine Learning</a></p><h3>2025 NBA Champion Predictions</h3><p><em>*Personal Pick — Boston Celtics or Oklahoma City Thunder*</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*UC69oylt4nPeKq95VkvUoA.png" /><figcaption>2025 NBA Playoff Machine Learning Predictions</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*YY7MIMYZsNMSjCdns3E3Sw.png" /><figcaption>Top 8 Player According to My Custom Player Rating Metric</figcaption></figure><p>For the 2025 season, <strong>both of my models project the Boston Celtics as your NBA Champions once again</strong>, followed closely by <strong>surprisingly the Cleveland Cavaliers, </strong>and the <strong>historically great, but young Oklahoma City Thunder</strong>.</p><h3>Models and Track Record 📃</h3><p>For this year’s projections, I used two different models:</p><ol><li>The <strong>Ridge Regression model</strong> has successfully predicted the last five NBA champions in a row. Over the last 20 seasons, the champion has always been ranked within the top 6 predicted teams. In 80% of those years, the champion was in the top 3, with an average predicted champion rank of 2.10.</li><li>The <strong>XGBoost model</strong> also correctly predicted the last five champions. Like the Ridge, the title team has always been inside the top 7, though slightly fewer (75%) were in the top 3. Its average predicted champion rank is 2.30.</li></ol><p>Both models have correctly predicted the <strong>NBA Champion as the #1 team 50% of the time.</strong></p><h4>Top 3 Predicted Teams (Per Model)</h4><p><strong>Ridge Regression Model (Top 5 Predictions):</strong></p><ol><li>Boston Celtics — 71.9 🍀</li></ol><p>2. Cleveland Cavaliers — 61.4 ⚔️</p><p>3. Oklahoma City Thunder — 56.0 ⚡</p><p>4. Houston Rockets — 45.3 🚀</p><p>5. New York Knicks— 40.7 🗽</p><p><strong>XG-Boost Regression Model (Top 5 Predictions):</strong></p><ol><li>Boston Celtics — 72.1 🍀</li></ol><p>2. Cleveland Cavaliers — 59.4 ⚔️</p><p>3. Oklahoma City Thunder — 48.0 ⚡</p><p>4. Los Angeles Lakers — 47.6 🌊</p><p>5. Houston Rockets — 38.2 🚀</p><h3>My Take on the 2025 Season</h3><p>If I had to pick based on intuition and qualitative factors, <strong>my order would be: Boston, OKC, then Cleveland </strong>slightly different from what the models suggest. Here’s why:</p><h4>Why The Thunder? ⚡</h4><p>The biggest reason I consider OKC a top contender is their <strong>record-setting dominance</strong> this season. The 2025 Thunder not only boast a historically elite defense ranking first in nearly every major defensive category, but they also have this years MVP winner (most likely) in <strong>Shai Gilgeous-Alexander</strong>, backed by a deep, balanced, and dynamic roster.</p><p>Most notably, the Thunder posted an <strong>adjusted average margin of victory (SRS)</strong> of <strong>12.70</strong> — the <strong>highest in NBA history</strong>. That’s not just impressive, it’s <strong>unprecedented</strong>. Since 1950, no team has ever recorded a score that high.</p><p>And when you look back at every team in NBA history that’s managed an adjusted margin of <strong>10.7 or greater</strong>, here’s what you’ll find:</p><p><strong>2025 Thunder</strong> with 12.70: <strong>?</strong></p><p><strong>1971 Bucks</strong> - 11.92: <strong>Champion 🏆:</strong> Playoff Record:<strong> 12–2 (.857)</strong></p><p><strong>1996 Bulls</strong> - 11.80: <strong>Champion 🏆:</strong> Playoff Record: <strong>15–3 (.833)</strong></p><p><strong>1972 Lakers - </strong>11.65: <strong>Champion 🏆:</strong> Playoff Record: <strong>12–3 (.800)</strong></p><p><strong>2017 Warriors</strong> - 11.35: <strong>Champion 🏆:</strong> Playoff Record: <strong>16–1 (.941)</strong></p><p><strong>2024 Celtics</strong> - 10.75: <strong>Champion 🏆:</strong> Playoff Record: <strong>16–3 (.842)</strong></p><p>Every team on this list <strong>won the title</strong> and <strong>dominated the playoffs</strong>. The only concern with OKC is their youth. While this is a fair point, it’s worth noting that only one team in NBA history has ever won the title with a roster younger than this year’s <strong>Thunder</strong> (average age 24.7): the 1977 Trail Blazers (24.5). So yes, OKC would be one of the youngest championship teams ever. But in today’s NBA, <strong>youth doesn’t carry the same weight</strong> when a team is <strong>this dominant</strong> and already has <strong>a taste of playoff experience</strong>, in my opinion.</p><blockquote>These are the reasons I <strong>can’t bet against the Thunder.</strong></blockquote><h4>Why The Celtics? 🍀</h4><p>Boston are the defending champions. They’ve had a strong regular season and, when fully healthy (particularly with <strong>Kristaps Porzingis</strong> in the lineup), I believe they are the best team in the NBA. They combine the league’s best starting lineup with solid depth and elite two-way players.</p><p>The Celtics check all the championship boxes. They rank first in total <strong>playoff experience</strong> with <strong>764 past combined playoff games</strong>, best road record in the NBA, third in adjusted point differential, second in offensive efficiency, and fifth in defense. With proven stars in <strong>Jayson Tatum</strong> and<strong> Jaylen Brown,</strong> Boston has once again cruised to another 60 win season for the second year in a row.</p><blockquote>A healthy Celtics squad is still my official pick.</blockquote><h4>Why The Cavs? ⚔️</h4><p>Cleveland and star guard Donovan Mitchell have been here before. Many people may not think this Cavs team is quite ready, but in my opinion, they are. They’ve had recent playoff success, and while Mitchell has never made it to the conference finals, he’s consistently led highly successful regular-season teams with strong statistical profiles. Pair that with the fact that the Cavs have the <strong>most potent offense</strong> in the NBA this season leading in both <strong>shooting efficiency </strong>and <strong>possession points per game</strong> and you’re looking at a real threat. Add in the presence of a potential Defensive Player of the Year in Evan Mobley, and this team becomes even more dangerous. The Cavs have a great mix of talent, but they are missing some proven star power. <strong>Donovan Mitchell</strong> and <strong>Darius Garland</strong> haven’t consistently been dominant forces in past seasons, and they haven’t risen on many MVP ladders in recent years, which raises a <strong>legitimate concern</strong> for me. Still, this is a team worth watching. You can’t completely discredit their potential.</p><h4>Let Me Talk About The Lakers Now… 🌊</h4><p>Now, one team in particular is the Los Angeles Lakers. First off, I still don’t understand how they managed to land Luka and honestly, I don’t think we’ll ever truly know. Maybe it had something to do with the Mavericks’ ownership situation, but regardless, they got him. And <strong>Luka has absolutely elevated this team</strong>, helping the Lakers secure a top three seed in the NBA. They now arguably have two of the greatest players ever in<strong> LeBron</strong> and <strong>Luka</strong>, making them an absolute nightmare matchup for any undersized opponent.</p><p>That said, I do have concerns. Despite their elite star power, strong record, and high seeding, the <strong>Lakers haven’t exactly been dominant</strong>. They’ve lost to some bad teams, they’re under .500 on the road, and while they’ve held their own against tougher opponents, they’ve looked inconsistent at times. I’ve always believed in their head coach (JJ Redick), but he himself is still a rookie, and their <strong>SRS ranks just 14th in the league which doesn’t scream championship caliber dominance.</strong></p><h4>Conference Landscape</h4><p>The Western Conference remains very competitive, but the <strong>Thunder</strong> clearly stand out above the rest. Meanwhile, <strong>Cleveland</strong> and <strong>Boston </strong>have dominated the Eastern Conference. All three teams have secured over 60 wins something we haven’t seen since 2009, when the Lakers, Cavaliers, and Celtics all did it. That year, the Lakers won the Finals, the Cavs reached the Eastern Conference Finals, and the Celtics reached the Semis.</p><p>I put this together pretty quickly out of boredom and curiosity, here’s a <strong>Matchup Matrix</strong> of all the 2025 NBA playoff teams. Here’s how it works: for example, if you look at the Cavaliers in the top-left corner, you’ll see they went <strong>1–3 against the Hawks</strong> and <strong>2–1 against the Pacers</strong> during the regular season. This matrix displays each playoff team’s <strong>losing matchup against other playoff teams</strong>, offering a quick way to see how they’ve fared head-to-head.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*AmEGT3vtgH5CgzIzL_X4dA.png" /><figcaption>2025 Regular Season Matchup Matrix</figcaption></figure><p>For exact information on these head-to-head matchups <a href="https://www.basketball-reference.com/leagues/NBA_2025_standings.html#all_expanded_standings"><strong>click here.</strong></a></p><h3>Champion Probability Tiers ⚖️</h3><p>When organizing teams based on championship probability, I use tiers that go beyond stats these buckets reflect style, experience, and storyline too.</p><h4>True Contenders — Top Shelf 🪙</h4><p>Boston Celtics, Oklahoma City Thunder &amp; Cleveland Cavaliers</p><h4>Secondary Contenders — Bottom Shelf 🪨</h4><p>New York Knicks &amp; Houston Rockets</p><h4>2011 Dallas Mavericks — Like Teams (Dark Horse) 🐎</h4><p>Los Angeles Lakers</p><h4>1995 Houston Rockets — Like Teams (How Did They Do It?) 🔮</h4><p>Golden State Warriors &amp; Denver Nuggets</p><h4>Bracket Busters— (Not Contenders, but Dangerous) 💀</h4><p>Minnesota Timberwolves, Los Angeles Clippers, Indiana Pacers</p><h3>Playoff Forecast &amp; Analytics 📈</h3><p>Lastly, here is my Playoff Round-by-Round Forecast, along with the Feature Importance Chart generated from the Ridge Regression model. This chart highlights how each playoff team’s raw statistics align with the stats that are influentially in predicting playoff wins (according to the ridge).</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*fQwmiZqxF_-aSepykeQFeg.png" /><figcaption>*6/23/25 Edit* 2025 Playoffs Round by Round Forecast | Pre Playoffs</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*kMkkXlXdwkO56ShjV2hAuQ.png" /><figcaption>Most Important Stats From The Prediction Model</figcaption></figure><h3>Feedback</h3><p>Any questions or feedback are always welcome, so feel free to post them. <strong>For full details</strong> of all steps done for this project check out my <a href="https://github.com/allenjake440/NBA_Champion"><strong>GitHub.</strong></a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=8d9afb8db533" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[2025 March Madness Machine Learning Prediction]]></title>
