This Gauge32 reports FallbackDelay, calculated as the average one-way delay from probe round trips (round trip sum divided by number of probes, divided by two) plus whichever of incoming or outgoing jitter is higher. It gives an administrator a composite estimate of the delay a real call would experience on a fallback path, factoring in both latency and jitter. For example, a high FallbackDelay value would warn the admin that even though a fallback path is reachable, users would likely perceive noticeable lag if calls were routed there.