Cloudflare Radar Analyzes How the 2026 World Cup Reshaped Global Internet Traffic

Cloudflare Radar Analyzes How the 2026 World Cup Reshaped Global Internet Traffic Cloudflare examined data from its global network across June and July 2026 to measure how the World Cup altered worldwide internet traffic patterns. Using normalized four-week baselines, the analysis revealed how kickoff times, team popularity, in-game breaks, and streaming habits drove traffic surges and drops across different regions. Traffic deviations were measured against a four-week rolling median using log₂ ratios to symmetrically evaluate surges and drops across high- and low-volume countries. Overnight matches between midnight and 8 a.m. local time produced the largest traffic spikes, whereas daytime matches often pulled traffic below normal as viewers watched the games. Argentina was the most-watched team with an average 1.17x traffic swing worldwide, and its quarterfinal win against Switzerland was the tournament's most disruptive match. During halftimes and hydration breaks, most countries experienced traffic spikes from viewers checking devices, while several nations relying heavily on streaming saw traffic drop as broadcasts paused. Online gambling traffic maintained elevated, steady activity throughout the tournament, and the final matches visibly altered global HTTP request volumes. Methodology and Kickoff Time Dynamics To evaluate the impact of individual matches, Cloudflare established a baseline using the median traffic of the preceding four weeks and converted deviations into symmetric log₂ ratios. Local kickoff times proved to be the primary driver of traffic behavior. Matches played between midnight and 8 a.m. local time yielded the highest deviations, frequently doubling normal traffic as viewers woke up or stayed up to watch. Conversely, games broadcast during normal waking hours often lowered device activity relative to baselines, as seen when Japan's traffic doubled during a late-night match against Brazil while Brazil's simultaneous midday traffic dipped around 40% below normal. Top Matches and Global Team Draw Measuring traffic displacement in the two-hour window following kickoff across stable national markets, Cloudflare found that the July 11 quarterfinal between Argentina and Switzerland caused the largest disruption, altering global internet traffic by a factor of roughly 1.26. This edged out the France versus Spain semifinal factor of 1.21. Argentina ranked as the most watched team overall, generating a median 17% deviation from baseline globally whenever they played, followed by high-profile sides such as France, Brazil, Portugal, Morocco, Spain, and Norway. In-Game Break Patterns and Streaming Discrepancies Cluster analysis of match-day traffic curves identified distinct behavioral patterns during breaks in play. In 44 countries representing 101 matches, internet activity rose sharply during halftimes and three-minute hydration breaks as fans turned to their mobile devices. In contrast, a cluster of eight countries—including Algeria, Tunisia, Jordan, Egypt, and DR Congo—experienced traffic drops during halftimes. MIME type breakdowns revealed that traffic in this second group was dominated by multimedia and streaming data, meaning viewers closed video streams during the 15-minute intermission rather than switching to general browsing. Iran represented a separate outlier due to baseline distortions following earlier internet shutdowns. Macro Trends and Tournament Finals Beyond individual fixtures, the tournament flattened the internet's standard weekly rhythm into a sustained daily profile, alongside a steady increase in HTTP requests to gambling websites. The tournament's footprint was also clearly visible at the global aggregate level during the closing stages. The final between Argentina and Spain drove traffic in both countries up by as much as 20 percentage points over expected levels, exhibiting distinct traffic fluctuations at kickoff, pauses in play, and the final whistle.

