Bouncebackability - fact or fiction?

General / Barry Millns / 16 August 2007 / Leave a Comment

"Magical" Matthew Walton takes a look at which players come back from the dead when going a set down and which lie down and die

In the recent final of the Rogers Masters, Roger Federer fought back from a set down to level the match with Novak Djokovic. At that point he regained favouritism on the exchanges and most punters would have expected him to go on and win the match.

And why not? He had won all four previous matches against the Serb.

However, one set of stats suggested that despite this fightback, Federer should still have been the overwhelming favourite to lose the match, not to win it.

In 2007, the world No.1 has played 46 matches - won 40, lost 6 - and has dropped the first set on 8 occasions (17%).

Of those 8 matches the Swiss player has fought back to win just twice but has gone on to lose 6 times (75%).

Even though Federer levelled the match against Djokovic in Montreal, giving him the momentum going into the deciding set and making him the favourite on the exchanges, the stats say the market had it wrong.

Furthermore, the market even had it wrong when you consider those matches when Federer loses the first set and then wins the second to level the match. That has happened 5 times this year and on 3 occasions the world No.1 has still lost the match (60%).

Both sets of stats, those when Federer loses the first set as well as those when he loses the first set but wins the second, marked him down as the favourite to lose not to win but did the exchanges reflect such findings? No, they didn't.

Of course, common sense tells us that any player who loses the first set is the most likely to end up losing the match but, highlighted by a case like Federer, it pays you to stay with a match beyond the first set. One player or another may take the lead but that, in itself, doesn't spell the end of decent betting opportunities.

It is open to more debate how much credence we should give to a player's respective opponents in these matches.

For example, Federer's only two comeback wins this year were to defeat Carlos Moya (not wholely unexpected) before returning from the brink to defeat Rafael Nadal (much harder to predict). Nadal, himself, during this year has hit back to claim wins over quality players such as Andy Murray and Lleyton Hewitt but then failed to do so against much weaker opponents like Nicolas Mahut and Fernando Gonzalez.

The best performers, based on match results when losing the first set, are :-

Rafael Nadal (won 7, lost 7 = 50%)
Andy Murray (won 6, lost 6 = 50%)
Paul-Henri Mathieu (won 8, lost 8 = 50%)
Lleyton Hewitt (won 6, lost 8 = 43)
Ivan Ljubicic (won 10, lost 14 = 42%)

Such bouncebackability is to be remembered and it's no surprise that both Nadal and Hewitt were rated as the first and second best fighters in our recent guide to the tour's toughest competitors (also in Montreal, Mathieu beat Guillermo Canas from 4-6 0-4 down!).

Richard Gasquet is another interesting case. The Frenchman was pegged as something of a flake in our last piece and his pathetic record (won 1, lost 14 = 7%) marks him down as more of a lover than a fighter. In fact, his only win was that stellar comeback against Andy Roddick in the Wimbledon QF's.

This raises the question that some players might not be worth backing before the match, rather wait for the first set to be completed and then apply the stats accordingly.

In short, the knee-jerk mentality of many backers to the losing of sets (even the breaking of serve) should be tempered with a more long-term familiarity with the stats and an understanding that the exchanges don't always reflect the true picture of how a match will ultimately unfold.

Next week's piece 'The Mysterious World of Seeding Committees'.