            <link>https://allenjake440.medium.com/2025-march-madness-machine-learning-prediction-e5b3a66b2c59?source=rss-b69642b334cd------2</link>
            <guid isPermaLink="false">https://medium.com/p/e5b3a66b2c59</guid>
            <category><![CDATA[march-madness]]></category>
            <dc:creator><![CDATA[JakeAllenData]]></dc:creator>
            <pubDate>Tue, 18 Mar 2025 01:44:26 GMT</pubDate>
            <atom:updated>2025-04-08T19:58:17.138Z</atom:updated>
            <content:encoded><![CDATA[<p>Using <strong>Machine Learning</strong> in preparation to predict the <strong>2025 Men’s NCAA March Madness Tournament Champion</strong> right before the playoffs <strong>(March 17, 2025).</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/600/1*s52T-vwnUCh9wusG70-EvA.png" /><figcaption>Photo by Author and <a href="https://dribbble.com/shots/4226205-March-Madness-Basketball">Stephen Piercy</a> on the Dribble</figcaption></figure><p>Well, it’s that time again… As the <strong>Men’s NCAA basketball season winds down</strong>, conference champions have been crowned, the committee has seeded the bracket, and so its time for <strong>March Madness </strong>to begin.</p><p>For those of you who are new, every year for the <strong>NFL, Men’s NCAA Division 1 March Madness Tournament, and NBA</strong>, I use <strong>detailed machine learning models</strong> to analyze teams, predict potential winners, and break down the tournament landscape. My approach relies purely on <strong>data and numbers</strong>, uncovering insights that traditional analysis might miss.</p><p>If there’s one thing I’ve learned from sports predictions, it’s that <strong>emotions can mislead us about a team</strong>. By leveraging massive amounts of <strong>data and machine learning</strong>, we can evaluate teams on a <strong>deeper, unbiased level </strong>free from gut feelings and personal biases.</p><p>Without further ado, here are my machine learning model predictions for the 2025 March Madness Tournament!</p><p>By the way, my predictions for last year’s tournament turned out <strong>great, </strong>it had <strong>UConn beating Purdue in the finals!</strong> Hopefully the results stay that way this year.</p><h3>For Full Details of the Predictions, Check out this Article!</h3><p><a href="https://allenjake440.medium.com/predicting-the-mens-march-madness-champion-with-machine-learning-892bd78997ca">Predicting The Men’s March Madness Champion With Machine Learning</a></p><h3>2025 Men’s NCAA March Madness Champion Predictions</h3><p><em>*Personal Pick — Florida Gators and Duke Blue Devils*</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*DxKpX7QiNVo9UinnFI9hIQ.png" /><figcaption>2025 NCAA Men’s MM Tour. Machine Learning Predictions</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*aNX9rL-hACm2xAvMfrtP6A.png" /><figcaption>2025 NCAA Men’s MM Tour. Machine Learning Predictions | Both Models Average</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*rDadBus2MQDJUzfRQmCvdg.png" /><figcaption>The Best Players in College Basketball (3/17/25)</figcaption></figure><p>This season, my primary model projects the <strong>Florida Gators</strong> as the <strong>National Champions</strong>, alongside the <strong>Duke Blue Devils</strong> as a top contender. If you’re interested in the context behind these predictions and the broader landscape of this season, stick around.</p><h4>Models and Track Record</h4><p>For this year’s projections, I used two different models:</p><ol><li><strong>Ridge Regression Model: </strong>This model has successfully predicted the last four tournament champions (primary model).</li><li><strong>Random Forest Regression Model</strong> : Over the past ten seasons, this model has <strong>always</strong> placed the eventual champion within its <strong>top 10 predictions</strong> (for example, it ranked the <strong>2014 UConn Huskies at #9</strong>, a historically unpredictable winner).</li></ol><p>Here are the results for the <strong>top two teams from each model</strong>:</p><p><strong>Ridge Regression Model (Top 2 Predictions):</strong></p><ul><li>Florida <strong>65.3%</strong></li><li>Duke <strong>62.6%</strong></li></ul><p><strong>Random Forest Model (Top 2 Predictions):</strong></p><ul><li>Auburn <strong>63.0%</strong></li><li>Florida <strong>62.1%</strong></li></ul><p>If you’re looking at the predictions, you’ll notice some teams have <strong>red numbers</strong> next to them this represents their pre-injury probability. Before running these projections, I researched key player injuries and adjusted individual player ratings accordingly.</p><h4>Duke Is Loaded…</h4><p>The <strong>Duke Blue Devils</strong> are absolutely <strong>stacked</strong> this year. If <strong>Cooper Flagg</strong> were fully healthy, they might have been my top pick. Flagg boasts the <strong>4th highest player rating in my database (since 2003)</strong> he’s the real deal. Additionally, Duke is playing in the <strong>East Region</strong>, which could play a factor in their path to the title since it is the easiest region of the bracket.</p><h4>The SEC Dominance in 2025</h4><p>This season, the <strong>SEC is historically strong</strong>, boasting the <strong>highest conference rating in my database (since 2003).</strong> With <strong>Auburn and Florida as #1 seeds</strong> and <strong>14 SEC teams</strong> making the tournament, the strength of schedule could either help or hurt these teams in March Madness. Playing in the toughest conference sharpens teams for deep runs, or could either wear you out.</p><h4>Bracket Region Strength Ratings</h4><ul><li><strong>South: </strong>13.27</li><li><strong>Midwest:</strong> 11.06</li><li><strong>West:</strong> 10.78</li><li><strong>East:</strong> 9.69</li></ul><h4>Top 5 Toughest Conferences This Season</h4><ol><li><strong>SEC</strong>: 5.15</li><li><strong>Big East: </strong>2.19</li><li><strong>A-Sun: </strong>1.58</li><li><strong>Big 12: </strong>1.33</li><li><strong>Big 10: </strong>1.07</li></ol><p>Lastly, here is <strong>my tournament forecast</strong> and the <strong>feature importance chart</strong> generated from the <strong>Ridge Regression model</strong>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/901/1*k7hPF1SyXfjpU_1GZ_9iKQ.png" /><figcaption>Regression Forecast Probabilities | *Post Tourney Update 4/8/25*</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*fzPFYgrn0gxVIPUnYpghrA.png" /><figcaption>My GitHub | Interactive Power BI Dashboard | 2025 Heat Table</figcaption></figure><h3>Feedback</h3><p>Any questions or feedback are always welcome, so feel free to post them. <strong>For full details</strong> of all steps done for this project check out my <a href="https://github.com/allenjake440/Mens_March_Madness_Champion"><strong>GitHub.</strong></a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e5b3a66b2c59" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[2025 NFL Champion Machine Learning Prediction]]></title>
            <link>https://allenjake440.medium.com/2025-nfl-champion-machine-learning-prediction-4fc82113cee2?source=rss-b69642b334cd------2</link>
            <guid isPermaLink="false">https://medium.com/p/4fc82113cee2</guid>
            <dc:creator><![CDATA[JakeAllenData]]></dc:creator>
            <pubDate>Fri, 10 Jan 2025 22:01:21 GMT</pubDate>
            <atom:updated>2025-02-11T03:56:15.653Z</atom:updated>
