Tuesday, July 17, 2012

England v South Africa - The Series Everyone is Looking Forward to But isn't Talking About

So we're just over a day away from the start of the most anticipated cricketing clash of the year. Well, for neutrals anyway. We have a couple of very similar looking teams playing in conditions their bowlers will really enjoy, especially with the kind of weather predicted for the next month.

England are the number one side in the world & have won 7 consecutive series at home. South Africa are number three but can get to number one with a series win (Click here to play around with the ICC Rankings Predictor). They are also unbeaten in 9 away/neutral series spanning almost 6 years, and are the last side to beat England in England. England, ofcourse, are a much improved side from that series in 2008. South Africa, meanwhile, haven't kicked on from their historic series in wins in England and Australia and seem to stumble at the sight of the number one spot.

There are plenty of series previews out there talking of the head-to-head matchups and who the key players will be. The general consensus seems to be that England have the better opening combination, lower middle order and spinner while South Africa have the edge in terms of their middle order and pace attack.

Being the stats geek that I am, I decided to look at how the two sides have performed over the last three years. I broke the teams down into 4 groups - Openers, middle order, lower middle order, and tailenders. I then looked at averages for these four groups - for and against these two sides. I don't see much sense in comparing the averages of their batting order & bowling attack to each other as they wouldn't have batted and bowled on the same pitches (Apart from the 4-test series in 2009-10). I think it makes more sense to see the difference in average runs scored and conceded for each team & then comparing them to each other.

Okay, I said that in my head and it sounds really confusing. Hopefully this will make it more clear.

P = Runs/dismissal for England openers
Q = Runs/dismissal for openers batting against England
X = Runs/dismissal for South African openers
Y = Runs/dismissal for openers batting against South Africa

Final Difference = (P-Q)-(X-Y). This difference gives the advantage/disadvantage England have over South Africa in the openers' slot. Similarly, we'll have numbers for the middle order, lower middler order & tailenders.

Below are the numbers for both teams in all conditions over the last three years:


Batting Position
England
South Africa
Final Difference
For
Against
Difference
For
Against
Difference
1-2
44.81
30.7
14.11
38.96
28.29
10.67
3.44
3-5
45.65
33.09
12.56
62.26
43.05
19.21
-6.65
6-9
31.44
25.38
6.06
27.16
22.66
4.5
1.56
10-11
11.33
11.1
0.23
15.2
9.58
5.62
-5.39

Predictably, England have the edge in the openers & lower middle order while South Africa are far superior when it comes to the middle order. They also seem to be better at wrapping up the tail. On the whole, South Africa do seem to be the better side although it remains to be seen how they will cope with Boucher's retirement.

Since the series is being played in England, I thought it worthwhile to compare England's numbers in home tests to South Africa's overall record. I would have liked to compare their numbers in conditions similar to England. Unfortunately, South Africa haven't played much cricket in such conditions since 2009. I guess you can call this an excercise in trying to see how much of a home advantage England have.


Batting Position
England (Home)
South Africa
Final Difference
For
Against
Difference
For
Against
Difference
1-2
40.96
30.91
10.05
38.96
28.29
10.67
-0.62
3-5
52.27
28.32
23.95
62.26
43.05
19.21
4.74
6-9
35.11
24.45
10.66
27.16
22.66
4.5
6.16
10-11
10.5
9.65
0.85
15.2
9.58
5.62
-4.77

It's interesting that while England's openers average less at home than away, their bowlers don't seem to bowl out the opposition openers cheaper than they do abroad. Their middle and lower middle order has a very good record at home, especially when you consider how cheaply their bowlers seem to dismiss the opposition's 3-9. At home, they have a better differential through positions 3-11 compared to when they play abroad. This is what the South Africans will have to overcome if they are going to win the series.

I think they have positions 3-5 covered. Amla, Kallis (His record in England notwithstanding) and de Villiers should average more than Trott, Pietersen and Bell. But it's positions 6-9 that will be a major problem for them, especially with Boucher not playing like I mentioned before. They can either play the specialist keeper and weaken a lineup that doesn't bat deep, or they can give the gloves to de Villiers and play both Duminy & Rudolph.

