Anyways, Raonic has many of the qualities Sampras had serving-wise: pace, placement, disguise, HUGE second serve (110 MPH on average in his first round match). His ace to double fault ratio this year is better than any of Sampras's years. He also gets a higher percentage of first serves in. Also, he leads the tour in just about every significant serving category.
Also, it's at leasaat SOMEWHAT impressive that his stats are better despite being in an era of homogenized surfaces, and with the rest of his game being inferior (by a lot) to Sampras's.
I hate to sound like a broken record, but the ace and DF counts can be very misleading, especially when we're comparing the contemporary stats to the previous eras'. This is particularly true for Sampras, whose serve is thought by some (my former self included) clearly inferior as a pure shot to that of a Goran, Karlovic or Krajicek, since his # of aces tends to be lower.
But that impression changes when you look at the % of unreturned serves (which basically means all the free points off one's serve). Here Pete is right up there with those guys, usually winning over 40%, often 45%, sometimes even 50%. You might have noticed some of these stats on the "Greatest Serves of All Time" thread. (It's a shame they aren't widely available.)
And it's a common fallacy to think the fewer DFs the better. Yes, you'll lose a few more points if you DF more, but then your 2nd serve will have more bite and will increase your chances of winning the rest of your 2nd-serve points, if not win a few more points outright in case of a Sampras or Raonic. If anything today's players are too afraid to DF.
If you're interested I once elaborated on this very topic here:
http://tt.tennis-warehouse.com/showthread.php?p=5968374#post5968374
1st-serve % doesn't favor Raonic a whole lot. He's been averaging 61% this year, just two points over Pete's 59% career average (this number is most likely off, like so many of the ATP stats from the '90s, but I'm fairly certain it isn't off by much as he used to average about 60%). And then there are the issues of strategy, mentality and technology to consider. Again you can follow these links to see what I've written on these very topics:
http://tt.tennis-warehouse.com/showthread.php?p=5267685#post5267685
http://tt.tennis-warehouse.com/showthread.php?p=6694094#post6694094
http://tt.tennis-warehouse.com/showthread.php?p=6842225#post6842225
If time is short at least try the last one, as it contains a few stats related to the surface issue. (That post has more to do with the supposed slowdown than homogenization, but the underlying idea is the same.)
Anyway I agree with you that Raonic's serve is probably one of the best ever already. What intrigues me more is whether he'll surpass the likes of Karlovic and Isner as a tennis player.
Sampras's 1st serve averaged 2500 RPM. Fact.
Again I hate to repeat myself, but it's very debatable whether this "fact" is based on a satisfactory number of samples. People really should be careful not to take these numbers as some sort of gospel. They're meant to serve as a starting point, like all the stats are.
That said, yes, it's quite probable that Pete's serve had an unusual amount of spin. But what really made it so tough for the returner was its topspin component--35% according to Yandell, much higher than that of other servers he measured, and which explains why it looked so "heavy" (Yandell compares it to Roddick's in his article).
BTW I wouldn't be surprised that Raonic's serve isn't too far behind in both RPM and topspin rate. He's given Fed fits in their meetings so far, the same Fed that has made many a big server's life miserable (Roddick should know). In fact their Halle match reminded me of the very Yandell article. Hopefully he'll get to Raonic's serve someday.