<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <channel>
    <title>DEV Community: CrystlQuant</title>
    <description>The latest articles on DEV Community by CrystlQuant (@crystlquant).</description>
    <link>https://dev.to/crystlquant</link>
    <image>
      <url>https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F4237449%2F7fb88535-d99b-41b8-b960-667a104484c9.png</url>
      <title>DEV Community: CrystlQuant</title>
      <link>https://dev.to/crystlquant</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/crystlquant"/>
    <language>en</language>
    <item>
      <title>NFL Teams That Score 40 Points Win 96% of the Time. The Cowboys Just Became the 26th to Lose</title>
      <dc:creator>CrystlQuant</dc:creator>
      <pubDate>Mon, 16 Nov 2026 16:02:12 +0000</pubDate>
      <link>https://dev.to/crystlquant/nfl-teams-that-score-40-points-win-96-of-the-time-the-cowboys-just-became-the-26th-to-lose-4m99</link>
      <guid>https://dev.to/crystlquant/nfl-teams-that-score-40-points-win-96-of-the-time-the-cowboys-just-became-the-26th-to-lose-4m99</guid>
      <description>&lt;p&gt;&lt;em&gt;This post first ran on &lt;a href="https://crystlquant.com/blog/nfl-teams-scoring-40-points-win-96-percent" rel="noopener noreferrer"&gt;CrystlQuant&lt;/a&gt;.&lt;/em&gt;&lt;/p&gt;

&lt;p&gt;On Sunday afternoon in Arlington, the Dallas Cowboys scored 40 points and lost. The San Francisco 49ers won 46-40, and the Cowboys walked off as members of one of the smallest clubs in football.&lt;/p&gt;

&lt;p&gt;Forty points is supposed to settle it. Our tables hold 7,091 NFL regular-season games with a final score, from September 12, 1999 through the late-afternoon games of November 15, 2026. In those games a team scored 40 or more 741 times. It won 713 of them, 96.2%. It tied twice, and both ties came in the same game. It lost 26 times. Dallas on Sunday was number 26.&lt;/p&gt;

&lt;h2&gt;
  
  
  Why 40 almost never loses
&lt;/h2&gt;

&lt;p&gt;An NFL team scores 23.0 points in an average game this season, through November 15, 2026. In 1999 it was 20.8. Forty is close to double a normal day. For a team that gets there to lose, the other side has to have its own extraordinary day in the same three hours, and finish ahead.&lt;/p&gt;

&lt;p&gt;That has happened in 27 of the 7,091 games, about one in 263: 26 with a winner and one 40-40 tie. The tie belongs to Dallas too. On September 28, 2025, the Cowboys and the Green Bay Packers finished 40-40 in Arlington. Twelve years before that, on October 6, 2013, the Denver Broncos won 51-48 in the same building. Since 1999 the Cowboys have scored 40 or more 36 times and won 33. All three that got away were at home.&lt;/p&gt;

&lt;h2&gt;
  
  
  It used to be rarer
&lt;/h2&gt;

&lt;p&gt;The club is growing. From the 1999 season through 2012, teams scored 40 or more 341 times and lost 8 of them, about one in 43. From 2013 through 2025 they did it 387 times and lost 17, about one in 23. The worst single year was 2018, when four teams scored 40 and lost, among them the Kansas City Chiefs in a 54-51 Monday-night defeat to the Los Angeles Rams on November 19.&lt;/p&gt;

&lt;p&gt;Why the jump happened is an argument, not a count. The rule changes of the last two decades protect quarterbacks and receivers more than they used to, and offenses have leaned into the passing game. When both teams can score fast, no lead is safe, and 40 stops being enough.&lt;/p&gt;

&lt;h2&gt;
  
  
  When both teams reach 40, the favorite stops mattering
&lt;/h2&gt;

&lt;p&gt;Here is the part that surprised us. In an ordinary NFL game, the market's favorite at kickoff wins about two times in three: 4,586 of 6,874 decided regular-season games from 1999 to 2025, or 66.7%. In the 25 decided games over the same years where both teams scored 40, the favorite won 12. That is 48.0%, worse than half.&lt;/p&gt;

&lt;p&gt;Some of those favorites were not close calls before kickoff. On September 9, 2018, the New Orleans Saints were 10-point favorites at home, and the Tampa Bay Buccaneers beat them 48-40. A year later the Buccaneers did it again in Los Angeles, winning 55-40 against a Rams team favored by nine.&lt;/p&gt;

