More on first and second serves

In my previous post, I said "I have never, ever watched any player who doesn't use different strategies for first and second serves", implying that all tennis players serve their second differently from their first.

Well, I was wrong. According to an article from Tennis SET (SET stands for science, engineering and technology), the young Venus Williams "basically used the same serve every time" in her first US Open in 1997. (Still, on the whole, Venus Williams serves a slower, safer second serve - see this example)

Another article from Tennis SET analyses the 2004 Wimbledom game between Jan-Michael Gambill and Sebastien Grosjean:

Gambill's first serve was working well but his second serve let him down. Gambill served 15 aces, he got 58% of his first serves into play, and he won 79% of his first serves when they went in. The problem was he won only 14 of his 45 second serves. What he should have done was to serve two first serves instead of one first and one second serve. That way he would have got 58% of the 45 second serves or 26 good second serves into play. He would have served 45 - 26 = 19 double faults instead of only 5, but he would have won 79% of those 26 second serves or 21 points instead of only 14.

The article concludes that:

The tactic of serving two first serves works best when the server is winning less than half the second serves and more than 70% of good first serves. There were at least 33 matches and possibly as many as 60 matches at Wimbledon this year where two first serves would have worked better than one first and one second.

The same reasoning would probably suggests that for some players, serving two second serves may be the best strategy.