August 23, 2008

Adjusting averages to account for bowling strengths

A stats piece that calculates effective batting averages by determining the strength of the bowling attack faced
113

Some of you may recall the quotient of BQI to bowling average discussed in this post. Roughly speaking, the idea is to reward bowlers who take the wickets of better batsmen. In this post, I'll flip the idea round, and reward batsmen who score against better bowling attacks.
***
Firstly, a digression on Ananth's post. The quotient was defined by summing up the batting averages of the batsmen dismissed by a particular bowler, and then dividing by the bowler's regular average. This is, to my mind, a very useful stat, perhaps the best of its kind for its simplicity (you can of course make it better by making it more complicated in appropriate ways). The only problem is that the numbers you get don't correspond to numbers we're used to in following cricket. How good is a 1.2 bowler? A 0.9 bowler?

Happily, there's an interpretation of this stat that puts the numbers on a scale we're familiar with. It's equivalent to the usual average (runs conceded divided by number of wickets taken), with each wicket weighted by the average of the batsman dismissed. You can set a 'benchmark' average (its value is arbitrary), and I'll set it at 31.5. Dismissing a batsman who averages 31.5 is worth 1 wicket. Dimissing a batsman who averages 47.25 is worth 1.5 wickets. A quotient Q is then equivalent to an average of 31.5 / Q. So, a bowler with a quotient of 1.2 has an 'adjusted average' of 31.5 / 1.2 = 26.25. This is the sort of number we're used to thinking about with bowlers' averages.

I don't know who first came up with the idea of weighting wickets in this way – it was first suggested to me by a friend of mine. Probably various people over the years have thought of it.

Working in the reverse direction (adjusting batsmen's averages) is more difficult, since apart from the last few years, we don't know which bowlers each batsman faced. But we can make a first attempt, by taking the average of the bowlers' averages for each innings, weighting each by the number of overs that they bowled.

To take an example, suppose that in one innings, four bowlers were used:

Bowler A, career average 28, bowls 30 overs.
Bowler B, career average 30, bowls 30 overs.
Bowler C, career average 35, bowls 25 overs.
Bowler D, career average 40, bowls 20 overs.

The "average average" is then (28*30 + 30*30 + 35*25 + 40*20) / (30 + 30 + 25 + 20) = 32.52.

Each batsman's runs for this innings would be multiplied by 31.5 / 32.52 – they'll all be slightly decreased, because the attack is slightly weaker than our benchmark average of 31.5.

(Note: if a bowler never took a wicket, or has an average above 100, then I set that bowler's average at 100. This seems reasonable to me.)

We do this for all innings, and we get adjusted averages for all batsmen.

One useful feature of this method (for both batsmen and bowlers) is that it adjusts across changes in the relative strength of bat and ball (as well as rewarding players who do well against strong opposition). In an era where averages are high (such as today), bowlers are rewarded more for wickets and batsmen less for runs. For players in the low-scoring years before 1900, the reverse is true. Of course, it's possible that in a given era, runs are low because there happen to be a lot of good bowlers and not many good batsmen, and in this case the bowlers are unfairly punished (and batsmen unfairly rewarded). But to my mind the results are better than raw averages.

So onto the results. Qualification: 20 Test innings. Here's the top 20.

name            inns  no  runs  avg   adj avg
DG Bradman      80    10  6996  99.9  90.4
GA Headley      40    4   2190  60.8  62.8
MEK Hussey      42    8   2325  68.4  59.4
CL Walcott      74    7   3798  56.7  58.3
ED Weekes       81    5   4455  58.6  55.9
FS Jackson      33    4   1415  48.8  55.4
JB Hobbs        102   7   5410  56.9  55.0
GS Sobers       160   21  8032  57.8  54.6
L Hutton        138   15  6971  56.7  53.8
H Sutcliffe     84    9   4555  60.7  53.6
AD Nourse       62    7   2960  53.8  53.4
KF Barrington   131   15  6806  58.7  52.6
GS Chappell     151   19  7110  53.9  52.3
GE Tyldesley    20    2   990   55.0  52.2
RG Pollock      41    4   2256  61.0  52.0
KS Ranjitsinhji 26    4   989   45.0  50.8
BC Lara         230   6   11912 53.2  50.4
J Ryder         32    5   1394  51.6  50.4
RT Ponting      197   26  9999  58.5  50.3
FMM Worrell     87    9   3860  49.5  49.5
AG Steel        20    3   600   35.3  49.0