&lt;p&gt;Our read is simple. A pregame price is mostly a judgment about which team is better, and a big part of what makes a team better is its defense. A game where both teams score 40 is a game where neither defense showed up, so the gap the price was measuring has quietly left the building. What is left is a race, and a race doesn't care much who was supposed to win it.&lt;/p&gt;

&lt;h2&gt;
  
  
  What CrystlQuant's numbers said on Sunday
&lt;/h2&gt;

&lt;p&gt;They say nobody beats the market. One model never sees the market price. The other fixes it. Before kickoff on Sunday, CrystlQuant had the 49ers at 65.4%. The market had them at 65.1%. Blindside, the model that never sees the price, had them at 61.9%. All three named San Francisco, and San Francisco won.&lt;/p&gt;

&lt;p&gt;Here is our take. A 65.4% number is not a promise about one afternoon. It is a statement about all the ways a game can go, and in nearly all of them nobody scores 40. Sunday was the rare version, the kind where history says the favorite holds only about half the time, and the 49ers held. They are 8-1 through November 15, the best record in the league; Dallas are 4-6. A shootout is the game a forecast can say least about, which is exactly why we grade every game in public, the shootouts included. Every NFL game, with CrystlQuant's number beside the market's before kickoff and graded after, is on &lt;a href="https://crystlquant.com/nfl" rel="noopener noreferrer"&gt;CrystlQuant's NFL page&lt;/a&gt;.&lt;/p&gt;




&lt;p&gt;Every game this week, with our number and the market's side by side, is free at &lt;a href="https://crystlquant.com" rel="noopener noreferrer"&gt;crystlquant.com&lt;/a&gt;.&lt;/p&gt;

</description>
      <category>nfl</category>
      <category>sports</category>
      <category>datascience</category>
      <category>analytics</category>
    </item>
    <item>
      <title>NFL Home Teams Are 66-68-1 After 135 Games, the Same Record as the Empty-Stadium 2020</title>
      <dc:creator>CrystlQuant</dc:creator>
      <pubDate>Thu, 12 Nov 2026 19:25:41 +0000</pubDate>
      <link>https://dev.to/crystlquant/nfl-home-teams-are-66-68-1-after-135-games-the-same-record-as-the-empty-stadium-2020-49gf</link>
      <guid>https://dev.to/crystlquant/nfl-home-teams-are-66-68-1-after-135-games-the-same-record-as-the-empty-stadium-2020-49gf</guid>
      <description>&lt;p&gt;&lt;em&gt;This post first ran on &lt;a href="https://crystlquant.com/blog/nfl-home-teams-66-68-after-135-games" rel="noopener noreferrer"&gt;CrystlQuant&lt;/a&gt;.&lt;/em&gt;&lt;/p&gt;

&lt;p&gt;On the second Sunday of November, the road team won almost everywhere. The Giants won 19-11 in Philadelphia. The Packers scored 45 in Foxborough and beat the Patriots by 18. The Texans put 44 points on the Chargers in Los Angeles, and the Bengals beat the Falcons 34-17 in Madrid, in a game that counted as an Atlanta home game. By the time the Bills finished off the Vikings 23-6 on Monday night, road teams had won 10 of the 15 games in Week 9.&lt;/p&gt;

&lt;p&gt;One strange week would be a footnote. This one tipped the whole season over. Through November 9, 2026, NFL home teams have won 66 regular-season games and lost 68. The one tie was the Giants and Commanders at 27-27 on October 11. Road teams are ahead.&lt;/p&gt;

&lt;p&gt;That is not supposed to happen. Our tables hold every NFL regular-season game since September 12, 1999, and from 1999 to 2018 home teams won 2,902 of 5,073 of them: 57.2%. No season in those twenty years finished below 52.8%. In 27 full seasons, road teams have won more games than home teams exactly once.&lt;/p&gt;

&lt;h2&gt;
  
  
  The season with nobody in the seats
&lt;/h2&gt;

&lt;p&gt;That once was 2020, the year the experiment ran itself. Because of the pandemic, most stadiums were empty or close to it all season, and home teams finished 127-128-1, the only losing home record of any full season since 1999.&lt;/p&gt;

&lt;p&gt;Here is the strange part. After 135 games in 2020, home teams were 66-68 with one tie. After 135 games in 2026, they are 66-68 with one tie. Same wins, same losses, same single tie. One of those seasons had cardboard cutouts in the stands. This one has crowds, and the count is identical.&lt;/p&gt;

&lt;p&gt;2020 is the best evidence there is that a big share of home-field advantage was the crowd: the noise that makes a road offense go to a silent count, and the pressure that may lean on the officials. Take the fans out and the advantage almost disappeared. So why has it gone missing in a season when the fans are back?&lt;/p&gt;