Cloudflare 发布 2026 年世界杯全球互联网流量深度分析 Cloudflare 结合其全球网络覆盖数据发布报告，揭示 2026 年世界杯如何通过开球时间、明星球队关注度以及用户观看习惯，深刻改变全球不同国家的网络流量形态与用户在线行为模式。 分析基于比赛前四周流量中位数建立基线，以 log₂ 比率衡量增跌幅度；开球时间对流量影响显著，深夜早晨比赛会引发流量暴涨（有时翻倍），日间比赛影响不大，晚间比赛则常促使流量降至常态以下。 7 月 11 日阿根廷对阵瑞士的四分之一决赛以约 1.26 倍的变动幅度成为影响全球流量最大的单场比赛，领先于法国对西班牙半决赛（1.21 倍）。 卫冕冠军阿根廷队以 1.17 倍的流量偏离度位列全球最受关注球队榜首，博彩网站流量也在揭幕战后呈现整体上升与平稳高频态势。 各国观赛行为聚类出截然相反的形态：多数国家用户在补水暂停及中场休息时拿起手机导致流量激增；而在阿尔及利亚等重度依赖流媒体直播的国家，中场休息时流媒体流量会显著下降。 基线设定与开球时间的影响 为消除不同国家总流量差异的影响，Cloudflare 将比赛前四周的每分钟中位数定义为“正常”基线，并采用以 2 为底的对数比率进行标准化分析。数据显示，本地开球时间是决定流量波动的首要因素：当比赛在午夜至清晨等原本低谷时段进行时，熬夜或早起观赛的用户会推动流量大幅激增甚至翻倍；而当比赛在白天进行时，由于用户本已在线，流量波动不明显；若在傍晚进行，放下设备专注观赛的行为反而会使常规网络活动降至基准以下约 70%。 最受关注的单场比赛与球队 通过对开球后两小时内各国绝对偏离度中位数的计算，阿根廷对阵瑞士的四分之一决赛以 1.26 倍的影响力位居首位，超越了法国对阵西班牙的半决赛。在各支球队的全球号召力方面，阿根廷以 1.17 倍的偏离度高居榜首，带动典型国家的网络流量产生约 17% 的波动。法国、巴西、葡萄牙、摩洛哥、西班牙及哈兰德领衔的挪威紧随其后；此外，海地与伊拉克因与强队交锋也成为高偏离度的特例。同时，博彩类网站访问量在赛事开幕后持续走高且日常波动趋于平缓。 两类截然相反的观赛互动行为 对比赛日流量曲线的聚类分析揭示了三种主要模式。最大群体（覆盖 44 国 101 场比赛）在补水暂停和中场休息时流量上升，反映出观众利用比赛间隙拿起手机的习惯；而第二类群体（如阿尔及利亚、突尼斯、埃及等 8 国）则在暂停时段出现流量回落。通过分析 MIME 媒体类型发现，第二类群体的高峰主要由多媒体流媒体驱动，观众倾向于在 15 分钟的中场休息时关闭直播流离开，导致流量暂时骤降，下半场开赛后又重新飙升。 微小暂停与决赛的全球印记 数据表明，即使是半场中间仅持续约三分钟的补水暂停，也能显著改变在线行为，多数国家在此期间迅速出现短暂的访问高峰。在 7 月 19 日的决赛及 7 月 18 日的季军赛期间，全球 HTTP 字节流清晰记录下了开球、补水、中场休息和终场哨响的每个节点；在阿根廷对阵西班牙的决赛期间，两国的流量水平相比对比时段激增了多达 20 个百分点。

For 96 years, the World Cup has been a global phenomenon, uniting nations and communities through a shared love of sportsmanship. While its popularity is nothing new, what is novel today is how rare a truly collective global experience has become. In an era defined by microtrends and algorithmic bubbles, it is increasingly uncommon for people across most countries to engage in the exact same event. That is precisely the unifying power of the World Cup. Fans from all over the globe reshape their daily routines around these once-in-a-lifetime matchups and storylines — and because Cloudflare operates a global network with 330+ points of presence worldwide, we are in a unique position to see exactly how this global ritual reshaped the world’s online activity throughout June and July 2026. Cloudflare Radar tracks HTTP traffic, DNS, security, and more to highlight global Internet trends. In this blog post we’ll use that data to explore how the World Cup impacted global traffic patterns throughout the tournament’s run. To understand how traffic changes throughout a match, we had to establish what it is “normally.” One way to do this is by looking at raw request volumes, or the amount of traffic we see on our network per country. But these amounts vary per country (the amount of daily traffic in the United States is always a larger number than the traffic in Portugal), which makes it difficult to establish a globally applicable baseline. Instead, we defined "normal" using the median traffic of the four preceding weeks: a month-long window that provided a stable, per-minute reference and smoothed out day-to-day noise. We also wanted to know whether traffic rose or fell relative to that baseline, but a plain difference wouldn't let us compare a high-volume country against a low-volume one. Instead, we used the ratio of current to baseline traffic, expressed as a log₂ value: the log makes increases and decreases symmetric around zero (+1 = twice normal, −1 = half). In other words, **a score of zero means traffic is perfectly normal, a positive number shows a spike, and a negative number shows a drop.** One factor shaping how traffic changes is simply what time the match kicks off locally. The largest changes in activity happen when a match is played in the overnight and early-morning hours — roughly midnight to 8am local time. These are the hours when very few people are normally online, so **fans staying up (or waking early) to watch push traffic well above its usual level, more than doubling it in some cases**. As the graph shows, this is where the deviation peaks on both workdays and weekends. By contrast, matches played during normal daytime and working hours — around 9 a.m. to mid-afternoon — don’t show such an impact: traffic stays close to its usual level, likely because the people watching would already have been online anyway. In the early evening there's a smaller, second lift, most visible on weekdays, as a match keeps people connected at a time when usage would normally start to wind down. Weekends follow a similar shape, with the strong early-morning rise but a gentler evening bump. The impact of kickoff time is easiest to see when comparing matches within a single country that take place at very different hours. Bosnia and Herzegovina provides a clear example. As seen in the graph shown above, when Bosnia played at 2 a.m. local time, people stayed awake to watch and traffic during the game jumped to well above its normal level, at times more than doubling. When Bosnia played in the evening, the opposite happened: traffic dipped below normal (falling to about 70% of typical value), as people put their devices aside and focused on the match itself. When Brazil played Japan in the Round of 32 (Brazil won 2–1 on June 29, 2026), the two countries watched the very same game 12 hours apart: kickoff in Houston (GMT−5) fell during normal waking hours in Rio de Janeiro (GMT−3), while in Tokyo (GMT+9) it landed in the dead of night. The result is two nearly parallel curves for the same 90 minutes: one higher than normal, one lower. Japan's traffic (red) sits well above normal, around +1, roughly double its usual level, because the match aired in the small hours, when almost no one would ordinarily be online. Brazil's traffic (green), by contrast, runs *below* normal, around −0.4, as the game fell in the middle of an ordinary active day. In this case, watching the match pulled people *away* from their usual browsing rather than adding to it. One of the most compelling aspects of the World Cup is seeing which storylines and teams capture the attention of fans across the world. We’ve discussed how regional traffic patterns change as a result of matches. But who are they watching? Which matches made the most impact on Internet traffic? Here's how we calculated this: for each match, we took the two-hour window after kickoff and, for every country with enough baseline traffic to give stable measurements (small, noisy markets are excluded), computed how far traffic strayed from normal. We then took the absolute value of each country's deviation, so we're measuring *how much* traffic changed, not in which direction (a surge and a drop both count as impact), and for each match we took the median of those absolute deviations across all countries. Because several group-stage matches were played simultaneously, making it impossible to attribute a country's traffic swing to one game or the other, we dropped those concurrent matches to avoid ambiguity. The result is this ranking of the matches that moved the Internet most, worldwide. And there's a surprise: **the very top spot wasn’t snagged by a final or semifinal. It was Argentina vs. Switzerland on July 11, a quarterfinal that saw Argentina win 3-1 — and that moved Internet traffic by a factor of about 1.26.** That put it ahead of the France vs. Spain semifinal, which had a factor of 1.21. The rest of the top matches were a mix of quarterfinals, round-of-16 and even round-of-32 ties. To decide which team the world watched most, we looked at each team's matches and aggregated the median worldwide impact across all countries. In other words, when a given team took the field, how much did the typical country's traffic move away from normal? Not surprisingly, **Argentina topped the list at 1.17x, meaning that when Argentina played, the typical country's traffic swung about 17% away from its normal level, the strongest global pull of any team.** This comes as no surprise, since they were the defending champions and each knockout game could have been Lionel Messi's last dance for his national team. Love them or hate them, people were watching them. Not far behind were nations packed with superstars such as France, Brazil, Portugal, Morocco, Spain — and Norway, fueled by the Erling Haaland phenomenon. Haiti and Iraq appear in the top as outliers due to their high deviation scores relative to their typical traffic, suggesting matches against major teams drove disproportionate engagement. Compared to HTTP request data in the month preceding the World Cup, there was an overall increase in requests to gambling industry websites since the opening game. Additionally, whereas pre-tournament traffic followed a clear weekly pattern, after the Cup’s opening game, the trend flattened into a more constant profile, likely a consequence of the high, near-daily regularity of matches. Because Cloudflare is present in 120+ countries and handles traffic from Internet users worldwide, we can see distinct behavioral patterns across the globe. For example, when examining the deviation trends during the Algeria vs. Austria group stage game on June 28, we noticed something peculiar: Austria’s traffic (in red) *increased* during halftime, while Algeria's (in green) decreased. The former follows the pattern described above of people spending more time online while not watching the game, while Algeria’s is the complete opposite — and they’re not the only ones. *Algeria, in green and denoted as DZ, saw a much higher uptick in Internet traffic during the match than Austria, in red.* To understand patterns in behavior across countries we grouped every country's match-day behavior by the shape of its traffic curve and let the patterns cluster together. Grouping match-day traffic shapes this way, three distinct patterns emerge. The largest group (44 countries playing 101 matches) shows Internet usage rising during hydration breaks and halftime, the natural pauses in play, as people reach for their phones. A second, smaller group (8 countries playing 18 matches)) is its near mirror image: traffic falls at exactly those same moments, dipping during the breaks instead of climbing. The third group is a clear outlier, made up entirely of Iran's three matches. The explanation is simple: the May baseline was measured while Iran was still coming back online after the shutdown, so its match-day traffic sits far above that depressed reference, producing a deviation unlike any other