            <content:encoded><![CDATA[<p>Using <strong>Machine Learning</strong> in preparation to predict the <strong>2024–25 NFL Champion</strong> right before the playoffs <strong>(January 2025).</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/600/1*uVrRAfX2pce6fBCuV3vp-A.png" /><figcaption>Photo by Author and <a href="https://dribbble.com/shots/9328989-Lombardi/attachments/1369972?mode=media">Zilligen Design Studio</a> on the Dribble</figcaption></figure><h3>Introduction</h3><p>The 2024–25 NFL regular season has come to close and the playoffs are under way. As per usually many fans love to fill out playoff prediction brackets using nothing but their own intuition, feeling, and emotions to decide what teams move on and what teams drop out. This prediction purely the opposite of that, the predictions presented in this article are fully based on just the numbers… and well my opinions too, so not really.</p><h3>Full YouTube Prediction Video!</h3><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FsiFcXUAwb3g%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DsiFcXUAwb3g&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FsiFcXUAwb3g%2Fhqdefault.jpg&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/4a55f49b811b19f27b77a53252a9e1f4/href">https://medium.com/media/4a55f49b811b19f27b77a53252a9e1f4/href</a></iframe><h3>2024–25 NFL Champion Prediction!</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*YpYi5r5TwRJ-mSjUI9fvJw.png" /><figcaption>2024–25 Forecast Table Round by Round</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*h0MSbfBgLAy5dnEngqTB4g.png" /><figcaption>2024–25 NFL Champion Projections</figcaption></figure><p>Well all of I got to this say is this one is going to be a tight one. Literally less half a decimal of a playoff win separates the top 6 favored teams by the model. And according previous machine learning forecast that backtested these predictions, you safe champion pick should be between the <strong>Ravens, Bills, </strong>and <strong>Vikings!</strong></p><p>Based on my overall research out of these playoff teams, honestly the <strong>Ravens🐦‍⬛</strong>, <strong>Bills 🦬</strong>, <strong>Lions 🦁</strong>, and <strong>Chiefs 🏹</strong>are/should be the Super Bowl <strong>Winner this Year, </strong>well one of the them of course .</p><p>Well to narrow that down some more, I believe the <strong>Kansas City Chiefs </strong>or the <strong>Baltimore Ravens </strong>will be the Champion 🏆.</p><p>A dark horse team I can see making noise this post-season is well, the <strong>Los Angeles Chargers </strong>and the <strong>Minnesota Vikings</strong>. Now beating the Chiefs is a tall task, but honestly they could. The Chargers have great defense, a experienced head coach, and a talented quarterback despite his lack of playoff success. The <strong>Vikings </strong>well are actually really good nearly were the 1 seed in the NFC, I believe the <strong>Vikings</strong> could beat <strong>Detroit</strong> and <strong>Philadelphia </strong>to be in the Super-Bowl.</p><p>Something else to mention is for this NFL season I’ve changed my modeling approach from last season. Instead using seeding to influence predictions (mostly), I’ve created a <a href="https://github.com/allenjake440/NFL_Champion/blob/main/custom_team_qb_stat.md"><strong>team rating metric</strong></a> that in short bakes in regular season success (wins) and the teams star power into one historical metric.</p><p>For all of my thoughts and more please check out the video above and for full details surrounding this project please check out <a href="https://github.com/allenjake440/NFL_Champion"><strong>GitHub.</strong></a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=4fc82113cee2" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Using Machine Learning To Find Every Missing DPOY in NBA History (1982–1947)]]></title>
            <link>https://allenjake440.medium.com/using-machine-learning-to-find-every-missing-dpoy-in-nba-history-1982-1947-0aa1e42da0fe?source=rss-b69642b334cd------2</link>
            <guid isPermaLink="false">https://medium.com/p/0aa1e42da0fe</guid>
            <dc:creator><![CDATA[JakeAllenData]]></dc:creator>
            <pubDate>Tue, 23 Jul 2024 02:12:01 GMT</pubDate>
            <atom:updated>2024-10-25T20:37:39.557Z</atom:updated>
            <content:encoded><![CDATA[<p>Using <strong>NBA data</strong> and <strong>Machine Learning</strong> to find out what <strong>players “would’ve” won</strong> the <strong>award before its origination in 1983.</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*pRVpRSzDnqe6M8KZi2pGvQ.png" /><figcaption>Photo by: Author</figcaption></figure><h3>Introduction</h3><p>The NBA Defensvie Player of the Year award (DPOY) orginiated in 1983 and was to be given to the “best” defensive player at the end of each NBA season. As a basketball fan I’ve always heard the noise surrounding the legendary Bill Russell and his defensive prowess, but today I wanted to take it to the test and find out how many DPOY’s he would’ve won if the award was around back then (1947).</p><p><strong><em>Well… I also just wanted to answer this question in general:</em></strong></p><blockquote><em>“What players would’ve won this award 🏆if it was already introduced when the NBA orginiated in 1947?”</em></blockquote><h3>Data</h3><p>All the data used for this project comes from <a href="https://www.basketball-reference.com/players/">basketball reference.</a> I used several data sets from <strong><em>Sumitro Datta | NBA Stats (1947-present) </em></strong>kaggle hub to piece together the data and build a dataset to use for analysis.</p><p><em>For full details feel free to investiage the code: </em><a href="https://github.com/allenjake440/NBA_Missing_DPOY"><strong>My GitHub</strong></a></p><h3>Model Prediction Value</h3><p>The value to be calculated by the model is award_dpoy_share. This statistic is the number of votes a player obtained divided by the maximum number of votes possible. For example, Rudy Gobert achieved 433 points out of a maximum of 495 points (votes), giving him a share of 0.875, the largest of the competitors, meaning he wins the award.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/613/1*aJCcYlrOaF8CGSaMOsvpvg.png" /><figcaption>Table Calculation of ‘award_dpoy_share’</figcaption></figure><h3>The Missing Data</h3><p>These two charts visualize the NaN’s “blanks” in each feature these ranging from two very different time period (1950–1980 &amp; 1980–2024). As you can see, many advanced statistics (vorp, bpm, dbpm, obpm, etc.) were not fully tracked until around 1974. The same goes for other advanced stats like basic percentage stats (orb, drb, ast, etc.).</p><p>With that said, I did include a couple of these advanced features in the model process because those features were still around for about 11 missing DPOY seasons.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*4fNqPnAa9gS7sb84M-1jZg.png" /><figcaption>NaN Counts (1950–1980)</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*-UJRM9CUMo2GV6KLM3Jp1A.png" /><figcaption>NaN Counts (1980–2024)</figcaption></figure><h4>What Happen Here?</h4><p>Something interesting I found is that there has only been 4 DPOYs ever, who won the award despite being listed on the 2nd team All-Defense for that season. This is very strange since 91% of the time the DPOY makes 1st team All-Defense. Can someone explain to me what happened here, in those seasons?</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/209/1*cTnLCnzeDGtAJJYlcbcidA.png" /><figcaption>DPOYS, not All-Defense 1st Team</figcaption></figure><h4>Analysis Notes</h4><ul><li>82% of DPOYs are top 8 in the league in (<em>dws</em>) defensive win shares.</li><li>100% of DPOYs are All-Defensive team (<em>All-Defense</em>), obviously.</li><li>52% of DPOYs make an All-NBA team selection (<em>All-NBA</em>).</li><li>91% of DPOYs are 1st team All-Defense (<em>All-Defense_1st)</em>, obviously.</li></ul><h3>Machine Learning Results</h3><p>I chose two different model archetypes for this experiment because, first I wanted to understand the key differences between the models, and secondly, I wanted to determine which is better in general for a problem like this.</p><blockquote>Logistic Regression (Classification) v. Random-Forest (Regression)</blockquote><ul><li>Logistic Regression -&gt; Target -&gt; ‘won_dpoy’ (The single DPOY winner).</li><li>RF Regresstion → Target -&gt; ‘award_dpoy_share’ (The votes of DPOY).</li><li>Logistic Regression -&gt; Error Metrics -&gt; Hit-Rate/BackTesting.</li><li>RF Regresstion → Error Metrics -&gt; AP/NDCG/BackTesting.</li></ul><blockquote>Which Model Performed Better?</blockquote><ul><li>Winner: <strong><em>Random-Forest Regression</em></strong></li></ul><blockquote>Specific Model Results</blockquote><h4><strong>Logistic</strong>: 3/5 (2024 – 2020) Current DPOYs were correctly predicted.</h4><ul><li>It labeled the 2020 winner, Giannis Antetokounmpo, as <strong>ranked 3rd</strong> in the projection.</li><li>It labeled the 2022 winner, Marcus Smart, as <strong>ranked 4th </strong>in the projection.</li><li>It didn’t get the exact voting (contenders) in the best order, compared to Random-Forest.</li></ul><h4>RF: 3/5 (2024–2020) Current DPOYs were correctly predicted.</h4><ul><li>It labeled the 2020 winner, Giannis Antetokounmpo, <strong>as ranked 3rd</strong> in the projection.</li><li>It labeled the 2022 winner, Marcus Smart, as <strong>ranked 3rd</strong> in the projection.</li><li>Random-Forest not only had better DPOY predictions, but it also assessed the other contenders for each season better as well.</li></ul><h3>Model Decision Making</h3><p>To understand the complex decision-making of the Random-Forest Regressor, here’s the general feature importance of the model.</p><ul><li>Players with large <em>All-Defense_1st_share</em> values contribute positively and massively towards the prediction.</li><li>Players with large values in blocks (<em>blk_percent/blk_per_game</em>) also contribute well towards the prediction.</li><li>And so on…</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/999/1*gyZH1wZPnKX5a4mCEjntwA.png" /><figcaption>Feature Importance | Random-Forest Regression</figcaption></figure><h3>The Missing DPOY’s Results🤨 | Plus My Analysis</h3><p>True Contender Definition:</p><blockquote>True Contenders = When model was tested on current day DPOY’s (2024–2020) it always had the correct DPOY winner within the top 3 predicted players.