THE CORDON HOME

BLOGS ARCHIVES
SELECT BLOG
September 11, 2009

Batting

Follow-up on comparing halves of players' careers

Anantha Narayanan

There were two very good suggestions to the above referenced article which were worth following up. One was by Arjun to have the datum of 80 innings (Bradman's career) and see what is/was the best streak in players' career. The other was Abhi/Kris's suggestion that I could look at the career in three parts, rather than two, since in most careers there is a slow start, a spurt and a slow finish. I have completed these two tables and presented these here.

The usual criteria apply. For the first table, the minimum is 80 innings and a batting average exceeding 25.00. For the second, I have retained the mid-point limits of 4000 runs and 45 Tests as the cut-off for batsmen.

Test Batsmen: Analyzing the three career splits

SNo.For Batsman         |<---Career---->|Start-third| Mid-third| End-third
|Mat  Runs  Avge|Runs   Avge|Runs  Avge|Runs   Avge
|               |           |          |
1.Aus Bradman D.G     | 52  6996 99.94|2229  96.91|2643 97.89|2124 106.20
2.Eng Sutcliffe H     | 54  4555 60.73|1805  78.48|1537 56.93|1213  48.52
3.Eng Barrington K.F  | 82  6806 58.67|2111  54.13|2379 62.61|2316  59.38
4.Win EdeC Weekes     | 48  4455 58.62|1602  66.75|1643 63.19|1210  46.54
5.Eng Hammond W.R     | 85  7249 58.46|2519  58.58|2396 61.44|2334  55.57
6.Win Sobers G.St.A   | 93  8032 57.78|2781  61.80|2783 60.50|2468  51.42
7.Eng Hobbs J.B       | 61  5410 56.95|1773  57.19|2019 63.09|1618  50.56
8.Eng Hutton L        | 79  6971 56.67|2193  56.23|2661 59.13|2117  54.28
9.Aus Ponting R.T     |136 11341 55.87|2535  40.89|4530 68.64|4276  57.01
10.Slk Sangakkara K.C  | 85  7308 55.36|1951  47.59|2258 48.04|3099  70.43
11.Pak Mohammad Yousuf | 82  7023 54.87|1712  40.76|2273 56.83|3038  66.04
12.Saf Kallis J.H      |131 10277 54.66|2678  43.19|4209 67.89|3390  52.97
13.Ind Tendulkar S.R   |159 12773 54.59|3617  50.24|5202 63.44|3954  49.42
14.Aus Chappell G.S    | 87  7110 53.86|2310  53.72|2394 53.20|2406  54.68
15.Slk Jayawardene D.P.|107  8750 53.35|2653  49.13|2469 46.58|3628  63.65
16.Win Lara B.C        |131 11953 52.89|3884  54.70|3504 44.92|4565  59.29
17.Pak Javed Miandad   |124  8832 52.57|3074  53.93|2817 52.17|2941  51.60
18.Ind Dravid R        |134 10823 52.54|3772  54.67|4001 61.55|3050  42.36
19.Zim Flower A        | 63  4794 51.55|1310  43.67|1488 46.50|1996  64.39
20.Ind Gavaskar S.M    |125 10122 51.12|3951  53.39|3362 54.23|2809  45.31

Average 45.91 44.28 46.84 45.10 (for all 101 batsmen)

The average of the averages figures indicates a clear move up of 5.7% from the first third to second third and a clear drop of 3.8% from the second to the third. Remember that these are on the grand average figure. Individual batsmen have clear move up and move down patterns.

Barrington, Hobbs, Hutton, Ponting (in a big way), Kallis (huge variations), Tendulkar, Dravid (again in a big way) are amongst the ones who have clearly identified low, up, low patterns.

Note the consistency across the complete career of Greg Chappell and Javed Miandad.

Sobers and Gavaskar are amongst those who have had great starts but fallen off drastically.

Bradman, Lara, Sangakkara, Mohammad Yousuf and Flower are those who have finished their careers very strongly.

To view the complete list, please click here.

