I completely agree with the article and disagree with most of the posters here. I have noticed the same trend with Roger since the beginning of his career.
This is a positive not negative thing, and has nothing to do with how he performs in the clutch (indeed various indicators of his performance in tight points are amongst the highest on the ATP tour, for instance tie break records).
It does on the contrary have to do with statistical variance. In other words, an inferior player who adopts a risky game plan (example, an Ivo Karlovic who ignores all return games and only holds serve to try to force a tiebreak where he goes for outright winners on all returns) will lose far more playing that strategy against standard ATP tour players, except when he is up against the elite players where that 5-10% locked in winning percentage is actually better than what his normal matchup would entail.
So of course you forget the many games where he is completely blown out by Roger, but only focus on the one time where he managed to win 3 tiebreaks in a row.
Indeed when you look at the great losses in Rogers career, they are almost universally performed by players who are having career games. You see, its wrong to say oh yes well he lost a close game, that proves he is not clutch. WHen you should be saying, oh wow, Roger actually made that a close match, where most players would have been blown out playing a guy on a hotstreak.
