RE:Factors..
by uday on Nov 29, 2005 10:28 AM
Some clarification: how is it that if a batsman faces 1 ball his score can be only 0 or 1? It can be 0,1,2,3,4, or 6 (or even 5, with an overthrow) Similarly if he has faced 2 balls ne can have a score of 1,2,3,4,5,6,7,8,9,10,12!!
Let us tabulate the possible strike rates a batsman can have v/s the number of balls he faces (for simplicity I'll assume that it is unlikely for a batsman to have a SR, say, above 100%)
# balls Possible #runs Possible SR 1 0 1 0 100 2 0, 1, 2 0 50 100 3 0 to 3 0 33 67 100 4 0 to 4 0 25 50 75 100 . N 0 to N 0 100/N 200/N
i.e. a batsman can acheive a SR of p% in an innings only if he faces some integer multiple of 100/p balls in that innings. So it is not true that he is equally likely to acheive any SR between 0 and 100 in an innigs. In particlar it would be very hard for him to have SR of say 59%. This effect is even more prominent for tail-enders who face a few number of balls and therefore have a small set of acheivable SRs.
On the other hand, he can always have SR = 0 or 100% irrespective of number of balls faced; SR=50% as long as he faces even number of balls; and so on.
Also since a typical batsman has a *career* SR of between 50-100% he is unlikely to have an *innings* SR much outside the range. These facts, IMO, explain the apparent conundrum
RE:An explanation
by uday on Dec 01, 2005 04:14 PM
Some clarification: how is it that if a batsman faces 1 ball his score can be only 0 or 1? It can be 0,1,2,3,4, or 6 (or even 5, with an overthrow) Similarly if he has faced 2 balls ne can have a score of 1,2,3,4,5,6,7,8,9,10,12!!
What can i say, a guy who picks Asish Nehra and Harbhajan Singh to write an analysis on batsman? The author, it is cited is a well educated man, i am surprised came to those conclusions without providing the reader with enough evidence. He could have presented simple bar graphs for the strike rates of some of the best batmsmen (ex: SRT, Viv, Gilchrist, Ponting etc) that could have convinced the reader to belive it. The next blunder was his failure to recognise and eliminate what can be termed as a statistical outlier(i.e ashish nehra's performance as batsman in very few innings). What interests cricket enthusiasts is performnace and excellence that is displayed over a period of time rather than a fluke show that happens once in a while.
I am very disappointed with the qulaity of the article and i hope the author comes with something better in his next try.
1. The batsmen considered have been essentially bowlers. This means that they usually face less number of balls. So the possiblity of getting a round figure for strike rate is very high. For example, if you consider that a batsman gets out(and there is a very good chance that a Harbhajan or Nehra will) on 1. The possible number of balls faced (as an example) lets say is 1, 2, 3, 4 or 5. (Its very unlikely that a tail-ender will have patience to stay without scoring for more than those number of balls). In 3 of the 4 cases the strike rate turns out to be a round number (considering 25 is a round number) ie. 100, 50, 33, 25. When you consider pure batsmen, they score maybe 47 or 81 or 103 in maybe 60 or 100 or 95 balls which gives them a higher probability of getting to a non-round strike rate. 2. Considering 1 is true, one should also consider that the probability of a batsman getting out (statistically) inscreases as her faces more number of balls. So the probability of getting a higher score and, in turn, non-round strike rate decreases. So even we consider a sample containing both batsmen and bowlers,we can safely say that there is definitely a reason for this coincidence.
on similar lines of thought....there should be a statistic of number of balls runs scored on to the number of balls the batsman faced. For argument sake lets call this ratio production rate. Production rate would kind of tell us how the batsman scores runs. So a batsman with high strike rate and production rate is an ideal for a ODI.
Im not convinced with your example. Harbhajan Singh comes in right at the death and faces a few deliveries at most. If he were to face ,say, only 2 balls then strike rates of 0,50,100 would realistically be the only possible strike rates he would have (given that the field is spread out and knowing Harbhajan's batting ability). If he faced 4 balls then his chances of hitting 0,2 or 4 runs would be between 50 - 60% realistically.
Do your observations hold out for top order batsmen who would have a much broader sample set?
I would be very interested in seeing your table for Tendulkar or Dravid.
Does sound pretty strange doesnt it. However, here is a thought that I had - although it seems that the strike rates shoudl be a continous (possibly normal) distribution - this may not be the case.
Here is what I thought, assuming that 0, 1 and 2 runs are the most common what will be strike rate be?
So to try and asnwer this - I wrote a little program to run X number of iterations (balls) - randomly scoring 0, 1 or 2 runs each (each with same probability). The results were intersting:
I ran each test 10 times varying the nuber of balls for 10 balls: strike rate of 100 occured 4 times !!! for 20 balls: 3 times!! for 30 balls: 2 times!!
Seems quite suggestive? Maybe some more data crunching is in order?
Sir, It would be really interesting to know about Dhoni with a career strike rate of over 108. In most of the matches his strike rate has been over 100. I think this analysis does not hold true for him. Regards, Arindam