&lt;h2&gt;
  
  
  It was shrinking before anybody noticed
&lt;/h2&gt;

&lt;p&gt;Because 2026 is not a bolt from the blue. It is the latest step in a long slide. From 1999 to 2018 hosts won 57.2% of games. From 2019 to 2025 they won 992 of 1,871, or 53.0%. Even 2019, the last season before the pandemic, came in at 51.6%, and 2021, the first full season with crowds back, at 51.5%.&lt;/p&gt;

&lt;p&gt;The market noticed. Measured in points, the closing price gave the home team an average head start of 2.4 points a game from 1999 to 2016. From 2021 to 2025 that head start was 1.6 points, and this season it is 1.7. Playing at home is still worth something, but what used to be close to two and a half points is now closer to one and a half.&lt;/p&gt;

&lt;p&gt;Why it shrank is an argument, not a count. Teams travel and recover far better than they did in 1999, with charter flights, sleep plans and staff whose whole job is getting players fresh for Sunday. Offenses practice for noise all week. Replay review fixes more of the calls a loud building might once have bent. None of those can be counted in a box score, but all of them push the same way.&lt;/p&gt;

&lt;h2&gt;
  
  
  Seven home games that were not
&lt;/h2&gt;

&lt;p&gt;Part of this season's number is in the schedule. Seven 2026 games so far have been played outside the United States: in Melbourne, Rio de Janeiro, three times in London, in Paris and in Madrid. The team listed as the host won one of them, the Jaguars' 19-17 over the Eagles at Tottenham on October 11. The visitors won the other six.&lt;/p&gt;

&lt;p&gt;Take those seven out and hosts are 65-62-1 at their own stadiums. Still ahead, but barely, and nowhere near the 57.2% of 1999 to 2018.&lt;/p&gt;

&lt;p&gt;A lead at this point of the season has also been lost before. After 135 games of 2021, road teams were ahead 70-65. Home teams then won 75 of the remaining 137 games, lost 61 and tied one, and finished the year 140-131-1. Half a season is a long time in the NFL.&lt;/p&gt;

&lt;h2&gt;
  
  
  What CrystlQuant's numbers say
&lt;/h2&gt;

&lt;p&gt;CrystlQuant publishes two win chances for every NFL game before kickoff. They say nobody beats the market. One model never sees the market price. The other fixes it.&lt;/p&gt;

&lt;p&gt;Over the 134 games of 2026 that had a winner, the closing price gave the home team about 55% on average, which works out to about 74 home wins. Home teams won 66. The CQ blind model, which builds its numbers without ever seeing the price, leaned on home field even harder, at more than 57% on average. CrystlQuant itself went the other way. Across the 104 games it has published since September 22, CrystlQuant has the home team at 51.2% on average, against the market's 54.1% in the same games. The hosts won 51 of those 104.&lt;/p&gt;

&lt;p&gt;Here is our take. Home field in the NFL is now a small thing, about a point and a half, and this season it has been smaller still. A forecast that still gives the home side an old-fashioned bump is paying for something the league stopped delivering years ago, and the one number in our three that shaved home field the most is the one that has been closest to the truth this season. Whether hosts climb back above .500, the way they did in 2021, is for the second half to settle. Either way, the days when playing at home won 57 games in 100 are over. Every NFL game, with CrystlQuant's number beside the market's before kickoff and graded after, is on &lt;a href="https://crystlquant.com/nfl" rel="noopener noreferrer"&gt;CrystlQuant's NFL page&lt;/a&gt;.&lt;/p&gt;




&lt;p&gt;Every game this week, with our number and the market's side by side, is free at &lt;a href="https://crystlquant.com" rel="noopener noreferrer"&gt;crystlquant.com&lt;/a&gt;.&lt;/p&gt;

</description>
      <category>nfl</category>
      <category>sports</category>
      <category>datascience</category>
      <category>analytics</category>
    </item>
    <item>
      <title>NFL home teams win only 45% on Thanksgiving (33 of 73 since 1999)</title>
      <dc:creator>CrystlQuant</dc:creator>
      <pubDate>Tue, 10 Nov 2026 17:01:27 +0000</pubDate>
      <link>https://dev.to/crystlquant/nfl-home-teams-win-only-45-on-thanksgiving-33-of-73-since-1999-2ccn</link>
      <guid>https://dev.to/crystlquant/nfl-home-teams-win-only-45-on-thanksgiving-33-of-73-since-1999-2ccn</guid>
      <description>&lt;p&gt;&lt;em&gt;This post first ran on &lt;a href="https://crystlquant.com/blog/nfl-thanksgiving-home-teams-45-percent" rel="noopener noreferrer"&gt;CrystlQuant&lt;/a&gt;.&lt;/em&gt;&lt;/p&gt;

