Friday, January 06, 2006

Twenty20

I belong to the (old) school of thought that says if you want to win a Test match, you first pick up a bowling attack that can get you twenty wickets before picking up the batsmen who can score 500 runs.

So, IMO, a better bowling attack is the one that has a "better" potential to take twenty wickets in a Test.

I made this point during the recent Ashes and showed how England had an edge over Australia in the bowling department and hence were better placed to take wickets. And in hindsight, it wasn't such a bad analysis :-)

Now, our very own "Bhai bhai Ashes" is being billed as the contest between the Pakistani bowlers and the Indian batsmen, thus giving the impression that the Pakistan has a "better" bowling attack than India.

Question is; is it? Lets see...

To determine which is the better, I have analysed both bowling attacks based on their "Wicket Taking Potential" (WTP, to coin my own abbreviation), which is the simple addition of the Wickets per match statistic of their front line bowlers.

I have assumed that the bowling attack for India will consist of Pathan, Kumble, Harbhajan and Zaheer (sorry Agarkar) and Pakistan will go with Akhtar, Sami, Naved and Kaneria.

Also, to get a better spread for analysis, I have determined the WTP based on the records from the last 5, 10, 15, 20 Tests and entire career. (Except in the case of Naved, who has only played in 7 Tests and Pathan, in 18).

Here it is...



As you can see, the Indians, on any given 5 days, can take about 3 more wickets that the Pakistanis thus making them a "better" attack.

Have I missed something here? Ah, yes! While I can chose to ignore part time bowlers like Malik, Shewag, Tendulkar and Yuvraj (ok Ganguly as well), Afridi is too good a bowler to ignore, as India found out at Bangalore last year.

So what does his inclusion do to the Pakistani WTP? This...



Though India still have their noses ahead, the difference becomes negligible. However, the balance now tilts in favour of Pakistan because (again old school thought) a five man bowling attack is better than fou
r.

Now, this WTP analysis ignores an important aspect that also determines the quality of a bowling attack. It is the bowling average (number of runs given per wicket). Needless to say, if both bowling attacks have the same WTP, the one which gives away less runs wins.

I have used the bowling averages to determine the potential runs a bowling attack is likely to give away while taking 20 wickets, i.e. I have multiplied the bowling average with wickets per Test for each bowler, added it for the entire attack and normalised it to twenty wickets (for example, if India gives 340 runs to get 17 wickets, they will give 400 runs to get 20).

So how do both the bowling attacks compare in this department?



So no matter what the pundits say, reverse swing or not, I think the Indian bowling attack is miles ahead of Pakistan.

Having said that, when Akhtar steams in to bowl to Sehwag, all this counts for nothing...

postscript: Sometimes you don't need 20 wickets to win as the Aussies demonstrated today. In fact, this is the 11th time that a team has declared in the 3rd innings of a Test and lost. And the third time when a team has declared both its innings and lost (including the famous famous Cronje game at the centurion when South Africa scored 248/8 and then forfeited one innings along with England, and England scored 251 for 8 to win).

16 comments:

Pankaj Tripathi said...

Saurabh,
I i had taken the liberty of posting your recent stats in Prems blog.

http://web.ics.purdue.edu/~swarup/forum/index.php?topic=877.0

If you have any concerns with that i can take it off.
Thanks

Crick_Love said...

Saurabh : Can I make one request. Please change the blog setting to a more neutral color in the background. The first thing, which strikes me is that - yikes - white letters on black- totally against convention, especially if you are not the kind of gifted with great vision.
Great analysis. May not agree with all of it though.
Do visit my blog. rants, raves , suggestions, criticisms - all welcome.
http://cricket4ever.blogspot.com/

Anonymous said...

what about umar gul. i think he is going to play a big role in this series..
-kishore

Born a Libran said...

Saurabh, very interesting analysis. There is so much talk in the Paki media about their bowlers being in better form this time around than last time that it would be great if the following analysis can be performed:

Compare the WTP and the number of runs given for 20 wickets in five to ten tests by Pakistan during that period and now, i.e., the England tour. Also, the same statistics compared to how India performed last time. In other words, compare how Indian bowlers bowled last time (and hopefully, will give a preview of how they will perform this time) to how the Pakistani bowlers are bowling currently.

Thanks

Anonymous said...