Test Batsmen: By average sustained in 80+ innings

SNo.For Batsman                Start       Finish    Inns No Runs   Avge
Ins  Year     Ins  Year

1.Aus Bradman D.G 1 (1928) to 80 (1948) 80 10 6996 99.94 2.Saf Kallis J.H 82 (2001) to 161 (2006) 80 19 4661 76.41 3.Aus Ponting R.T 87 (2002) to 178 (2006) 92 14 5904 75.69 4.Win Sobers G.St.A 28 (1958) to 111 (1968) 84 13 5283 74.41 5.Ind Dravid R 66 (2000) to 149 (2005) 84 14 4809 68.70 6.Eng Barrington K.F 34 (1961) to 121 (1968) 88 12 5154 67.82 7.Pak Mohammad Yousuf 42 (2000) to 122 (2006) 81 7 5008 67.68 8.Ind Tendulkar S.R 69 (1996) to 148 (2002) 80 8 4782 66.42 9.Eng Hutton L 42 (1947) to 123 (1954) 82 11 4687 66.01 10.Aus Hayden M.L 23 (2001) to 102 (2004) 80 8 4744 65.89 11.Eng Hammond W.R 15 (1928) to 97 (1936) 83 12 4672 65.80 12.Aus Waugh S.R 82 (1993) to 176 (1999) 95 23 4699 65.26 13.Slk Sangakkara K.C 61 (2004) to 142 (2009) 82 6 4899 64.46 14.Aus Border A.R 88 (1982) to 168 (1988) 81 14 4295 64.10 15.Win Lara B.C 126 (2000) to 205 (2005) 80 2 4985 63.91 16.Eng Hobbs J.B 15 (1910) to 95 (1930) 81 5 4827 63.51 17.Pak Inzamam-ul-Haq 91 (1999) to 175 (2005) 85 9 4795 63.09 18.Win Chanderpaul S 123 (2004) to 202 (2009) 80 17 3947 62.65 19.Eng Sutcliffe H 1 (1924) to 80 (1934) 80 9 4425 62.32 20.Pak Javed Miandad 72 (1982) to 152 (1989) 81 6 4604 61.39

Leaving the colossus outside the discussions, there is a surprise in the second position. I have kept repeating myself many a time. In all the discussions centering around Lara, Tendulkar and Ponting, Kallis has been ignored completely. People point to his lack of wicket-taking ability, forgetting the outstanding batting skills. He and Ponting are the only two batsmen who have averaged over 75 in a consecutive 80+ innings stretch. These two are closely followed by Sobers whose stretch obviously includes the 365*.

Dravid's purple patch comes next, followed by the recent stretch of Yousuf and the mid-career brilliance of Tendulkar. Hutton (not including his 364) and Hayden (including his 380) complete the top-10.

It can be seen that the 80+ innings stretch averages of the last 15 batsmen in the table are within 6 runs.

To view the complete list, please click here.

Test Batsmen: By average sustained in exactly 80 innings

SNo.For Batsman                Start       Finish   Inns No Runs   Avge
Ins  Year     Ins  Year

1.Aus Bradman D.G 1 (1928) to 80 (1948) 80 10 6996 99.94 2.Saf Kallis J.H 82 (2001) to 161 (2006) 80 19 4661 76.41 3.Aus Ponting R.T 102 (2003) to 181 (2006) 80 13 5048 75.34 4.Win Sobers G.St.A 28 (1958) to 107 (1968) 80 12 4969 73.07 5.Ind Dravid R 96 (2002) to 175 (2006) 80 12 4652 68.41 6.Pak Mohammad Yousuf 42 (2000) to 121 (2006) 80 7 4884 66.90 7.Ind Tendulkar S.R 69 (1996) to 148 (2002) 80 8 4782 66.42 8.Aus Hayden M.L 23 (2001) to 102 (2004) 80 8 4744 65.89 9.Eng Hutton L 44 (1947) to 123 (1954) 80 10 4555 65.07 10.Eng Barrington K.F 27 (1961) to 106 (1966) 80 11 4462 64.67 11.Slk Sangakkara K.C 61 (2004) to 140 (2009) 80 6 4740 64.05 12.Eng Hammond W.R 15 (1928) to 94 (1936) 80 11 4416 64.00 13.Aus Border A.R 88 (1982) to 167 (1988) 80 14 4220 63.94 14.Aus Waugh S.R 77 (1993) to 156 (1998) 80 18 3963 63.92 15.Win Lara B.C 126 (2000) to 205 (2005) 80 2 4985 63.91 16.Eng Hobbs J.B 15 (1910) to 94 (1930) 80 5 4753 63.37 17.Win Chanderpaul S 123 (2004) to 202 (2009) 80 17 3947 62.65 18.Eng Sutcliffe H 1 (1924) to 80 (1934) 80 9 4425 62.32 19.Pak Inzamam-ul-Haq 100 (2000) to 179 (2006) 80 8 4470 62.08 20.Pak Javed Miandad 73 (1982) to 152 (1989) 80 5 4578 61.04

Arjun Hemnani wanted a table in which the stretch is exactly equal to 80 innings. I have created a different table and displayed the same here.

