Attention behaves less like an on/off switch and more like a limited resource that gets allocated, depleted, and recovered. Sustained attention on a demanding task shows measurable declines over time (vigilance decrement) well before a person consciously feels tired, and attentional capacity is affected by the same factors that affect willpower-adjacent processes generally: sleep, stress, hunger, and cognitive load from unrelated sources.
The practical implication is that 'just focus harder' has real physiological limits, not just motivational ones. Structuring demanding cognitive work around attention's actual patterns — front-loading hard tasks when attentional resources are freshest, building in real breaks rather than pushing through a decline — produces more total output than trying to sustain peak focus for an extended, unbroken stretch.
During World War II, radar operators watching for enemy activity showed a puzzling pattern: their detection accuracy dropped substantially within the first 30 minutes of a watch, well before anyone reported feeling tired. Mackworth built a simplified lab version — the "Clock Test," a clock hand that made occasional irregular double-jumps participants had to spot — and found the same decline reliably reproduced: sustained attention on a monitoring task degrades measurably within half an hour, even in alert, motivated, well-rested people who had no reason yet to feel fatigued.