Monday, October 03, 2011

Michael Hussey and the Myth of Padding Averages

Disclaimer: This post will be entirely stats-based so anyone who finds them repulsive is better off not reading it.

Just how good is Michael Hussey? He isn't the most talked about batsman in world cricket and after a Bradmanesque start to his test career his test average has settled at a more human 53.26. Having played all his international cricket in an era which has seen an increase in averages and scoring rates, it is a very good but not a great number. What is amazing, though, is that this is still the tenth highest average among batsmen who have played 60 or more tests. Three of the batsmen above him  - Kallis, Sangakkara, Tendulkar - have played a lot of their cricket in the same era, so they have benefited just as much as Mr. Cricket. Infact, even with a very generous cut-off of 60 innings in the last 6 years, Hussey has the sixth best average. The 3 batsmen mentioned earlier are ahead of him, a testament to their calibre and their status in the pantheon of batting greats.

But this isn't what makes Hussey so good. It is his ability to seamlessly go from one format to another taking hardly any time to change his approach to batting based on the format that sets him apart from any cricketer of this generation. Among batsmen who have played atleast 160 ODIs, Hussey has the second best average and the eleventh best strike rate. The argument used against most batsmen who bat in the lower middle order is that they pad their averages by virtue of remaining not out. It is one of the most cliched excuses given against batsmen who time and again show the skill and composure to score quickly at the end of an innings and yet remain unbeaten. Historically, as well as since Hussey's debut, batsmen in the top 4 have averaged more than those batting from 5 to 7.


Batting Position
All-time
Innings
Average
Strike Rate
Not Outs
Openers
12666
32.3
70.17
4.74%
3-4
12413
33.85
71.69
10.36%
5-7
16526
26.61
75.13
19.56%
Batting Position
Since Hussey's Debut
Innings
Average
Strike Rate
Not Outs
Openers
4366
32.23
78.19
4.54%
3-4
4267
34.34
75.11
10.05%
5-7
5731
28.57
79.15
18.32%
Batting Position
Hussey's Numbers
Innings
Average
Strike Rate
Not Outs
Openers
2
36
69.23
50%
3-4
20
62
85.71
25%
5-7
113
48.51
88.91
30.97%

As one can see, the batting averages of batsmen in the top 4 have remained more or less same. However, there has been a considerable increase in their strike rates. It's not something unknown to us, most of the changes in playing conditions in ODI cricket have been to the benefit of batsmen. The fact that batsmen in the lower middle order pad their averages due to greater frequency of remaining unbeaten is also offset by the fact that they can't score as big as batsmen before them and at times have to sacrifice their wicket in trying to score quickly.

Finally, the shortest format of the game and once again Hussey is right up there in the list of batsmen. He has the third best average and fourth best strike rate among batsmen who have played 15 or more innings in T20 Internationals. What is amazing about this is he usually bats in the middle order for Australia (he has batted at 4 or below in 19 of his 20 T20 innings) while he usually opens the innings for Chennai Super Kings, for whom he averages 43.27 at a strike rate of 122.

I don't know how much longer Hussey is going to be around. Any talk of his place in the list of batting greats would be premature and should happen once he has retired. Until then, we can just marvel at his performance across all formats, and how on earth did 9 IPL teams allow Chennai to buy him for just $425,000.

Wednesday, August 17, 2011

Does Anybody Understand the Rankings?

It is always fun to see two sets of fans taking potshots at each other. The ICC Rankings have given them one more reason to display their ignorance. Depending on where your team is, the rankings are either a joke or an affirmation of a team's performance.

From December 2009 to August 2011 the Indian fans were pointing to the rankings and gloating about how their team was better than the rest. On the other hand, some thought India were number one only on paper. Some pointed to how India had never won a series in South Africa or Australia, how they didn't have a proper fast bowler, how the players didn't look hot in G-Strings. Now that England have taken that spot from India, you have scores of articles on how they are such a ruthlessly professional side and their rise to the number one position was inevitable. And the ones who were gloating a few weeks ago think England aren't really number one because they haven't won anything of note in the subcontinent for a decade. While that is indeed true, what is not true is that 9 of England's last 12 series have been at home. Some even think that they have the right to tell people that they do not have the right to belittle the Indian team after a bad series.

I wonder how many of these people actually know how the ICC Rankings System works. Jrod has pointed that England could lose their number one spot even before they play their next series. Those of you interested in finding out how can go to the ICC Website and fill in the boxes in the ICC Rankings Predictor to see how it's possible. If you want to dig deeper into how points are awarded to teams based on the series scoreline, here's the exact formula used to calculate it.