It can be seen that the exactly-80-innings average is slightly lower than that when more than 80 innings are considered since there is more flexibility in the extra innings. A below-average stretch can be more than made up with a very good sretch.

The tables look somewhat similar.

Anantha Narayanan has written for ESPNcricinfo and CastrolCricket and worked with a number of companies on their cricket performance ratings-related systems

RSS Feeds: Anantha Narayanan

© ESPN Sports Media Ltd.

Posted by AJ on (September 22, 2009, 20:50 GMT)

Since Sir Don is a given at the top of any batting lists, would it be better to compare the best 80 (or let us say all innings greater than 50) to compare a worthy challenger to the best batsman of all times title?

Posted by Zeeshan Ahmed Siddiqui on (September 22, 2009, 11:25 GMT)

Dear Shafiq, I can give your answer for not 80 innings but in different manner. Just like 80 innings with four opponents of Sir Bradman's case in which we have to include England too.

If we count Yousuf batting average only against four opponent that are England, Bangladesh, Zimbabwe and W. Indies. It is 86.81 in 50 innings with 17 centuries and 12 half centuries. His batting average against W. Indies is 101.16 with least batting average against Zimbabwe that is 68.44.

If we include New Zealand too then his average changed to 83.33 with five opponents with 18 centuries and 15 half centuries in 59 innings.

He scored in one year 1788 runs with nine hundred, world record with highest no. of centuries in one year. It was amazing and also I was very happy at that time how he scored nine hundred in one year.

I think above is enough to prove himself world class batsman of 21st century.

Posted by Shafiq on (September 21, 2009, 19:26 GMT)

Thanks, i am a onsistant reader of ur articles...it is always fantastic to find few pakistani names in record sheets. Now a question-- Can you please ind the best consecutive 80 inings table to compare with Bradman? I am sure Yusaf is still in the mid of a 'great' consecutive 80 innings.

Posted by Aditya Jha on (September 20, 2009, 13:53 GMT)

Ananth, I agree with the "time span" being a factor when comparing the 80-innings stretch. I do hope that you'll do a detailed analysis of the "best 80 innings" by factoring in things like quality of opposition, match performance, peer average - as well as the time span factor. thanks.

Posted by Zeeshan Ahmed Siddiqui on (September 16, 2009, 20:47 GMT)

Nice once again and good idea once more. It is impossible to compare any one with Bradman for 80 innings. He scored 42 fifty or plus scores in his career span from 1928-1948 in 52 test matches. Is any batsman scored 42 fifty or plus scores in his first 52 test matches.

Yes, Gavaskar scored 5007 in his first 52 test matches as he had also managed 42 (42 are fifty plus and 20 are hundred or plus) fifty or plus innings at 27 different grounds in four continents in which seven grounds are related to home grounds with 50 or plus runs scored 4205.

Gavaskar faced bowlers like Derek Underwood, Imran Khan, Micheal Holding, Ian Botham and Malcom Marshall, CM Old, Willis, Clark, Qasim, Lever, Hadlee and Sinkandar He lost his wickets 82 (12+11+11+8+8+6+5+5+4+4+4+4) times by these bowlers in his entire test career.