&lt;p&gt;Since 1999, NFL teams have played &lt;strong&gt;73 games on Thanksgiving&lt;/strong&gt;. The home team won &lt;strong&gt;33 of them: 45.2%&lt;/strong&gt;. In the other 6,828 regular-season games over the same years, home teams won 3,842: 56.3%.&lt;/p&gt;

&lt;p&gt;That is every regular-season game between September 12, 1999 and January 4, 2026: 27 seasons. Thanksgiving here means the NFL's Thursday games in the fourth week of November. There were two a year until 2005 and three from 2006, except 2020, when the third game moved off the holiday.&lt;/p&gt;

&lt;h2&gt;
  
  
  Who has been hosting, and how it went
&lt;/h2&gt;

&lt;div class="table-wrapper-paragraph"&gt;&lt;table&gt;
&lt;thead&gt;
&lt;tr&gt;
&lt;th&gt;Host&lt;/th&gt;
&lt;th&gt;Games&lt;/th&gt;
&lt;th&gt;Home wins&lt;/th&gt;
&lt;th&gt;Share&lt;/th&gt;
&lt;/tr&gt;
&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;
&lt;td&gt;Detroit Lions&lt;/td&gt;
&lt;td&gt;27&lt;/td&gt;
&lt;td&gt;8&lt;/td&gt;
&lt;td&gt;29.6%&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;Dallas Cowboys&lt;/td&gt;
&lt;td&gt;27&lt;/td&gt;
&lt;td&gt;15&lt;/td&gt;
&lt;td&gt;55.6%&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;Night game (rotating host)&lt;/td&gt;
&lt;td&gt;19&lt;/td&gt;
&lt;td&gt;10&lt;/td&gt;
&lt;td&gt;52.6%&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;All Thanksgiving games&lt;/td&gt;
&lt;td&gt;73&lt;/td&gt;
&lt;td&gt;33&lt;/td&gt;
&lt;td&gt;45.2%&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;Every other regular-season game&lt;/td&gt;
&lt;td&gt;6,828&lt;/td&gt;
&lt;td&gt;3,842&lt;/td&gt;
&lt;td&gt;56.3%&lt;/td&gt;
&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;&lt;/div&gt;

&lt;p&gt;&lt;em&gt;NFL regular season, 1999 to 2025 seasons. Counted November 5, 2026, from the public nflverse &lt;code&gt;games.csv&lt;/code&gt;.&lt;/em&gt;&lt;/p&gt;

&lt;p&gt;The holiday number is mostly a Detroit number. The Lions won 8 of their 27 Thanksgiving home games in this period, and lost nine in a row from 2004 to 2012, then seven in a row from 2017 to 2023. Dallas sit close to the league's usual home rate, and so do the night-game hosts. Take Detroit out and Thanksgiving home teams won 25 of 46 (54.3%), which is ordinary.&lt;/p&gt;

&lt;h2&gt;
  
  
  The years it went one way
&lt;/h2&gt;

&lt;p&gt;In five seasons every Thanksgiving host lost: 2001, 2012, 2019, 2020 and 2021. Only three times did every host win: 1999, 2013 and 2024.&lt;/p&gt;

&lt;h2&gt;
  
  
  Thursdays that are not Thanksgiving
&lt;/h2&gt;

&lt;p&gt;The other Thursday games tell the opposite story. Home teams won 156 of the 256 non-holiday Thursday games in this period, 60.9%, a little better than the usual 56.3%.&lt;/p&gt;

&lt;h2&gt;
  
  
  What would make this wrong
&lt;/h2&gt;

&lt;p&gt;73 games is a small sample, and 27 of them belong to one team going through some very bad years. The holiday hosts are fixed before anyone knows who is good: a Thanksgiving home game is usually Detroit's or Dallas's, not a reward for a strong season. Seasons before 1999 are not in this count.&lt;/p&gt;




&lt;p&gt;Every game this week, with our number and the market's side by side, is free at &lt;a href="https://crystlquant.com" rel="noopener noreferrer"&gt;crystlquant.com&lt;/a&gt;.&lt;/p&gt;

</description>
      <category>nfl</category>
      <category>datascience</category>
      <category>sports</category>
      <category>analytics</category>
    </item>
  </channel>
</rss>
