Chang d. Lendl, 4-6, 4-6, 6-3, 6-3, 6-3
Statistically what I find most interesting is the H2H. This was their first meeting. Lendl ended up winning all 5 of their best-two-of-three matches, all in straight sets. Chang won both of their best-three-of-five matches, each of them after nearly being swept in straights (1989 RG, 1991 GS Cup).
That's according to their ATP page, anyway.
Washington Post and Los Angeles Times:
Would he have won if Chang hadn't had the cramps? "That we will never know," said Lendl, who had 45 unforced errors.
ESPN gave Lendl a lot more unforced errors (83 as of 4-3 in the final set). I don't know why there's a discrepancy, and such a large one.
As of 5-2 in the fourth, Lendl had 73 unforced errors, Chang 48.
Earlier, Barry Tompkins said that Lendl was up to 63 ue's, and he broke them down as 43 FH, 20 BH. He didn't make any mention of Lendl's df's, so it's unclear whether ESPN was including df's in their UE totals.
In the 1990s on Google News, I've turned up a lot of phrases such "___ unforced errors, including ___ double-faults." I've hardly seen any such phrases in articles from the 80s. The earliest I've found so far is from July 1987, at the U.S. Clay-Court Championships, from the Chicago Tribune:
The unseeded Kuhnen, ranked only 131st in the world, made 42 unforced errors--including four double-faults on serves--to just 15 for Carlsson.