By a well-known property of the
incenter I (see page 1 of Mathematical
Excalibur, vol. 11, no. 2), we have AC1
= C1I and AB1 = B1I. Hence, ∆AC1B1 ≅
∆IC1B1. Similarly, ∆BA1C1 ≅ ∆IA1C1
and ∆CB1A1 ≅ ∆IB1A1. Letting [⋯]
denote area, we have
[AB1CA1BC1] = 2[A1B1C1].
If ∆ABC is not acute, say ∠ BAC is
not acute, then