</blockquote><p>Here is a table displaying the true contenders with the <strong>true probability</strong> of winning the <strong>DPOY each season (1982–1947)</strong>. As you can see, there are some seasons in which the model has major difficulty in assessing due to a lack of relevant data (i.e. 1952).</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/959/1*-kKEDELut67Xc58y6Q9i_w.png" /><figcaption>Random-Forest | Missing DPOY | Projections</figcaption></figure><h4>Takeaways</h4><p>Here is a list, <strong>first</strong> sorted by which unique players accumulated the most percentage chance to recieve each award over the missing seasons. And <strong>secondly </strong>showing the counts of that award that unique player would’ve won according to the model.</p><ul><li>1st: Bill Russell | 4.39 | <strong>5 Awards</strong></li><li>2nd: Wilt Chamberlain | 4.16 | <strong>6 Awards</strong></li><li>3rd: George Mikan | 3.66 | <strong>6 Awards</strong></li><li>4th: Bobby Jones | 2.78 | <strong>3 Awards</strong></li><li>5th: Dave DeBusschere | 2.55 | <strong>4 Awards</strong></li><li>6th: Dennis Johnson | 1.89 | <strong>3 Awards</strong></li><li>7th: Kareem Abdul-Jabbar | 1.66 | <strong>2 Awards</strong></li><li>8th: Walt Frazier | 1.64 | <strong>2 Awards</strong></li><li>9th: Dolph Schayes | 1.37 | <strong>1 Award</strong></li><li>10th: Maurice Stokes | 1.22 | <strong>2 Awards</strong></li></ul><h3>The Missing DPOY’s… According to My Analysis 🙄</h3><p>Now here is the same table again, but this time highlighting yellow to show my picks based on my analysis (Power-BI Dashboard and Google Searches).</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/881/1*ST4aFkwsEb4ijPHblWCgDA.png" /><figcaption>My DPOY Picks</figcaption></figure><p>According to my personally researched-based picks, here are the total award counts for each unique player throughout the missing seasons:</p><ul><li>1st: Bill Russell | <strong>8 Awards</strong></li><li>2nd: George Mikan | <strong>5 Awards</strong></li><li>3rd: Chamberlain, Frazier, and Abdul-Jabbar| <strong>3 Awards</strong></li><li>4th: Schayes, Stokes, DeBusschere, and Johnson | <strong>2 Awards</strong></li><li>5th: Miasek, Dallmar, and Sloan | <strong>1 Award</strong></li></ul><h3>Conclusion</h3><p>Overall this project was quite challenging to say the least. The lack of data from 1968–1947 became more apparent as the predictions rolled into the past, for instance in 1947 you can see both John Logan and Stan Miasek have the exact same probability to win the award 🤨simply because of the few statistics available back then. To conclude realistically more data is needed to be present back then in order to truly make a more accurate prediction.</p><blockquote>“With that, said I want to know what you guys think about all of this. Who do you think are the missing dpoys, let me know in the comments.”</blockquote><h3>Feedback</h3><p>Any questions or feedback are always welcome, so feel free to post them. <strong>For full details</strong> of all steps done for this project check out my <a href="https://github.com/allenjake440/NBA_Missing_DPOY"><strong>GitHub.</strong></a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=0aa1e42da0fe" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[2024 NBA Champion Machine Learning Prediction]]></title>
            <link>https://allenjake440.medium.com/2024-nba-champion-machine-learning-prediction-6bea90855891?source=rss-b69642b334cd------2</link>
            <guid isPermaLink="false">https://medium.com/p/6bea90855891</guid>
            <category><![CDATA[nba-championship]]></category>
            <category><![CDATA[nba-playoffs]]></category>
            <dc:creator><![CDATA[JakeAllenData]]></dc:creator>
            <pubDate>Wed, 17 Apr 2024 01:23:12 GMT</pubDate>
            <atom:updated>2024-06-18T13:02:42.811Z</atom:updated>
            <content:encoded><![CDATA[<p>Using <strong>Machine Learning</strong> to predict the <strong>2024 NBA Champion</strong> right before the playoffs <strong>(April 2024).</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/600/1*9Mr4fDoJtxf7G8d3tzHj1A.png" /><figcaption>Photo by Author and <a href="https://dribbble.com/shots/14301082-NBA-Finals-2020">Michael Meek</a> on the Dribble</figcaption></figure><h3>Introduction</h3><p>I am excited! As the NBA season whines down, the conference seedings are locked up and now fans are being to swirl around playoff brackets and picks on who they think can win it all.</p><p>Now just like I always say emotions can often times blind us from the truth. By using mass amounts of past NBA data we can evaluate every team on a complete and unbiased level allowing us to see cracks in teams that we never have noticed before. Without any further hold up, here are my 2024 NBA machine learning playoff predictions!</p><h3>For Full Details of the Predictions, Check out this Artice!</h3><p><a href="https://allenjake440.medium.com/predicting-the-nba-champion-with-machine-learning-25e3a45a82f9">Predicting The NBA Champion With Machine Learning</a></p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FVwjmGrNQPzg%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DVwjmGrNQPzg&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FVwjmGrNQPzg%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/0dde36812bdd453f43f441fb130a4553/href">https://medium.com/media/0dde36812bdd453f43f441fb130a4553/href</a></iframe><h3>2024 NBA Champion Predictions</h3><p><em>*Personal Pick —Denver Nuggets*</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*FY9SbTveAjuLtXHHxw6E7A.png" /><figcaption>2024 NBA Playoff Machine Learning Predictions</figcaption></figure><p>Just like anyone expected the <strong>Denver Nuggets and Boston Celtics</strong> top the charts as basically the co-title favorites. What I personally find surprising is my model values the <strong>Nuggets, Bucks,</strong> and <strong>TimberWolves</strong> much more so than the odds makers. Something else to note is the odds makers give the <strong>Los Angeles Clippers </strong>the 3rd best chances to win the title, while my forecast believes they are not even close.</p><p>To summarize my research on this season, I will be very surprised if the <strong>Denver Nuggets </strong>or the <strong>Boston Celtics</strong> do not win the NBA title this season. The western conference is wild and very competitive which could be very tough on the<strong> Nuggets</strong>. The <strong>Thunder </strong>and <strong>TimberWolves</strong> have had a winning record against the <strong>Nuggets </strong>this season. The <strong>Dallas Mavericks </strong>are by far the hottest team in the league right now and even the <strong>Phoenix Suns</strong> have shown to have a bite lately.</p><p>The eastern conference on the other hand, as been completely controlled by the <strong>Boston Celtics</strong>. The only other team I can see beating the <strong>Celtics </strong>in the east is the slept on <strong>Milwaukee Bucks</strong>. From many coaching changes and adjustments many fans have completely abandoned the Bucks this season as championship contenders. Despite the concerns the numbers and myself believe the <strong>Buck’s </strong>are still a contender. In 2 out of the 4 models the <strong>Bucks </strong>actually appeared in 3rd place easily locking them into contention.</p><p>The historical results of the model ensures an<strong> 90% chance</strong> that the NBA champion is with-in the <strong>top 3 predicted teams</strong>, with the ridge regression in particular ensuring a <strong>55.2% chance</strong> of champion being the <strong>1st place</strong> predicted team <strong>(Denver Nuggets)</strong>.</p><p>The exact results of these predictions were from a combination of 4 machine learning models: Random Forest, XGBoost, Ridge, and Neural Network. In terms of model performance (top-to-bottom | best-to-worst) and their top NBA champion prediction are as follows:</p><p><em>Random-Forest (Top Pred.) | 1990-&gt; Data | </em><strong><em>Boston Celtics</em></strong></p><p><em>Ridge Linear (Top Pred.) | 1950-&gt; Data | </em><strong><em>Denver Nuggets</em></strong></p><p><em>Neural Network (Top Pred.) | 1950-&gt; Data | </em><strong><em>Denver Nuggets</em></strong></p><p><em>XGBoost Model (Top Pred.) | 1970 -&gt; Data | </em><strong><em>Denver Nuggets</em></strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*5tD3-Y3i0_rD6duPZkeeBA.png" /><figcaption>*Post-Tourney Update* | Regression Forecast Probabilities</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*kWkZX9c5uwRymbX4HvHKBQ.png" /><figcaption>My GitHub | Interactive Power BI Dashboard | 2024 Heat Table</figcaption></figure><h3>Feedback</h3><p>Any questions or feedback are always welcome, so feel free to post them. <strong>For full details</strong> of all steps done for this project check out my <a href="https://github.com/allenjake440/Mens_March_Madness_Champion"><strong>GitHub.</strong></a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=6bea90855891" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[2024 March Madness Machine Learning Prediction]]></title>
            <link>https://allenjake440.medium.com/2024-march-madness-machine-learning-prediction-081d6688ea71?source=rss-b69642b334cd------2</link>
            <guid isPermaLink="false">https://medium.com/p/081d6688ea71</guid>
            <dc:creator><![CDATA[JakeAllenData]]></dc:creator>
            <pubDate>Mon, 18 Mar 2024 20:53:08 GMT</pubDate>
            <atom:updated>2026-07-28T03:16:06.076Z</atom:updated>