If possible, kindly arrange any research that showing the strength of any batsman that how many leading wkt takers he had faced like in his case near to 20 bowler joined 200 clubs

Posted by Kris on (September 15, 2009, 14:06 GMT)

Ananth, Further to my previous comment how about a “peer comparison” ratio for the best 80 inn. streak in order to judge the “intrinsic value”? [[ Kris Your comment is a very valid one. Problem is that the peer comparisons between, say test no 1135 and 1332, vary from player to player. In other words, Ponting will have a different peer value to Tendulkar for the same set of tests. I have to do a new methodology in which I build a database segment for every test combination and then remove the concerned players' own figures from that. No problems except that I am involved in two other major projects and have very limited time. Rest assured that I will do it when I can breath a bit more easily (both literally and figuratively considering that I am down with an infection !). Once again many thanks. Ananth: ]]

Posted by Kris on (September 15, 2009, 3:41 GMT)

Ananth, Besides using best “3 / 4 year period” or some such, another means of judging the true value of the best 80 inn. streak may be to include the “peer comparison” ratio during this particular streak. This would at least reduce certain advantages which may have benefited some batsmen such as conditions more suited to batting, weaker bowling, team composition etc.

Posted by Abhi on (September 14, 2009, 13:53 GMT)

Well, my point is simply this: in ANY sport a streak is easier to maintain the shorter the period of time involved. So if you have say a 40 match “winning streak”- it is easier if the 40 matches are played over a short period of time when you are at your peak. It is much more difficult maintaining a similar streak (with the same numbers) over a much longer period of time against more varied opponents and conditions. The whole idea of all your analyses is comparing players (batsmen, bowler, allrounders) over either different or same eras. So, my point is that when any particular player had a major bulk of matches due to good fortune when he was at his peak in good conditions- then how can we compare to another who has not had the similar good fortune? Even though prima facie the numbers may look similar? So, as someone mentions, perhaps a “time” factor and not only a “number of matches” etc factor may be a better yardstick for comparisons.

Posted by Abhi on (September 14, 2009, 10:03 GMT)

Ananth, Do you still disagree with the logic? What if McEnroe had the opportunity to play 10 grand slams in ’84 when he played as well as anyone ever has? He may well have won 8. So, if Ponting gets to play 20 Tests a year for 4 yrs at his peak, in good batting conditions, he naturally gets a much better opportunity to put on better numbers than another batsman who played 5 tests a year for 16 years in varied circumstances. Unless some sort of allowance is made for this the whole thing is completely tilted in favour of the most recent batsmen and any direct comparisons are most odious. [[ Abhi I still feel you are following a wrong premise. How can you compare a 2-week individual Grand slam with a 5-day team-based test. These are totally different types of sporting events. We can only come to an understanding and common ground if you stop comparing totally different sporting events. Make your point purely based on Cricket. Anyhow what has happened that you have such a problem. This is one of many such analysis I have done. You can accept one analysis, not another one and so on. There is no doubt that the modern batmen have played through great batting times. It is true that an average of 50+ is more likely nowadays than earlier. It is true that a 10000 run-aggregate is far more easily reachable than earlier. But it does not make the modern batsmen lesser players. They may be lesser than their figures indicate, that is all. So I suggest you take note of these facts and consider the results accordingly. I cannot be tweaking every analysis I do with all possible corrections. Ananth: ]]

Posted by Kris on (September 13, 2009, 9:55 GMT)

Since the 2000s batsmen have a clear advantage as far as concentration of Test matches is concerned, there are perhaps two ways to neutralize this lopsidedness: 1) Use “Time” instead of no. of innings. Suppose, Ponting (or Kallis) have taken for eg.3yrs 8 mths to complete their best 80 inn. run…use this time period as a base for all batsmen instead of no. of innings which show an incredible degree of variance over time. Also, instead of aggregate runs, we simply have a minimum base runs and show only averages, since aggregate runs will again vary considerably depending on no. of Tests played. 2) Reduce the no. of innings to say 20/30.

Otherwise, we will continue to see batsmen who have done well relatively recently dominate most such “best batsmen” lists.

Comments have now been closed for this article

ABOUT THE AUTHOR

Anantha Narayanan
Anantha spent the first half of his four-decade working career with corporates like IBM, Shaw Wallace, NCR, Sime Darby and the Spinneys group in IT-related positions. In the second half, he has worked on cricket simulation, ratings, data mining, analysis and writing, amongst other things. He was the creator of the Wisden 100 lists, released in 2001. He has written for ESPNcricinfo and CastrolCricket, and worked extensively with Maruti Motors, Idea Cellular and Castrol on their performance ratings-related systems. He is an armchair connoisseur of most sports. His other passion is tennis, and he thinks Roger Federer is the greatest sportsman to have walked on earth.

All articles by this writer