SUDDEN LATENCY INCREASES
Online Lag Spike Test
Monitor response times and flag only substantial increases against preceding successful readings. Review when each detected spike occurred and how far it rose above its observed baseline.
INTERACTIVE SPIKE DETECTOR
Watch for sudden response-time jumps
Loading PingWatch endpoints…
Round-trip timing to the selected PingWatch server.
Latency and detected spikes
Orange markers identify detector events · response time in milliseconds
Waiting for the first measurement
The chart and event list will use only measurements collected during this test.
Spike events
Events are listed in the order detected.
- No detected spikes yet At least five successful readings are needed to establish a baseline. Small fluctuations do not automatically count as spikes.
Results are preliminary while the test is running. No detected spikes is not proof that the connection is perfectly stable.
What counts as a detected spike?
PingWatch uses its existing detector: after at least five preceding successful readings, it compares a new latency reading with the median of its recent baseline. A spike is flagged when it is at least 50 ms above baseline and at least twice the baseline. Consecutive readings in the same elevated incident do not create repeated events. These are product-defined detection heuristics, not a universal networking standard.
Understand the limits
Measurements are browser-to-PingWatch endpoint HTTP or WebSocket round trips. This is not an ICMP ping test and does not measure latency to specific game servers. Browser scheduling, local device load, and remote endpoint performance can affect results; the readings cannot alone establish the cause of instability. Failed probes are application-level monitoring failures, not confirmation of packet loss.
For troubleshooting steps, read the guide to diagnosing an unstable internet connection. You can also run the full Ping Test or the interactive Jitter Test.