            <content:encoded><![CDATA[<p>Using <strong>Machine Learning</strong> in preparation to predict the <strong>2024 Men’s NCAA March Madness Tournament Champion</strong> right before the playoffs <strong>(March 17, 2024).</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/600/1*KQBP6OpkltJpdx0y7iQTjQ.png" /><figcaption>Photo by Author and <a href="https://dribbble.com/shots/4226205-March-Madness-Basketball">Stephen Piercy</a> on the Dribble</figcaption></figure><h3>Introduction</h3><p>As the men’s NCAA basketball season whines down, the conference champions are in the books, the committe seeds the bracket,all setting up the craziest and one the most exciting sport spectacles in the United States.</p><p>As many people know, this tournament is all about the brackets, predicting upsets, and finding out who will win this crazy tourney. Many analysts and die-hard fans make reasons why this or that will win it all, but if there is anything I’ve learned from sports it’s that emtions can blind you. So, by assessing every team by using purley numbers can allow us to analyze teams on deep and unbiased level. Without further a-do here are my machine learning model predictions for this 2024 March Madness tourney.</p><h3>For Full Details of the Predictions, Check out this Article!</h3><p><a href="https://allenjake440.medium.com/predicting-the-mens-march-madness-champion-with-machine-learning-892bd78997ca">Predicting The Men’s March Madness Champion With Machine Learning</a></p><h3>2024 Men’s NCAA March Madness Champion Predictions</h3><p><em>*Personal Pick — Arizona Wildcats*</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*lWdz2Iwyzp2tOk5_Db_gMg.png" /><figcaption>2024 NCAA Men’s MM Tour. Machine Learning Predictions</figcaption></figure><p>To no surprise the algorithm, thinks the defending champion <strong>Uconn Huskies</strong> are very, very dominant. In fact the Huskies have the highest team rating ever <strong><em>(team_rating_average_custom)</em></strong> recorded in my database (since 2003) and from looking at historical data based Uconn’s team rating I would say the Huskies are a lock for making the <strong>Elite 8</strong>. Something else to note is that five out of the eight (5/8) true contenders are <strong>blue bloods. </strong>From what I’ve heard the Kansas Jayhawks are quite banged up (Dickinson &amp; McCoullor), so I wouldn’t feel to good about them.</p><p>Now these specific results were combination of two models averaged together <strong>(RandomForest &amp; Ridge)</strong>, which showed better results surrouding the top 6 predicted teams. Basically this model assures an <strong>88% chance</strong> that the champion is with in the t<strong>op 6 predicted teams</strong>.</p><p>Now the model is a regression type model, so it’s not necessarily built for probabilities like a classification model so a round by round forecast isn’t entirely accuracte for the top 8 teams or so (top 8 would be greater). But nevertheless I wanted to see what it would look like, so here’s a full forecast below.</p><p><em>Random-Forest Model (Top Pred.) — Purdue</em></p><p><em>Ridge Regression Model (Top Pred.)-Uconn</em></p><p><em>Xgboost Regression Model (Top Pred.)-Arizona</em></p><p><strong>Region Difficult (Per Team Pred. Wins Avg.):</strong></p><p><strong>South 1.21, East, 1.11, MidWest 1.04, and West 1.03</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/879/1*nnbGj0PJhDMHN-RbnZppfw.png" /><figcaption>*Post-Tourney Update* | Regression Forecast Probabilities</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ZllOJkwl7jyVL1AmBMWtuQ.png" /><figcaption>My GitHub | Interactive Power BI Dashboard | 2024 Heat Table</figcaption></figure><h3>Feedback</h3><p>Any questions or feedback are always welcome, so feel free to post them. <strong>For full details</strong> of all steps done for this project check out my <a href="https://github.com/allenjake440/Mens_March_Madness_Champion"><strong>GitHub.</strong></a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=081d6688ea71" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[2024 NFL Champion Machine Learning Prediction]]></title>
            <link>https://allenjake440.medium.com/2024-nfl-champion-machine-learning-prediction-f76d17eae962?source=rss-b69642b334cd------2</link>
            <guid isPermaLink="false">https://medium.com/p/f76d17eae962</guid>
            <category><![CDATA[2024nflchampion]]></category>
            <dc:creator><![CDATA[JakeAllenData]]></dc:creator>
            <pubDate>Fri, 12 Jan 2024 14:48:18 GMT</pubDate>
            <atom:updated>2024-05-09T18:54:45.012Z</atom:updated>
            <content:encoded><![CDATA[<p>Using <strong>Machine Learning</strong> in preparation to predict the <strong>2024 NFL Champion</strong> right before the playoffs <strong>(January 2024).</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/600/1*Pq9zZE41PHTeD8VAZVQblQ.png" /><figcaption>Photo by Author and <a href="https://dribbble.com/shots/9328989-Lombardi/attachments/1369972?mode=media">Zilligen Design Studio</a> on the Dribble</figcaption></figure><h3>Introduction</h3><p>The 2024 NFL regular season has come to close and the playoffs are under way. As per usually many fans love to fill out playoff prediction brackets using nothing but their own intution, feeling, and emtions to decide what teams move on and what teams drop out. This prediction purely the opposite of that, the predictions presented in this article are fully based on just the numbers.</p><p><a href="https://www.youtube.com/shorts/l2hfh-CEgNw">I Know Who&#39;s Going to Win the 2024 NFL Championship... Maybe</a></p><h3>2024 NFL Champion</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*OOhNde8ldif-bBq5K5HVGA.png" /><figcaption>*Post-Tourney Update* | Regression Forecast Probabilities</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/630/0*yM37FcVSiD9T7ODJ.png" /><figcaption>2024 NFL Champion Machine Learning Predictions</figcaption></figure><p>Despite most fans picking the <strong>Baltimore Ravens</strong> to <strong>win</strong> the <strong>Super-Bowl</strong>, the <strong>San Francisco 49ers</strong> are deemed to be the best pick according to the numbers 😯. The <strong>Kansas City Chiefs</strong> are still shining through as a contender despite the shakey season. Now a potential dark horse team is the <strong>Los Angeles Rams</strong>, this team is among a top 6 contender despite being a 6 seeded team.</p><h3>Feedback</h3><p>Any questions or feedback are always welcome, so feel free to post them. <strong>For full details</strong> of all steps done for this project check out my <a href="https://github.com/allenjake440/NFL_Champion"><strong>GitHub.</strong></a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f76d17eae962" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Predicting The NBA Champion With Machine Learning]]></title>
            <link>https://allenjake440.medium.com/predicting-the-nba-champion-with-machine-learning-25e3a45a82f9?source=rss-b69642b334cd------2</link>
            <guid isPermaLink="false">https://medium.com/p/25e3a45a82f9</guid>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[nba-playoffs]]></category>
            <category><![CDATA[nba-champion]]></category>
            <dc:creator><![CDATA[JakeAllenData]]></dc:creator>
            <pubDate>Sat, 30 Dec 2023 17:54:48 GMT</pubDate>
            <atom:updated>2025-06-23T23:45:10.932Z</atom:updated>
            <content:encoded><![CDATA[<p>Using <strong>Machine Learning</strong> in preparation to predict the<strong> NBA Champion</strong> right before the playoffs <strong>for every up-coming year!</strong></p><blockquote>Article Orginally Published: <strong>12/30/2024</strong></blockquote><blockquote>Article Updated: 10/25/2024</blockquote><blockquote>Article Inspiration: <a href="https://thejk.medium.com/"><strong>TheJK</strong></a>, please check this guy out… <strong>He’s the Goat!</strong></blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/600/1*Zbig2BhJaj9W_oI7ug8IZg.png" /><figcaption>Photo by Author and <a href="https://dribbble.com/shots/14301082-NBA-Finals-2020">Michael Meek</a> on the Dribble</figcaption></figure><h3>Introduction</h3><p>The NBA playoffs is the post-season tournament held at the end of the regular NBA season. This tournament takes the top 8 teams from each conference (east/west), 16 teams total into a best-of-7-game elimination tournament. The playoffs confer glory among the best regular-season teams in the NBA every season. The playoffs are where players make names for themselves, it is where they show their worth to a team if it’s star players proving the narratives wrong, or unexpected role players making a surprising impact on the game. The unpredictable events affected by high-pressure situations are why the playoffs are so entertaining.</p><p>Many fans each year make predictions about why they believe this or these certain teams truly have the best chances to win it all.</p><p>In this article, I give an overview of how <strong>I created a machine learning model that has managed to predict the last 5 NBA Champions Correctly!</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*KJNrK_bYN0OIgY3ziiTyQA.png" /><figcaption>New Machine Learning Predictions as 8/1/2024</figcaption></figure><h3>Year By Year Predictions</h3><p><em>4/16/2025 | Personal Pick: Thunder/Celtics</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*UC69oylt4nPeKq95VkvUoA.png" /><figcaption>2025 Machine Learning NBA Champion Predictions | Date: 4/16/25</figcaption></figure><p><a href="https://allenjake440.medium.com/2025-nba-champion-machine-learning-prediction-8d9afb8db533">2025 NBA Champion Machine Learning Prediction</a></p><p><em>4/25/2024 | Personal Pick: Nuggets/Celtics</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*FY9SbTveAjuLtXHHxw6E7A.png" /><figcaption>2024 Machine Learning NBA Champion Predictions | Date: 4/12/24</figcaption></figure><ul><li><a href="https://youtube.com/shorts/jdaYqlurfyM?si=W8IZsuAjBxcVMUdX">The Boston Celitcs Are Champions!! Nearly Perfect Video... #nbachampionship #nbaplayoffs #nbafinals</a></li><li><a href="https://allenjake440.medium.com/2024-nba-champion-machine-learning-prediction-6bea90855891">2024 NBA Champion Machine Learning Prediction</a></li></ul><h3>Data</h3><p>All the data used for this project comes from <a href="https://www.basketball-reference.com/players/">basketball reference.