Kuiper, a mobile game created by Borja de la Osa, became unreachable from a Vodafone connection in Zaragoza during a LaLiga-related IP block on September 5, 2026. The game’s Cloudflare Worker resolved normally, but a TCP connection to 188.114.96.5:443 timed out after 20 seconds. A Hetzner reverse proxy in Germany then returned HTTP 200 in 0.2 seconds—a sharp demonstration that the failed path was the network route, not a general game-server outage.
The concrete Kuiper case
Kuiper’s backend runs on a Cloudflare Worker using the shared range 188.114.96.0/22. Its monitoring data recorded blocks affecting that range on 12 distinct days between August 15 and September 6, 2026, across DIGI, MasMóvil, Movistar, Orange and Vodafone.
The September 5 test from Zaragoza helps separate the failure stages. DNS resolution returned the expected Cloudflare addresses through the operator’s resolver, 1.1.1.1 and 8.8.8.8. The connection failed later, when the Vodafone path attempted to reach 188.114.96.5:443.
That distinction matters: changing a DNS resolver would not address the documented failure, because the name already resolved. The timeout happened at the TCP connection stage.
Why shared IP blocks affect unrelated services
An IP address is a network destination, not necessarily a single website. Services delivered through Cloudflare can share address space while remaining separate at the domain and application layers. Blocking an IP range can therefore cut off unrelated traffic that reaches the same infrastructure.
That is the practical problem exposed by Kuiper. LaLiga’s enforcement aims at IP addresses associated with unauthorized access to its football content, but an IP-level measure can also affect a legitimate Cloudflare Worker when both services occupy the same shared infrastructure. The block does not need to identify Kuiper as a target for Kuiper’s connection to fail.
OONI’s measurements provide broader context. During its January 1–June 1, 2026 study period, the organization found that more than 500,000 domains in its test population were affected at some point by LaLiga-related IP blocking. That historical measurement describes the study period; it is not a live count of services blocked on September 12.
The German fallback that restored access
Kuiper’s client waits five seconds for the Cloudflare Worker. If the primary request does not respond, it routes traffic through a Hetzner CX22 VPS in Germany. The VPS acts as a reverse proxy: it forwards requests to the Worker and does not store game data.
During the documented September 5 test, the direct Vodafone path timed out after 20 seconds, while the German proxy returned HTTP 200 in 0.2 seconds.
| Path | Connection context | Observed result | Trigger or function | Cost |
| Direct Cloudflare Worker | Vodafone connection in Zaragoza, Spain | TCP timeout after 20 seconds to 188.114.96.5:443 | Primary game-backend path | Not stated |
| German reverse proxy | Hetzner CX22 VPS forwarding traffic to the Worker | HTTP 200 in 0.2 seconds | Activated after a five-second primary-request timeout | €4 per month |
The fallback is a small but revealing engineering response. A €4-per-month server became the bridge between Spanish players and a service whose underlying Worker was still reachable through another network path. It does not prove that every user, operator or match-day block behaves identically, but it shows how shared infrastructure can turn an anti-piracy measure into an application-availability problem.
LaLiga’s position and the operator response
LaLiga says its measures target IP addresses used to provide illegal access to its content and rejects the characterization of the action as broad or indiscriminate. That is the league’s stated position.
Movistar support gave a different piece of the picture in response to a customer reporting access problems. The representative said the disruption appeared related to blocks applied under a LaLiga-linked judicial resolution and warned that shared Cloudflare ranges could be affected. Movistar also said it could not create individual exceptions.
Those statements describe the intended target and the operator’s explanation. Kuiper’s measurements document the observed consequence: a legitimate game’s Cloudflare-backed path became unreachable from a Spanish connection while an alternate route succeeded. The technical dispute is therefore not about whether shared infrastructure exists; it is about whether blocking at that layer produces unacceptable collateral effects.
The same issue reaches beyond games. Users have reported failures involving game authentication, Docker image pulls, self-hosted services and development tools during football match windows. Those accounts illustrate the practical stakes, while Kuiper’s dated connection test supplies the clearest controlled example in this case.
The bottom line
Kuiper’s September 5 test shows exactly how LaLiga-related IP blocking can disrupt a legitimate service without a DNS failure or a general backend outage. The Spanish route timed out; the German reverse proxy worked. As long as multiple services share address space, an enforcement system aimed at an IP can affect more than the service it intends to reach—and developers may be left paying to build the detour.