It looks like a really complicated calculation, doesn't it? Well, even if it is, it's quite clear that there is no extra weightage for away wins (or home losses). This means that if a team plays a dispropotionate amount of cricket at home, it can go up the rankings. This is quite possible over a 2 or 3 year period as the current FTP mandates teams to play each other home and away over 6 years and allows boards to add on series as they please. Hence, England can host the West Indies in 2009 just two years after a 4-test series between the 2 countries in 2007. Likewise, India hosted Australia for a 2-test series in 2010 (It was just an ODI series initially) despite hosting them for a 4-test series in 2008. Ofcourse these teams have to still ensure they win at home in order to improve/maintain their position, but I don't think many people would contest the argument that teams usually perform better at home than away.

Also, the rankings don't take into account the margin of victory/defeat. As far as they're concerned, a 1-run win is just as good as a win by an innings and 400 runs. Similarly, a draw with both teams scoring an equal number of runs is the same as a draw in which a team was 1 wicket away from an innings defeat.

One can debate the ratings system but it is what it is. There are other rating/ranking systems out there for those who want to look at alternatives. I quite like the system Russ from idlesummers has come up with. I agree with the parameters he uses for calculating the points although I don't know how much weightage he assigns to them.

I guess what I'm trying to say is there is way too much talk of ratings & rankings when all teams and fans should be worried about is their team trying to win every test they play. It doesn't matter if you're ranked 1 or 10, if you pull out of a chase with 15 overs to bat/bowl just because you're tired or think it's impossible to score 6 an over you're cheating the fans of a proper result.

Tuesday, May 24, 2011

IPL 2011 Points Table

Blame it on its to the World Cup or the heartburn that comes from watching an obviously weak Daredevils side, but it took me a while to warm up to the IPL this season. I was so indifferent toward the start that during one of the games I decided to go grocery shopping. 3 of the 4 teams to make the playoffs were decided almost before the start of the tournament. The owners of Mumbai and Chennai had got their way with the player retention rule while Rajasthan and Punjab were docked a certain amount of their payroll. Bangalore had bought very wisely as well. It was just a matter of finding out which would be the fourth team to make the playoffs, although the playoff system was a very good move to ensure teams didn't ease off once qualification was confirmed.

But I did get into it as the season went on and there have been some really good performances. I'm not going to get into who are the best and worst players this season as that's not what this post is about.

This post is about the final points table and a statistical assessment of where teams should have been based on their performance. I should add a caveat here. A 14-game sample for teams isn't really big and there will always be large variations because of some lopsided matches. This was especially the case for the Bangalore, as the team's performance improved remarkably with the arrival of Chris Gayle. Just to give you a sense of perspective, Bangalore's estimated win percentage in matches Gayle didn't play was 29.43. In matches Gayle played, their estimated win percentage was 68.57. It would have probably been even more had they not had that huge loss against Punjab, but since a 9-game sample is small it will be prone to such odd results.

The estimated win percentage is based on the teams' average (batting and bowling), strike rate (batting and bowling), economy rate, and balls per wicket lost. I would have loved to incorporate the percentage of dots, singles, boundaries as well but unfortunately that data isn't available very easily. Also, the weight of each statistic was based on data from the previous IPL seasons. Below is the table depicting the actual and estimated win percentage of the 10 teams.

Team

Mat

Won

Lost

Win %

Est Win%

Bangalore

14

9.5

4.5

67.86%

55.95%

Chennai

14

9

5

64.29%

61.93%

Mumbai

14

9

5

64.29%

63.10%

Kolkata

14

8

6

57.14%

62.96%

Punjab

14

7

7

50.00%

44.15%

Rajasthan

14

6.5

7.5

46.43%

41.91%

Deccan

14

6

8

42.86%

50.09%

Kochi

14

6

8

42.86%

38.73%

Delhi

14

4.5

9.5

32.14%

35.24%

Pune

14

4.5

9.5

32.14%

42.49%


* Bangalore and Rajasthan, Delhi and Pune shared points because their respective matches did not produce a result due to rain.

The 4 statistically best sides have made the playoffs, albeit not in the order of their statistical edge. This is the first time in four seasons that it's happened - Mumbai missed out in 2008 and 2009 while Delhi didn't make the semis in 2010.

There really isn't much to choose between Chennai, Mumbai and Kolkata. It's interesting that Bangalore who have the worst estimated win percentage of the four have topped the table. It's due to the fact that teams play just 14 games and all it takes is a run of 5-6 good matches to make it to the top 4. In this case, Bangalore reeled off 7 consecutive wins thanks in large part to Chris Gayle.