</a> For this project I used player, coach, and team data that traces back to the 1950 season. For more information on the statistics <a href="https://www.basketball-reference.com/about/glossary.html">click here</a> for the glossary of the statistics.</p><p>All of the detailed steps of this project, including scraping, creating, and analyzing the data, is on my <a href="https://github.com/allenjake440/NBA_Champion">GitHub</a>.</p><p><em>Python Libraries: np, pd, BeautifulSoup, selenium, StringIO, Commet, webdriver</em></p><h3>Model Prediction Value</h3><p>The value to be calculated by the model is <strong><em>champion_share</em></strong>. This is a custom statistic is calculated from the ending post-season standing. An example of this calculation is the 2023 title runner-up <strong>Miami Heat </strong>who achieved 13 playoff wins out of the potential 16 wins it takes to win the title <strong>(13 / 16 = 0.8125)</strong>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/760/1*TAPV3SPNDAsDLROn6J-McA.png" /><figcaption>Table Calculation of Champion Share</figcaption></figure><h3>Machine Learning Model Results</h3><blockquote>Machine Learning Model Used: <strong>XG-Boost Regression (Complex)</strong></blockquote><blockquote>For full details feel free to investiage the code: <a href="https://github.com/allenjake440/NBA_Champion"><strong>My GitHub</strong></a></blockquote><p>This section covers <strong>4</strong> <strong>charts</strong> that captures every aspect of what it takes to become a <strong>Modern NBA Champion.</strong></p><ol><li>Season Playoff Comparison Chart (2024 Season)</li><li>Feature Importance Chart (Most Important Features)</li><li>Feature Importance Heat Table (2024 Season)</li></ol><p><em>Python Libraries: np, pd, sklearn, matplotlib, xgboost, statsmodels.api, shap, lime</em></p><p><strong>The Season Playoff Comparison Chart</strong> shows the differences between the prediction and actual results of the 2024 playoff.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*rp28vBwCuyh7xv8QO2Pcpw.png" /><figcaption>Season Playoff Comparison Chart (2024 Season)</figcaption></figure><p>One way to assess which features are most important for predicting the NBA playoffs is by examining the feature importance scores of the model. In this project, a complex tree-based model (Random-Forest Regression) was used, which calculates feature importance based on how much each feature contributes to reducing error in the predictions.</p><p>However in this models case, the <strong>Feature Importance</strong> <strong>Chart</strong> doesn’t indicate whether a feature has a positive or negative effect it simply shows how relevant each feature is for the model’s overall performance. This is a common trait of tree-based models, which split the data based on different features without assigning direct positive or negative weights to them.</p><p>XG-Boost builds one main decision tree, then sequentially adds additional decision trees that focus on correcting the errors from the previous tree. Each new tree tries to capture what the previous tree missed, which reduces the overall error. By iteratively building and adjusting these trees, XG-Boost effectively minimizes prediction errors and achieves a highly accurate final prediction.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/999/1*o4-s0FgDtg16xogYR0FwHg.png" /><figcaption>SHAP Feature Importance (Most Important Features)</figcaption></figure><p>For more information on how tree-based models, <a href="https://www.ibm.com/topics/random-forest">click here.</a></p><p>Now, for a more detailed view, here is the <strong>Feature Importance Heat Table</strong>, which displays the top 10 most important features in a tabular format. This table not only highlights how these key features contribute to the prediction but also reveals how they relate to one another.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*AtuV54xZJ1dKlbmbbdXq4A.png" /><figcaption>Feature Importance Heat Table (2024 Season)</figcaption></figure><h3>What Makes An NBA Champion?</h3><p>With all of the information presented, this article can briefly be summarized in these 4 points (in ascending order of importance).</p><ol><li>To become an NBA champion, a team must dominate throughout the season, consistently defeating opponents by large margins and securing a top 3 seed heading into the playoffs <strong><em>(top_3_conference, top_6_SRS).</em></strong></li><li>The team must also be highly regarded by oddsmakers, as every NBA champion has ranked within the top 9 in preseason odds <strong><em>(pso).</em></strong></li><li>Star power is essential, particularly MVP-caliber talent. The roster should feature one or more players with strong performances in past MVP races <strong>(<em>sum_mvp_shares</em>)</strong>. Having players who have excelled in Defensive Player of the Year (DPOY) contests <strong><em>(sum_dpoy_shares)</em></strong> also provides an advantage. Moreover, the presence of players who made an impact early in their careers, particularly in Rookie of the Year (ROY) races <strong><em>(sum_roy_shares)</em></strong>, underscores the team’s overall talent.</li><li>Experience plays a critical role in achieving postseason success. Championship teams typically include players with substantial playoff experience over the past three seasons <strong><em>(sum_franchise_L3S_cs).</em></strong> The team’s average age is also crucial players in their prime <strong><em>(Age)</em></strong> are generally best equipped to compete for a title.</li><li>Lastly, a successful team must be strong offensively and on the boards. The last 10 NBA champions have ranked in the top 5 in offensive effective field goal percentage <strong><em>(top_5_offensive_eFG%).</em></strong> Additionally, being one of the league’s top rebounding teams is crucial for withstanding the physical demands of the postseason <strong><em>(TRB).</em></strong></li></ol><h3>Conclusion</h3><p>Overall, I was really impressed with my results. By creating an expansive and creative dataset capturing all aspects of the players, coaches, and teams I was able to capture the complexity of the NBA playoffs using a complex tree-based regressor.</p><h3>Feedback</h3><p>Any questions or feedback are always welcome, so feel free to post them. <strong>For full details</strong> of all steps done for this project check out my <a href="https://github.com/allenjake440/NBA_Champion"><strong>GitHub.</strong></a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=25e3a45a82f9" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Predicting The Men’s March Madness Champion With Machine Learning]]></title>
            <link>https://allenjake440.medium.com/predicting-the-mens-march-madness-champion-with-machine-learning-892bd78997ca?source=rss-b69642b334cd------2</link>
            <guid isPermaLink="false">https://medium.com/p/892bd78997ca</guid>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[ncaa-mens-basketball]]></category>
            <category><![CDATA[ncaa-championship]]></category>
            <category><![CDATA[bracketology]]></category>
            <category><![CDATA[march-madness]]></category>
            <dc:creator><![CDATA[JakeAllenData]]></dc:creator>
            <pubDate>Tue, 26 Dec 2023 03:13:24 GMT</pubDate>
            <atom:updated>2025-05-07T02:20:30.583Z</atom:updated>
            <content:encoded><![CDATA[<p>Using <strong>Machine Learning</strong> in preparation to predict the <strong>NCAA Men’s March Madness Champion</strong> right before the tournament <strong>for every up-coming year!</strong></p><blockquote>Article Originally Published: 12/25/2023</blockquote><blockquote>Article Updated: 3/18/2025</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/600/1*CLBr7Eu9JzKanuvdUgISig.png" /><figcaption>Photo by Author and <a href="https://dribbble.com/shots/4226205-March-Madness-Basketball">Stephen Piercy</a> on the Dribble</figcaption></figure><h3>Introduction</h3><p>March Madness, also known as the NCAA Men’s Division 1 basketball post-season tournament, takes in the best college basketball teams from each conference into a massive 64-team single-elimination tournament to determine the winner of the national college basketball season. A tournament this large being all single-game elimination brings out incredible performances and many upsets every single year, which is what makes this tournament so exciting to watch.</p><p>Because of the unpredictability of the tournament, thousands of people fill out many brackets each year in order to both predict the winner and attempt to make a perfect bracket. The odds of having either a perfect bracket or predicting the champion correctly are very long and in most cases impossible.</p><p>Well, hold up and wait a minute because this machine learning model I made <strong>nearly</strong> predicted the <strong>last 4 tournament champions correctly!