Thu, Aug 18, 2011 - Page 18 News List

Actuary explains how rankings work

COMMON SENSE:After decades of what seemed like unfair cricket rankings, the ICC has now adopted a system of weighted results over the previous three to four years

Reuters, LONDON

England players celebrate after the dismissal of India’s Shanthakumaran Sreesanth, defeating India by an innings and 242 runs in the third cricket Test match at Edgbaston cricket ground in Birmingham on Saturday. The series win propelled them into the No. 1 position of the International Cricket Council rankings.

Photo: Reuters

England headed to The Oval for the final Test of the summer against India yesterday as the world’s top side courtesy of a rankings system which is now widely accepted but not generally understood.

The International Cricket Council (ICC) rankings are the brainchild of former Wisden Cricketers’ Almanac editor Matthew Engel.

“Test cricket crucially depends on context,” Engel wrote in the 1995 edition. “It needs a five-Test series (six is too long) for the personalities to emerge and the battle to capture the public imagination. These half-hearted one-off tests rarely work.”

Engel’s proposal, which was adopted by the ICC, awarded points on the basis of home and away series between each of the Test-playing nations.

It was replaced in 2003 by the present system designed and introduced by English actuary, scorer and cricket statistician David Kendix.

“What you had, unfortunately, was a situation some years ago when South Africa were top of the table and yet it was clear intuitively to anyone following the game that Australia was the No. 1 team,” Kendix said this week.

“Indeed, Australia had just beaten South Africa comprehensively home and away. This was clearly not sustainable,” he added.

“And the reason you had that situation was that South Africa at that point had just beaten Bangladesh and Zimbabwe home and away, whereas Australia had not recently played them at all,” he said.

“So you basically had four series victories for South Africa which counted for just as much as victories over the likes of England, India or Pakistan. Since Australia happened not to have the played the two weakest teams at the time, their series points were only enough for second place in the rankings. At that point it became difficult to believe it was a fair reflection of the relative strength of the teams,” Kendix said.

“Matthew’s advocacy of Test rankings was great and I have huge admiration for him as he started the process. He had the great idea of trying to establish a league table that gives that context to test cricket and I said to him soon after my system was adopted by the ICC, how many prototypes of anything ends up being the final version? He had the idea of having a ranking system, the ICC recognized that this was something they ought to make official, but they also realized that to be accepted, it needed to give sensible results all the time,” he added.

Kendix’s system awards points for individual matches in a series as well as a bonus for the series winners. The model is based on results over the previous three to four years, weighted to give greater relevance to more recent results.

“That reflects the reality that in a Test series people are looking to win each match and they are also looking to win the series. I wanted a points mechanism that reflected the aims of the competing teams and what they are actually trying to achieve,” Kendix said.

“They are trying to win the match and win the series and therefore it seemed appropriate that the points system reflected those twin goals,” he added.

The next step in the process depends on the current ratings of the two sides.

“If you are playing a team that have a similar rating to you then they are by definition around the similar strength and therefore the points you get for beating them will be roughly the same as the points they get for beating you,” Kendix said.

This story has been viewed 3836 times.
TOP top