Terrific, ok, so with R128 out, Fed's numbers are closer to his Masters numbers, which were:
I was terribly busy again today, but I finally had time to run some numbers.
There are several things we may want to consider:
1. When we look at the first round of any group of tourneys, included in that first round will be a lot of early losses also early in careers. Looking at all tourneys, Fed lost 7 times in R128. 6 of those times were at major - where else would be find R128? But it turns out that somehow he lost at Miami in R128. I didn't know that this tournament ever HAD a Round 128. BUT: all of those losses were June of 2003. He never lost in that round again, ever.
2. The same thing happens in R64. At the majors this is round 2, so when we look at this round Fed has only one loss in majors, to:
Sergiy Stakhovsky [UKR] d. Federer
That's a famous loss, so everyone know about it.
But because this was round 2, apparently by the second round he built up some steam and never lost when he was young.
In contrast, he has a LOT of losses early in his career in R64 at masters. But only one loss in 2007, at IW. Another at Cincinnati in 2004. That's it in masters. But SEVENTEEN losses in R64 before 2003. Now, R64 is not a round. It's a level.
3. R32
At this point we are well into big tournaments but only in the 1st round of smaller tourneys. We can expect some losses in majors, early. By R32 a young player is hitting his limit. He washed out 3 times in 2001 and 2000. But after that only twice, once in 2015, AO, Seppi. And Kuerten got him at RG, 2004.
In contrast, he's gone out quite a few times in R32 at M1000s. Wouldn't we expect a better record here than in masters? But more seeds may still be in M1000s at this point.
Now, here is the kicker. R32 in M500s and M250s. Seems like it should be a lock. Smaller fields, fewer top players. Pretty much true, only one loss in Dubai to Murray, 2008. 12 more losses, through to 2003, mostly in very early years. That's when he would have been breaking into the ATP.