Shutdown Corner
Local Club Star
- Joined
- Jan 31, 2014
- Runs
- 2,078
This thread is at the request of Slog and Stallion who asked me to expand my data to a more meaningful sample size for the "flawed ICC rankings" thread.
So now the expanded data includes all Test series starting after Jan 1, 2013 and ending before Dec 31, 2014. Thus India's last tour of Australia is NOT included although some matches took place within the cutoff period.
I will try to run a regression analysis comparing the (ratings, win-loss ratio) of both the official ICC rankings and my system. I am not sure my formulas will pass this test, but we'll see.
I'll list the teams with their win-loss-draw record and official ICC ranking as well so you can have a better idea.
1. Australia | 7-10-2 | 4.671 | ICC rank 2nd
Despite their win loss record, they beat England 5-0 at home, and won 2-1 in South Africa - the only team to win a series in South Africa.
2. India | 7-5-3 | 4.533 | ICC rank 5th
Destroyed Australia 4-0 at home. Beat South Africa at home 1-0, lost to England 3-1 in England.
3. South Africa | 10-3-2 | 4.115 | ICC rank 1st
Harsh ranking, given their vastly superior win-loss ratio, but they've beaten mostly lower ranked sides at home. Still they are within close range of the top spot.
4. England | 8-7-7 | 2.539 | ICC rank 3rd
Beat India 3-1 at home.
5. West Indies | 5-6-1 | 2.210 | ICC rank 8th
They are the only team that are vastly out of position, then again their win-loss ratio is also a lot better than most people would expect.
6. Pakistan | 6-9-2 | 2.181 | ICC rank 4th
They were torrid throughout all of 2013 and most of 2014 only coming into good form late in the year. Just 2 series prior, they were ranked 8th.
7. Sri Lanka | 6-2-5 | 2.045 | ICC rank 7th
Like South Africa they have a harsh rating, but 2 of their series wins have been against Bangladesh. Did beat England 1-0 in England, but that was by mouse's whisker.
8. New Zealand | 6-6-8 | 1.586 ICC rank 6th
Another harsh rating, but they also had a wretched start in 2013 by losing their first 5 Tests or so.
9. Bangladesh | 4-5-4 | 1.345 | ICC rank 9th
Were below Zimbabwe, until the 3-0 win against them late in 2014. A handful of draws also helps to keep them above their African rivals.
10. Zimbabwe | 2-8-0 | 0.310 | ICC rank 10th
A big series draw against an-at-the-time-****-poor-Pakistan is not enough to keep them from the bottom of the table. However, being only a point behind Bangladesh, a hypothetical 3-0 win against the Tigers will see them swap places.
____________________________
Overall my ratings don't differ much from the ICC's rankings. Its also important to bear in mind two things about my rankings formulas:
1) The rating is a cumulative average, and like all averages can be skewed by statistical outlier results. However, as the number of matches increases the rating becomes more immune to such skewing.
2) Teams aren't directly ranked against their opponents ratings, which is good because each and ever series should be (and in reality is) an independent event unrelated to prior rankings and performances. However, teams are rated against the batting/bowling strength of their opponents and because a highly rated team will almost always have good averages, teams end being indirectly rated against the rating of their opponent. A case of being able to eat your cake and have it too.
So now the expanded data includes all Test series starting after Jan 1, 2013 and ending before Dec 31, 2014. Thus India's last tour of Australia is NOT included although some matches took place within the cutoff period.
I will try to run a regression analysis comparing the (ratings, win-loss ratio) of both the official ICC rankings and my system. I am not sure my formulas will pass this test, but we'll see.
I'll list the teams with their win-loss-draw record and official ICC ranking as well so you can have a better idea.
1. Australia | 7-10-2 | 4.671 | ICC rank 2nd
Despite their win loss record, they beat England 5-0 at home, and won 2-1 in South Africa - the only team to win a series in South Africa.
2. India | 7-5-3 | 4.533 | ICC rank 5th
Destroyed Australia 4-0 at home. Beat South Africa at home 1-0, lost to England 3-1 in England.
3. South Africa | 10-3-2 | 4.115 | ICC rank 1st
Harsh ranking, given their vastly superior win-loss ratio, but they've beaten mostly lower ranked sides at home. Still they are within close range of the top spot.
4. England | 8-7-7 | 2.539 | ICC rank 3rd
Beat India 3-1 at home.
5. West Indies | 5-6-1 | 2.210 | ICC rank 8th
They are the only team that are vastly out of position, then again their win-loss ratio is also a lot better than most people would expect.
6. Pakistan | 6-9-2 | 2.181 | ICC rank 4th
They were torrid throughout all of 2013 and most of 2014 only coming into good form late in the year. Just 2 series prior, they were ranked 8th.
7. Sri Lanka | 6-2-5 | 2.045 | ICC rank 7th
Like South Africa they have a harsh rating, but 2 of their series wins have been against Bangladesh. Did beat England 1-0 in England, but that was by mouse's whisker.
8. New Zealand | 6-6-8 | 1.586 ICC rank 6th
Another harsh rating, but they also had a wretched start in 2013 by losing their first 5 Tests or so.
9. Bangladesh | 4-5-4 | 1.345 | ICC rank 9th
Were below Zimbabwe, until the 3-0 win against them late in 2014. A handful of draws also helps to keep them above their African rivals.
10. Zimbabwe | 2-8-0 | 0.310 | ICC rank 10th
A big series draw against an-at-the-time-****-poor-Pakistan is not enough to keep them from the bottom of the table. However, being only a point behind Bangladesh, a hypothetical 3-0 win against the Tigers will see them swap places.
____________________________
Overall my ratings don't differ much from the ICC's rankings. Its also important to bear in mind two things about my rankings formulas:
1) The rating is a cumulative average, and like all averages can be skewed by statistical outlier results. However, as the number of matches increases the rating becomes more immune to such skewing.
2) Teams aren't directly ranked against their opponents ratings, which is good because each and ever series should be (and in reality is) an independent event unrelated to prior rankings and performances. However, teams are rated against the batting/bowling strength of their opponents and because a highly rated team will almost always have good averages, teams end being indirectly rated against the rating of their opponent. A case of being able to eat your cake and have it too.