country's. You can read more about Iran’s Internet shutdowns and partial restoration throughout 2026 [on our blog](https://blog.cloudflare.com/tag/internet-shutdown/). To better understand the second cluster, which included Algeria, Tunisia, Jordan, Egypt and DR Congo, we looked more closely at the traffic mix for these countries. We broke down traffic patterns by Multipurpose Internet Mail Extensions, or MIME type, and grouped it in families to easily distinguish clusters of content types. MIME types act like digital labels that tell browsers exactly what kind of file they are receiving, whether it's an HTML page, a JPEG image, or an MP4 video stream. By tracking these labels, we can infer what kinds of content users are consuming. Our hypothesis was that this behavior could be explained by a disproportionate amount of people watching the games via streaming in those countries. To test this, we compared traffic pattern distribution in games with teams of both clusters. In the following example, we see traffic distribution of Algeria and Austria respectively in the match between both countries. *In Algeria, traffic was far above normal, then dipped at halftime. Note the large increase in streaming traffic, in orange.* *In Austria, where streaming services were used less, Internet traffic increased at halftime.* In the Algeria graph above, we can see that the bulk of the increase during the match window indeed was driven by requests to multimedia and streaming services. This supports our hypothesis that the traffic trendlines correlate with use of streaming to watch the match. In Algeria, traffic rose sharply at kickoff, dropped during half-time, and returned to elevated levels once the second half began. Hydration breaks, by contrast, had little to no visible effect, which suggests that viewers don't meaningfully change their Internet or social behavior for short, in-play pauses, but do so during the longer halftime interval. Other countries in this cluster show similar behavior. This might be because a viewer is unlikely to close a stream for a three-minute cooling break, but a fifteen-minute halftime is long enough to close the stream and step away. A minority of countries, including Tunisia and Algeria, disconnect during halftime, with traffic dropping below its in-play level (the blue boxes, sitting under the 1.0 line). The majority of countries go the other way: traffic rises during the break as people pick up their phones the moment play stops, then settles again when the second half begins. Croatia and Bosnia and Herzegovina show this most strongly, with halftime traffic running well above their in-play baseline. Halftime is a long, familiar pause, but what about the much shorter hydration breaks? These last only about three minutes, taken midway through each half. Is three minutes really enough to change how people behave online? It turns out it is. Just as at halftime, the moment play briefly stops, traffic in most countries ticks up before falling again when the game resumes. We measured this by taking, for each match, the peak number of requests in a window around the middle of each half (where the hydration break falls) and comparing it to the five minutes immediately before the break. The pattern lines up with everything we've seen so far: the same audiences that surge at halftime also spike during these brief pauses, while the few countries that tend to disconnect during breaks show little or no lift. Even a three-minute gap in play is enough for a large share of viewers to glance back at their phones. At the scale of Radar’s global HTTP requests, it is genuinely hard for any single event to leave a visible mark. Even so, the 2026 World Cup Final, which pitted Europe’s and American football champions against each other, had enough social impact to affect the Internet’s footprint. When looking at the volume of HTTP bytes from July 18 to July 20 we can immediately identify the final match kicking off at 21:00 UTC on July 19, as well as England vs. France for the Bronze medal at 23:00 UTC on July 18. During the final match, Argentina and Spain's traffic volume increased up to 20 percentage points when compared to a similar period. The bronze medal match also coincided with a traffic increase, although at a smaller scale. HTTP request volume during the final match also appeared correlated not only with the timeline of the game, but with each individual stage as well. Looking at the graph, we can roughly pinpoint moments such as the kickoff, halftime break, hydration breaks, as well as the final whistle. Across every time zone, match, and goal, Cloudflare Radar provides a front-row seat to how the world connects during landmark cultural moments. To explore more interactive traffic insights and track how major worldwide events shape internet activity every day, visit [Cloudflare Radar](https://radar.cloudflare.com/) or follow us on social media at [@CloudflareRadar](https://twitter.com/CloudflareRadar) (X), [https://noc.social/@cloudflareradar](https://noc.social/@cloudflareradar) (Mastodon), and [radar.cloudflare.com](http://radar.cloudflare.com) (Bluesky).