</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Af0PCsATgW_33ZDy_Gh5qg.png" /><figcaption>New Machine Learning Predictions as of 8/2/2024</figcaption></figure><h3>Year By Year Predictions | 2 for 2</h3><h3>2025 Men’s NCAA March Madness Champion Predictions</h3><p><em>*Personal Pick — Florida Gators and Duke Blue Devils*</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/700/0*96RphA9slcO6oVim.png" /><figcaption>2025 NCAA Men’s MM Tour. Machine Learning Predictions | Date: 3/18/25</figcaption></figure><p><a href="https://allenjake440.medium.com/2025-march-madness-machine-learning-prediction-e5b3a66b2c59">2025 March Madness Machine Learning Prediction</a></p><h3>2025 Men’s NCAA March Madness Champion Predictions</h3><p><em>*Personal Pick —Uconn Huskies and Arizona Wildcats*</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*lWdz2Iwyzp2tOk5_Db_gMg.png" /><figcaption>2024 Machine Learning MM Predictions | Date: 3/18/24</figcaption></figure><p><a href="https://allenjake440.medium.com/2024-march-madness-machine-learning-prediction-081d6688ea71">2024 March Madness Machine Learning Prediction</a></p><h3>Data</h3><p>The data used for this project was from <a href="https://basketball.realgm.com/ncaa/team-stats">basketballrealgm.com</a>. For this project I used player and team data that traces back to the 2003 season. For more information on the statistics <a href="https://basketball.realgm.com/info/glossary"><strong>click here</strong></a> for the glossary of the statistics.</p><h4>IMPORTANT:</h4><blockquote>I chose this site over <strong>college basketball reference</strong> because many sites that track college basketball data take into account the post-season tournament (March Madness) in their regular season database. <strong>Not knowing this</strong> can cause your data to be <strong>overfitted or faulty</strong>, as post tournament data is already included.</blockquote><p><a href="https://basketball.realgm.com/ncaa/team-stats">NCAA Team Stats - RealGM</a></p><h3>Model Prediction Value</h3><p>The value to be calculated by the model is <strong>Champion Share</strong>. This is a custom statistic calculated from the ending post March Madness standing. An example of this calculation is the 2023 title runner-up <strong>San Diego State </strong>who achieved 5 tournament wins out of the potential 6 wins it takes to win the tournament (5 / 6 = 0.83).</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/746/1*tpXS6QIbi2-sBp53X-RtTQ.png" /><figcaption>Table Calculation of Champion Share</figcaption></figure><h3>Machine Learning Model Results | Ridge Regression</h3><blockquote>Machine Learning Model Used: <strong>Ridge Regression (Simple)</strong></blockquote><blockquote>For full details feel free to investiage the code: <a href="https://github.com/allenjake440/Mens_March_Madness_Champion"><strong>My GitHub</strong></a></blockquote><p>In this section will cover<strong> 3 charts</strong> that capture every aspect of what it takes to become a Modern March Madness Champion.</p><ol><li>Season Tournament Comparison Chart (2024 Season)</li><li>Feature Importance Chart (Most Important Statistics)</li><li>Feature Importance Heat Table (2024 Season)</li></ol><p><strong>The Season Tournament Comparison Chart</strong> shows the differences between the prediction and actual results of the 2024 March Madness tournament.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*SOtEtwh6aXH6EltbOMdaYw.png" /><figcaption>Season Tournament Comparison Chart (2024 Season)</figcaption></figure><p>One way to assess which features are most important for predicting the Men’s March Madness Division 1 Tournmanet outcomes is by examining the coefficients of the ridge regression model. In this project, ridge regression was used, which provides direct insight into how each feature impacts the target variable. The coefficients represent the strength and direction (positive or negative) of the relationship between each feature and the target outcome.</p><p>Unlike tree-based models, which show overall feature importance without indicating whether the impact is positive or negative, the coefficient bar chart in ridge regression clearly shows both the magnitude and direction of each feature’s influence on the predictions. Features with positive coefficients push predictions higher, while those with negative coefficients reduce the predicted value.</p><p>Ridge regression operates like linear regression by fitting a line that minimizes the sum of squared errors (distance between target and prediction). However, it includes an <strong>L2 regularization term</strong>, which penalizes large coefficients, effectively shrinking them and preventing any one feature from dominating the prediction. This regularization helps especially when features are highly correlated, as it distributes the weight more evenly and reduces overfitting, leading to a more stable model.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/997/1*7DqETw4mJzzozuL-vrHsww.png" /><figcaption>Feature importance (Most Important Features)</figcaption></figure><p>For information on linear Models, <a href="https://www.ibm.com/topics/ridge-regression">click here.</a></p><p>For example, features <strong><em>top_4_seed</em></strong> and <strong><em>AdjOE</em></strong> are showing massive and positive direct correlations to tournament success. While features like <strong><em>Ast/TO</em></strong> and <strong><em>TOV%</em></strong> show negative direct correlation to tournament success.</p><p>Now, for a more detailed view, here is the <strong>Feature Importance Heat Table</strong>, which displays the top 10 most important features in a tabular format. This table not only highlights how these key features contribute to the prediction but also reveals how they relate to one another.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*-Mfa4iBR8_URil0zE_larg.png" /><figcaption>Feature Importance Heat Table (2024 Season)</figcaption></figure><h3>What Makes A Men’s March Madness Champion?</h3><p>With all of the information presented, this article can briefly be summarized in these <strong>5 points</strong> (in ascending order on importance).</p><ol><li>To be a true contender, a team must be recognized by the selection committee as a dominant force. This includes securing a top 4 seed and ranking analytically among the top 6 teams in the country, as measured by KenPom’s efficiency margin <strong><em>(top_4_seed, top_6_AdjEM).</em></strong></li><li>In addition, the team must operate a highly efficient offensive scheme while playing disciplined basketball. This means minimizing turnovers, capitalizing on free throws, and making smart, calculated plays <strong><em>(AdjOE, TOV%, Ast/TO, FTM).</em></strong></li><li>Experience is crucial for both the team and the coach. Teams with a strong history of participation in previous March Madness tournaments have a competitive edge <strong><em>(sum_player_champions, sum_mm_tour_games).</em></strong></li><li>The team must also play physically, with an above-average defense and strong performance on both the offensive and defensive boards <strong><em>(DRtg, DRB, ORB%).</em></strong></li><li>Lastly, the team should come from a highly competitive conference that consistently produces strong March Madness performers. Champions often establish their dominance by mid-season, with nearly all past champions being ranked in the top 12 of the AP Polls by Week <strong><em>(sum_conf_L5S_cs, top12_AP_Week 6).</em></strong></li></ol><h3>Conclusion</h3><p>I’m pleased with the results, especially considering the unpredictable nature of the NCAA postseason tournament. By leveraging a large dataset and engineering a variety of features that highlight the impact of players, coaches, and the historical success of each school, I was able to capture key factors influencing performance. Surprisingly, even a simple linear model delivered impressive results, demonstrating the power of thoughtful feature creation and data-driven insights.</p><h3>Feedback</h3><p>Any questions or feedback are always welcome, so feel free to post them. <strong>For full details</strong> of all steps done for this project check out my <a href="https://github.com/allenjake440/Mens_March_Madness_Champion"><strong>GitHub.</strong></a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=892bd78997ca" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Predicting The NFL Champion With Machine Learning]]></title>
            <link>https://allenjake440.medium.com/predicting-the-nfl-champion-with-machine-learning-7c6eede5a4d2?source=rss-b69642b334cd------2</link>
            <guid isPermaLink="false">https://medium.com/p/7c6eede5a4d2</guid>
            <category><![CDATA[data]]></category>
            <category><![CDATA[nfl]]></category>
            <category><![CDATA[nfl-super-bowl]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[predictions]]></category>
            <dc:creator><![CDATA[JakeAllenData]]></dc:creator>
            <pubDate>Fri, 10 Nov 2023 18:09:51 GMT</pubDate>
            <atom:updated>2025-01-10T22:47:22.865Z</atom:updated>
