The Mysterious World of Seeding Committees

Truths, Lies and Tennis Statistics / Matthew Walton / 23 August 2007 / Leave a Comment

"Magical" Matthew Walton on how seedings are determined and how much us punters should read into them

The work of Seeding Committees is a complex matter, involving the delicate combination of several key criteria. A fundamental knowledge of form, an accurate assessment of ability, an idea as to potential improvement and an overall perspective as to how a particular tournament will develop on account of their opinions.

And us, as backers, have to assess the same factors when we approach any tournament ourselves.

But ask yourself this question. How often do the Seeding Committees get it right and how often do you get it right? Answer: Not as often as you'd like!

Now, it's simplistic to expect the No.1 seed to win every tournament but the stats show that Seeding Committees don't even get close to the mark.

There have been 49 events this year on the ATP Tour and of these the top eight seeds have the following combined records :-

No.1 seed - 9 tournament wins
No.2 seed - 12 tournament wins
No.3 seed - 9 tournament wins
No.4 seed - 3 tournament wins
No.5 seed - 3 tournament wins
No.6 seed - 1 tournament win
No.7 seed - 0 tournament wins
No.8 seed - 1 tournament win

That makes a total of 38 tour wins by the seeds No's 1-8 (a healthy 77%) but it still means that 11 tournaments (23%) have been won by unseeded players.

Only one event (Acapulco) has produced the Holy Grail of all eight seeds making the QF's and only one event (Memphis) has seen the four top seeds making the SF's.

The No.1 seed has beaten the No.2 seed in only two finals this year (Hamburg and Wimbledon) where both Roger Federer and Rafael Nadal were involved.

The two most common reasons for this are (1) when Seeding Committees strictly apply official ranking positions to produce their own seedings and take no account of the surface on which their event is being played or (2) the opposite, where tournament officials use 'surface form' as opposed to official rankings.

However, other variables often get thrown into the mix such as previous form in the event, home-bias to give local fans the chance of seeing their man progress through the event (the Henman effect) or instances where players returning from injury are either overlooked or fast-tracked depending upon the whim of 'the Committee'.

What results is a mish-mash of seedings which combines to make a blind faith in their guidance extremely risky.

We can see from the figures above that Seeding Committees are no more accurate than racing handicappers - when was the last time you saw all 16 runners in a sprint handicap finish in a straight line across the track?

The simple fact of the matter is that seedings provide little more than a rough guide as to the way in which an event will develop. If anything, as a backer on the exchanges, you should watch out for any mistakes in the market which assume that 'Player X is seeded so he should be the favourite to win his match'.

Just to confirm this view, in the last five tour events (Indianapolis, Sopot, Washington, Montreal and Cincinnati) take a look at the cumulative performance of the top 8 seeds in each tournament. That's 40 players in total.

Of those 40 players, 21 lost in the first or second round when they were all facing non-seeded opponents. That's 52%. Just consider what prices their opponents were and you can see how great the potential profit is when you take on the Seeding Committees.

Next article 'Calm down! It's only a break of serve'.