Roger Federer: Would You Back Him to Win at Indian Wells?

I wish I had a crystal ball to predict the results at Indian Wells. Or that I owned Paul the Octopus who so accurately predicted the results of the World Cup! No, I take it back. Owning Paul would have been no use as the most famous octopus in the world died last October.

So, what are the tools we have to predict the results (apart from wishful thinking)?

We have the ATP ranking system which says Nadal is the man to beat. He is No.1 and sitting atop a huge pile of points.

The year-to-date ranking backs Djokovic to win. He won the Australian Open and is unbeaten this year.

If you rely on the age model you would say a young super-talented player aged between 23 and 25 is more likely to win than Federer. So it could be Nadal or Djokovic or Murray.

Just when you are about to revert to the thought that only wishful thinking could predict Federer as the winner, you hear of a guy called Jeff Sackmann who has evolved what he claims is a superior ranking system for prediction.

According to this system detailed in "Indian Wells Projections," Federer is the clear favorite. This model takes two years' results into account, adjusts for surfaces, gives additional weight to recent results, some small additional weight to slams, adjusts for quality of opponents, takes the draw as well as players' head-to-head records into account and so on.

Apparently, when back-tested for the past two years' results, it is 69.5 percent accurate, which is stated to be four percentage points better than the accuracy of the ATP ranking system. However, the real test of its efficacy comes when we use it to predict future results. So Indian Wells is a test to start with.

Here's the summary of predictions as per the above tools:

ATP ranking system: Rafael Nadal

Year-to-date Ranking: Novak Djokovic

Age-Performance model: Not Federer

Jeff Sackmann's ranking system:

Roger Federer (25 percent chance), Novak Djokovic (15 percent chance), Rafael Nadal (13 percent chance) and Andy Murray (11 percent chance).