            <content:encoded><![CDATA[<p>Using <strong>Machine Learning</strong> in preparation to predict the <strong>NFL Champion</strong> right before the playoffs <strong>for every up-coming year!</strong></p><blockquote>Article Orginally Made: 11/10/23</blockquote><blockquote>Article Updated: 1/10/2025 “Final 24–25 Predictions In!”</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/600/1*aN7ov7ntbvGtYdMWjIxWqg.png" /><figcaption>Photo by Author and <a href="https://dribbble.com/shots/9328989-Lombardi/attachments/1369972?mode=media">Zilligen Design Studio</a> on the Dribble</figcaption></figure><h3>Introduction</h3><p>The NFL playoffs are a 14 team single game elimination tournament, and one of the most exciting sporting events in the world. The <strong>Super Bowl</strong> itself has become an American tradition as it is the most watched sporting event every year in the United States. The complexity of the game, physicality, athleticism, and strategy is what makes American football one of the most popular sporting leagues.</p><p>Now in regards to the playoffs, many fans and analysts every year make predictions of why they believe this or that team has the best chance to win it all. As a sports fan, often times I find our feelings and intution can deceive us. Look the NFL playoffs is crazy, single-elimination format, not many regular season games, trades, I guess the question I am asking is…</p><blockquote>“Can numbers even accurately measure the vast and wild NFL landscape…</blockquote><blockquote>Well that answer is…. Yes, kind of!”</blockquote><p>I have came-back and spent more time analyzing the NFL post-season, identified some new trends, and managed to create a new model that has captured a better understanding of the NFL post-season. The model in particular has ensured that the eventual champion is with in the top 3 predicted teams…<strong>for the last 10 seasons.</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*yGXCRJG5PllWbU31_t9W1w.png" /><figcaption>New Machine Learning Predictions as 10/24/24</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*y07zzmuxwE6RvIdOtwMAhA.png" /><figcaption>New Machine Learning Predictions as 10/24/24</figcaption></figure><h3>Year-by-Year Predictions | 1 for 2 | 25 in progress…</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*h0MSbfBgLAy5dnEngqTB4g.png" /><figcaption>2025 Machine Learning NFL Champion Predictions | Date: 1/10/25</figcaption></figure><p><a href="https://allenjake440.medium.com/2025-nfl-champion-machine-learning-prediction-4fc82113cee2">2025 NFL Champion Machine Learning Prediction</a></p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FsiFcXUAwb3g%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DsiFcXUAwb3g&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FsiFcXUAwb3g%2Fhqdefault.jpg&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/4a55f49b811b19f27b77a53252a9e1f4/href">https://medium.com/media/4a55f49b811b19f27b77a53252a9e1f4/href</a></iframe><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*i2SI8SgIlH-xizhXrYs6Vg.png" /><figcaption>2024 Machine Learning NFL Champion Predictions | Date: 1/11/24</figcaption></figure><p><a href="https://allenjake440.medium.com/2024-nfl-champion-machine-learning-prediction-f76d17eae962">2024 NFL Champion Machine Learning Prediction</a></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*h2qSQmzgfgNce7DLv2uRLg.png" /><figcaption>2023 Formula-Based NFL Champion Predictions | Date: 1/8/24</figcaption></figure><h3>Data</h3><p>All the data used for this project comes from <a href="https://www.pro-football-reference.com/">football reference</a>. For this project I used player and team data that traces back to the 1991 season. For more information on the statistics <a href="https://www.pro-football-reference.com/about/glossary.htm">click here</a> for the glossary of the statistics.</p><p><a href="https://www.pro-football-reference.com/">Pro Football Stats, History, Scores, Standings, Playoffs, Schedule &amp; Records | Pro-Football-Reference.com</a></p><h3>Model Prediction Value</h3><p>The value to be calculated by the model is <strong><em>champion_share</em></strong>. This is a custom statistic calculated from the ending post-season standing. An example of this calculation is the 2023 title runner-up <strong>Philadelphia Eagles</strong> who achieved 3 playoff wins out of the potential 4 wins it takes to win the title <strong>(3 / 4 = 0.75)</strong>.</p><p>Something else to mention is that “bye” seeds 1 or 2, it depends on season, automatically get a playoff win in the calculation. This keeps it consisent amongst all other teams in the playoff field.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/668/1*KbHGDBr6xmRb1JF7diUb9Q.png" /><figcaption>Table Calculation of Champion Share</figcaption></figure><h3>Machine Learning Model Results</h3><blockquote>Machine Learning Model Used: <strong>Random-Forest Regression (Complex)</strong></blockquote><blockquote>For full details feel free to investiage: <a href="https://github.com/allenjake440/NFL_Champion"><strong>My GitHub</strong></a></blockquote><p>This section covers<strong> 3 charts</strong> that captures every aspect of what it takes to become a <strong>Modern NFL Champion</strong>.</p><ol><li>Season Playoff Comparison Chart (2024 Season)</li><li>Feature Importance Chart (Most Important Statistics)</li><li>Feature Importance Heat Table (2024 Season)</li></ol><p><strong>The Season Playoff Comparison Chart</strong> shows the differences between the prediction and actual results of the 2024 playoff.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Sw7cwnPO3uPOyd_PRBW55w.png" /><figcaption>Season Playoff Comparison Chart (2024 Season)</figcaption></figure><p>One way to assess which features are most important for predicting the NBA playoffs is by examining the feature importance scores of the model. In this project, a complex tree-based model (XGBoost Regression) was used, which calculates feature importance based on how much each feature contributes to reducing error in the predictions.</p><p>However in this models case, the <strong>Feature Importance</strong> <strong>Chart</strong> doesn’t indicate whether a feature has a positive or negative effect — it simply shows how relevant each feature is for the model’s overall performance. This is a common trait of tree-based models, which split the data based on different features without assigning direct positive or negative weights to them.</p><p>Random-Forest builds multiple decision trees independently, with each tree trained on a random subset of the data and features. By averaging the predictions from all these trees, Random Forest produces a nice generalized and stable final prediction.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/998/1*fIkTm4rLmja0zE-pyqs3yA.png" /><figcaption>Feature Importance Chart (Most Important Statistics)</figcaption></figure><p>For more information on how tree-based models, <a href="https://www.ibm.com/topics/random-forest">click here.</a></p><p>Now, for a more detailed view, here is the <strong>Feature Importance Heat Table</strong>, which displays the top 10 most important features in a tabular format. This table not only highlights how these key features contribute to the prediction but also reveals how they relate to one another.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*fAlpdwlu-xdEB_8_wC4uKw.png" /><figcaption>Feature Importance Heat Table (2024 Season)</figcaption></figure><h3>What Makes An NFL Champion?</h3><p>With all of the information presented, this article can briefly be summarized in these 4 points (in ascending order on importance).</p><ol><li>The team should rank highly in my custom team rating metric, which assesses overall performance for the season. This rating considers the team’s regular season record, player star power, margin of victory, and kicker’s deep range efficiency for that season. <a href="https://github.com/allenjake440/NFL_Champion/blob/main/custom_team_qb_stat.md"><strong>Click here</strong></a> for full details of the custom team season ratingstatistic<strong><em>(team_rating_custom, top_5_team_rating_custom, 40_50+_perc_x_makes).</em></strong></li><li>A strong offensive line and an experienced quarterback — ideally one who has received MVP votes or is a seasoned veteran — are essential. The quarterback should rank among the top seven in my custom quarterback metric, demonstrating efficiency, particularly in yards per completion, throughout the regular season. <a href="https://github.com/allenjake440/NFL_Champion/blob/main/custom_team_qb_stat.md"><strong>Click here</strong></a> for full details of the custom quarterback season rating statistic <strong><em>(passing_offense_Y/C, QBSRC_rk).</em></strong></li><li>A championship team also requires a strong, disciplined defense and offense <strong><em>(team_offense_Penalties_Pen)</em></strong>, proven either in previous seasons or during the current season <strong><em>(team_defense_EXP</em></strong>). Importantly, the current season’s defense should excel at stopping the run game <strong><em>(team_defense_Rushing_Yds)</em></strong> and be equally adept at defending the pass <strong><em>(team_defense_Tot Yds &amp; TO_Ply)</em></strong>.</li><li>Finally, the team must have a coach with extensive playoff experience. This includes coaching several past playoff games, which is crucial for navigating the high-pressure environment of the post-season <strong><em>(sum_coach_playoff_games).</em></strong></li></ol><h3>Conclusion</h3><p>Overall, out of all of my projects this was by far the most unpredictable when comparing it to my NBA and the Men’s NCAA Division I Tournament Champion projects. By leveraging a large dataset, creating numerous features to evaluate the strength of individual players and coaches, and lastly using a complex tree-based regressor. The model to my surprise delivered promising results despite the inherent unpredictability.</p><h3>Feedback</h3><p>Any questions or feedback are always welcome, so feel free to post them. <strong>For full details</strong> of all steps done for this project check out my <a href="https://github.com/allenjake440/NFL_Champion"><strong>GitHub.</strong></a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=7c6eede5a4d2" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>