Pondering some numbers.
Disclaimer, this is a bit rubbery with data available, I’m making a few assumptions.
Adam Gilchrist played 96 tests, fielded in 191 innings, 379 catches. Let’s assume for simplicity that all those innings were completed. (There probably wasn’t too many innings defeats Australia had in that era, but there would have been draws etc where 20 wickets weren’t taken).
379 catches from 191 innings is 2 catches per innings, 4 per test. The article below had the best keepers (who played a significant volume of games) dropping 10% of chances. Gilly was 12%, So, in rough numbers, he got 4.5 chances per test, and caught 4 of them. Or, more elegantly, 9 chances every 2 tests and took 8 catches.
There was another band of keepers who were around the 20%-22% range. So, those players, if they were presented the same 9 chances in 2 tests, would catch ~7 of them.
So, let’s say Foakes drops 10% and Smith drops 20%. (I’m making those numbers up). The difference is roughly one catch per 2 tests. A dodgy google search suggests the long term test batting average is 30. Assuming drops are evenly distributed through an innings, on average each drop would cost 15 runs. Worst case, if the dropped catch occurred when the batter is on 0, each drop would be worth 30 runs.
So, on average, the extra drop per 2 tests that a 20% keeper has over a 10% keeper is worth 7.5 runs a test. If Smith was a 30% keeper, it would cost the team 15 runs per test. Not as much as I’d have guessed, however, there is also the impact of team mate morale, and presumably a similar increase in byes. I think there is also an impact whereby the runs scored by the not out batsman are increased too, but too hard basket to work that out.
TLDR, I’d pick the better keeper still.
(This all depends a lot on conditions - quicks have a much lower drop rate than spinners, so sub continental keepers are going to suffer from more drops due to greater volume of spin bowling they have to keep to. Off spin drop rates are more like 30% apparently.)