By all accounts, Tsolekile is a better keeper than de Villiers who is also a crucial fielder for South Africa. I think they are better served to go with the regular keeper and allow de Villiers to focus on his batting. It will mean they get less runs at 7, but I'd rather get 20 less in that position than see batsmen scoring big hundreds after they were dropped early in their innings. If they do indeed pick Tsolekile, they will hope their bowlers don't allow Prior, Bresnan, Broad and Swann to have a combined average of more than 25. I think they will win the series if they're able to manage that.

My Prediction: It's hard to predict anything with the rain being such a major factor, but I'm going with 1-0 South Africa. It's about time they stopped making me look silly for saying they're the best side in the world.

Monday, May 21, 2012

IPL League Stage Review


The Story So Far

The league stage of the fifth IPL season is over. In my previous post I had taken a shot at predicting where the nine teams would finish. As a Daredevils fan I'm happy my prediction for them has proven wrong. That said, I think they are the least likely of the four playoff teams to win the title as the early season form of Shahbaz Nadeem didn't carry over to the latter stages (They have dropped him!) and the two venues they will be playing at have had pitches which didn't have much for fast bowlers.

This has been a strange season in many ways. I think most people would agree that the league stage this year has been the best in five seasons. 2010 was good, but not this competitive. I try and keep track of team performances every season and determine their likelihood of winning a match based on the teams' average (batting and bowling), strike rate (batting and bowling), economy rate, and balls per wicket lost. The last IPL was pretty straightforward in that the 4 teams that finished at the top also had the best estimated win percentage. However, this season has been different. Only two of the teams that finished in the top four had their estimated win percentage in the top four as well. Below is the table depicting the actual and the estimated win percentage of the 9 nine teams.


Team 
Matches
Won 
Lost 
Win %
Est Win %
Delhi
16
11
5
68.75%
59.94%
Kolkata
16
10.5
5.5
65.63%
62.80%
Mumbai
16
10
6
62.50%
45.84%
Chennai
16
8.5
7.5
53.13%
49.03%
Bangalore
16
8.5
7.5
53.13%
49.71%
Punjab
16
8
8
50.00%
43.81%
Rajasthan
16
7
9
43.75%
55.25%
Deccan
16
4.5
11.5
28.13%
43.34%
Pune
16
4
12
25.00%
34.51%

With just a 16-game season, a couple of odd results are enough to see sides qualify or get eliminated. Even though Mumbai finished third, they have been the least convincing of the final four. They have been blown away in quite a few games and won three matches off the last ball. They have the strongest and the most balanced squad on paper, but have grossly underperformed (More on this later). Rajasthan, meanwhile, lost a game they had no business losing. Bangalore finished below Chennai because tie-breakers only consider net run-rate and don't take into account some of the other factors that are part of my model, which is fine because a model is just that and nothing more.

Looking Forward

Given that the remaining matches will be played in Pune, Bangalore and Chennai, I think Kolkata are the favourites to win the title. The pitches in Pune and Chennai have been good for slow bowlers and the Knight Riders have made it to the playoffs on the back of their spinners. Their only concern is the dependency of the batting on Gautam Gambhir who has scored twice as many fifties as the rest of the batsmen put together! But with the cushion of having two shots at making the final, I can't see them not playing on 27th May.

Delhi had a brilliant start to their IPL but have slowed down since Pietersen's departure. The batting lineup is top-heavy and their fast bowlers are unlikely to do as well in Pune or Chennai. Infact, they managed to take just three wickets in the two games they played at these venues in the league stage. If they have to win the title, it will have to be on their batsmen. Mahela Jayawardene will be the key man as he's their best batsmen on such pitches.

I had picked Mumbai to finish top of the table before the IPL because I felt they had done the right thing by taking the captaincy away from Tendulkar. However, they've created an even bigger headache for themselves because Harbhajan Singh is not going to sit out now that he's captaining the side. He's taken just 6 wickets at an average of over 60. His economy rate - 7.13 - isn't too flash either. They really should be playing Pragyan Ojha and/or Robin Peterson but they won't, and that is why I can't see them winning three in a row to win the title.

Chennai have made it this far because of some outrageously favourable results since their last game. Considering their record in the IPL a lot of people think they're destined to win the IPL this season too. I don't know about destiny, but I do know their biggest test will come in their next game. Mumbai have beaten them twice this season although one of those wins was freak result. They will be playing in Bangalore which is a batting beauty and only five times (Yes, five!) has a Chennai batsman scored a fifty this IPL.

My prediction: Kolkata to win the IPL.