ICC Test Cricket Champions - small improvements

This ranking follows the basic elements of the ICC rating, but with two minor improvements of things I find wrong with that rating:

  1. The way previous years are dropped from consideration. The ICC drop some series from the rating, thereby causing jumps when the august recalculation is done. I deduct 1/3 of the points, so that no jumps occur. In total, the previous years are counted in the same weight as the ICC.
  2. The value of the opposition. The ICC use the value just prior to the series in question. That makes that the final rating is dependent on the order in which two series are played. If England first win against New Zealand and then lose against West Indies, the West Indies rating goes up more than it would if those series had been reversed. For series within the same season, such an outcome is unacceptable. In my table, the value of the opposition is constant, and arrived at after a number of iterations. The values from previous years are not changed.
    As a consequence, in my table, South Africa did not drop from second in August 2004
  3. a third change is needed because of the second one. The maximum gap of 40 points is not witheld, since this would mean Bangladesh drop to minus infinity. I have included a system however, that the minimum number of points for a win match is equal to the team's own average (this is needed when two teams are more than 50 points apart). Also the maximum points for a loss is the team's average. The corrected points are placed in brackets.
  4. and a fourth one concerned the rating for Bangladesh. In the ICC system, a rating keeps dropping if one keeps losing. In fact, the rating for Bangladesh ought to have gone below 0, but for some reason the ICC did not do this. In my system, a rating does not drop if one loses to a team more than 50 points in front. So all ratings below 33 could have been applied to Bangladesh originally. I chose to award Bangladesh a rating of 33 at the start of my calculations, and they rose to 35 because of two drawn matches in 2003-04, and later to 42 after their first-ever win.
  5. Finally, I have not taken into account matches that were drawn because of very limited playing time, such as Chennai and Napier 2005-06

The table during 2003-04. The table during 2004-05. The table during 2005-06. The table during 2006-07

Current Ranking (August 2007):
sides points M rating
Australia 3513 25 141
England 3635 31 117
South Africa 2789 26 107.3
India 2570 24 107.1
Sri Lanka 2568 24 107.0
Pakistan 2423 23 105
New Zealand 1564 15 104
West Indies 1855 23 81
Bangladesh 787 18 44

the series of 2006-07:
sides opp
(rat)
nr of
tests
+1
points
from opp
(1)
result points
from res
(2)
points
gained

(1) equals the opponent's rating times the number of matches

(2) difference between results, multiplied by 50.

The calculation:
sides points
begin
M points
year
M points
total
M rating
Australia 3513 32 0 0 3513 32 141
Bangladesh 787 20 0 0 787 20 44
England 3635 31 0 0 3635 31 117
India 2570 24 0 0 2570 24 107
New Zealand 1564 23 0 0 1564 23 104
Pakistan 2423 23 0 0 2423 23 105
South Africa 2789 26 0 0 2789 26 107
Sri Lanka 2568 24 0 0 2568 24 107
West Indies 1855 23 0 0 1855 23 81
Zimbabwe 526 10 0 0 526 10 53

Start (1/3 deduction on 1 August 2007):
sides points M points' M' rating
Australia 5339 38 3513 25 141
England 5511 47 3635 31 117
South Africa 4184 39 2789 26 107.3
India 3855 36 2570 24 107.1
Sri Lanka 3852 36 2568 24 107.0
Pakistan 3687 35 2423 23 105
New Zealand 2398 23 1564 15 104
West Indies 2823 35 1855 23 81
Zimbabwe 789 15 526 10 53
Bangladesh 1180 27 787 18 44


Written : 2003-05-21 - Last Modified : 2007-07-17

hermandw@hdw.be / Copyright ©2004-07