按 \(x\) 和 \(y\) 是从区间中选取,还是从 \(0\) 与 \(1\) 中选取来分类。由于这取决于两次抛硬币,每种情况发生的概率都是 \(\frac{1}{4}\)。
情况 \(1: x\) 和 \(y\) 都是 \(0\) 或 \(1\)
\(x\) 和 \(y\) 必须不同,发生概率为 \(\frac{1}{2}\)。
情况 \(2: x\) 是 \(0\) 或 \(1\),而 \(y\) 从 \([0, 1]\) 中选取
若 \(x = 0\),则 \(y\) 必须从 \(\left(\dfrac{1}{2}, 1\right]\) 中选取;若 \(x = 1\),则 \(y\) 必须从 \(\left[0, \dfrac{1}{2}\right)\) 中选取。
所以 \(y\) 从正确区间中选出的概率总是 \(\frac{1}{2}\)。
情况 \(3: x\) 从 \([0, 1]\) 中选取,而 \(y\) 是 \(0\) 或 \(1\)
由对称性,这种情况的概率与情况 \(2\) 相同。
情况 \(4: x\) 和 \(y\) 都从 \([0, 1]\) 中选取
由于要考察无穷多个 \((x, y)\) 数对,可以使用几何概率。画出 \(|x - y| \gt \frac{1}{2}\) 的区域。

阴影面积占整个图形的 \(\frac{1}{4}\),所以这种情况成功的概率为 \(\frac{1}{4}\)。
把四个加权概率相加,得到 \[\frac14\left(\frac12+\frac12+\frac12+\frac14\right)=\frac7{16}\text{。}\]
所以正确答案是 B。
English
We can case on whether \(x\) and \(y\) are chosen from the interval or from \(0\) and \(1.\) Each case has a \(\frac{1}{4}\) chance of happening, since they depend on two coin flips.
Case \(1: x\) and \(y\) are either \(0\) or \(1\)
\(x\) and \(y\) need to be different, which happens with a \(\frac{1}{2}\) probability.
Case \(2: x\) is either \(0\) or \(1,\) and \(y\) is chosen from \([0, 1]\)
If \(x = 0,\) then \(y\) has to be chosen from \(\left(\dfrac{1}{2}, 1\right],\) and if \(x = 1,\) then \(y\) has to be chosen from \(\left[0, \dfrac{1}{2}\right).\)
This means that \(y\) always has a \(\frac{1}{2}\) probability of being chosen from the correct interval.
Case \(3: x\) is chosen from \([0, 1],\) and \(y\) is either \(0\) or \(1\)
This has the same probability as case \(2\) due to symmetry.
Case \(4: x\) and \(y\) are chosen from \([0, 1]\)
We can use geometric probability since we are working with an infinite number of \((x, y)\) pairs. We graph \(|x - y| \gt \frac{1}{2}.\)

The shaded area covers \(\frac{1}{4}\) of the graph, showing that there is a \(\frac{1}{4}\) probability of this case working.
Adding the four weighted probabilities gives \[\frac14\left(\frac12+\frac12+\frac12+\frac14\right)=\frac7{16}.\]
Thus, B is the correct answer.