Good one. I think India's #s are high in recent tests because they were on home soil .. so if you compare india-away vs pak-home in 5-test intervals..Pak should be ahead. Is there "home field advantage"? the #s will tell the story

Saurabh Wahi said...

Pankaj, No problems in you posting on Prem's (or any other blog) mate :-)

Cric_love, point take about the B&W colours on the blog. Some one else suggested it as well. Will do it over the weekend. More neutral colors it is for the main blog. But the graphs remain Black and colourful:-)

Born a libran, yes I planned to do the analysis of India's past performance in Pakistan with Pakistans recent performance, but given constraints with time, etc (considering I have a full time job and family).

I will probably do another analysis before the start of the Test match.

Saurabh Wahi said...

Anon, as I mentioned in my previous post, yes, i will extend this analysis to include additional "scenarios" hopefully before the Test...

Saurabh Wahi said...

Devatha,

I see your point. Frankly, I would like to exclude the stats against BD & Zimbabwe as well.

But my gut feeling is, India will still be ahead of Pakistan...

Unknown said...

I like going through statistical analysis time and again, but I don't like forging an opinion purely on their basis. There are several things one could argue against the above analysis. First, which I strongly believe in, is the fact that Afridi is a bowling all rounder - so, very different to Malik, Sehwag, Tendulkar or the like - they are all batting all rounders or batsman who bowl part time spin (or medium pace sometimes in case of Sachin). When some one asks me what the likely Pakistan attack would be, there is no way Afridi would not feature in it as one of the main bowlers. Secondly, as some else pointed out, the assumptions could be wrong. There is a very strong chance Sami might not get selected. Umar Gul might get a chance. And if Razzaq is fit (I have no idea if he's recovered from his typhoid as yet) then he will be under consideration as well, and he too, is a bowling all rounder. As far as my gut feeling is concerned, I have to admit I think Kumble is going to be a big threat, not so sure about Harbajjan, we tend to play offies really well and it's the leggies that usually cause as all the trouble. And I don't rate any Indian quick bowler barring Balaji and Pathan (sorry Zaheer, Agarkar and Nehra), but thankfully for us, India have decided to discard Balaji all together. I think my conclusion is that Pakistan have the definate age in fast bowling, with India just ahead in the spinning ranks thanks to Kumble.

Saurabh Wahi said...

Zainub, Firtly these stats will mean nothing when the Series start (something I have mentioned on this post). But the reason I did this analysis was to question the hype that this series is Pakistani bowlers Vs Indian batsmen!!! Why? Because Pakistan had one good series against the English?

Fact is, time and again (in the recent past) the Pakistani bowlers for all their talent, have failed to deliver, whereas the India bowlers have been consistently performing well in the last 3-4 years (and hence India's move from Nr 8 to Nr.2 in the ICC ranking).

I acknowledge Afridi's importance in the analysis and I mention that his presence is key to Pakistan's attack.

But I think the series hinges on the performance of Kumble & Akhtar. Whereas I can be pretty sure about Kumble, I have no idea "which" Akhtar will turn up for this series.

Do you? And more importantly, does he?

Saurabh Wahi said...

Vivek, point taken. But for the record, Pakistan isnt exactly "abroad" as the conditions are not so alien to the Indians.

Besides, Kumble's overseas record in the recent past hasn't been too bad. Remember his 24 wickets in Australia not long ago.

I think (as always) Kumble's importance is being underestimated...

anantha said...

Saurabh: One quick question. I think your analysis assumes equal probability of a bowler bowling. I mean, it assumes that every bowler in the attack would bowl the same number of overs each day. So, I think that a better indicator would also incorporate a factor based on the number of overs each bowler bowled.
hmmm.. mebbe I should do this....

anantha said...

Rather, that should have read "one quick point"!

Anonymous said...

buy xenical


valium diazepam



buy prozac


cheap cialis

Anonymous said...

nikon digital camera


camera digital polaroid software



nikon digital camera


digital camera review

Anonymous said...

Hello vitalstats.blogspot.com ! :)
A young couple on the brink of divorce visit a marriage counsellor. The counsellor asks the wife what is the problem.
She responds "My husband suffers from premature ejaculation."
The counsellor turns to her husband and inquires "Is that true?"
The husband replies "Well not exactly, it's her that suffers not me."


computer blog
laptop and battery