The modern-day greats are surprisingly low down. That their averages should be heavily reduced is not surprising, since the bat has been very dominant over the ball in the past few years. But they're still further down that I had expected. Perhaps there is some bias in the method, or perhaps we should pay more attention to Neil Harvey when he compares modern players to those of his day.

(There's another possibility worth thinking about, and that is a gradual increase in competitiveness of the sport, so that today there are fewer players on the high and low extremes and more players towards the middle. I don't know how big an effect this would be.)

Here is a list of players from recent years:

name            inns  no  runs  avg   adj avg
MEK Hussey      42    8   2325  68.4  59.4
BC Lara         230   6   11912 53.2  50.4
RT Ponting      197   26  9999  58.5  50.3
KP Pietersen    80    3   3890  50.5  48.8
JH Kallis       207   32  9678  55.3  48.3
V Sehwag        100   4   5074  52.9  48.0
Moh'd Yousuf    134   12  6770  55.5  47.8
SR Tendulkar    244   25  11877 54.2  47.4
RS Dravid       214   26  10223 54.4  47.3
A Flower        112   19  4794  51.5  46.5
KC Sangakkara   125   9   6356  54.8  46.5

Tendulkar's low position is a bit of a surprise. It's an anomaly that jars with most people's impressions. But remember that averages are not perfect indicators of a batsman's 'true talent' – there's some inherent uncertainty with them.

A full list of batsmen (with an adjusted average of at least 25), click here.

Some further comments:

- Opening batsmen face the opening bowlers disproportionately often, and this isn't taken into account.

- The conditions or characteristics of the batsmen on a given day can change the effectiveness of the bowlers, and the captain would use his bowlers accordingly. So the simple weighting by career average is not a perfect reflection of the overall skill of the attack. But in the long run the above method should be pretty close.

- There's no allowance for ground or pitch conditions, etc.

- I've ignored not-outs. This is worthy of a post of its own, but not-outs don't affect averages much in Test cricket.

- I've used career averages of the bowlers, mainly because it's easy to do. Career-to-date averages can be unstable. It would be reasonable to add a correction factor for the experience of each bowler. But while I've done a small amount of work in this area, I don't have enough results for it to be usable.

Comments have now been closed for this article

  • Mohanlal on September 8, 2008, 15:38 GMT

    Lara scores some 25 more runs on the average than Sachin in all 99+ scores if * is excluded.He scores at 10 more str: rate on avg: in these inns too.All agreed.But I have 1 point which by instincts prompts me to place Lara a touch below Sachin.Lara didn't have a 100 to show vs Akr-Wqr,Don-Pol,Kum-'x'in India.And these 5 bowlers along with Mc-gra-Warne,Amb-Wal,Murali-Vas to some extent formed the bowling force in the era when these 2 bats played.Others such as Lee in his start days, Gillespie,Srinath,Whitney,Zoysa,Saqlain,Raju etc can only be classified as good bowlers to the utmost extent.So to me it is a very vital factor to have as much as big knocks against this bowling force.Here Sachin far outweighs Lara.Only when Akram,Waqar& Donald retired did Lara start scoring heavily against their respective nations.Sachin due to his better technique had more varied 100s against this bowling force.Lara used to give up when going was tough at both ends.Sachin found ways to overcome .

  • Satyajit on September 3, 2008, 6:10 GMT

    @Barry "not-outs don't affect averages much in Test cricket." Hmm that's a very subjective statment. I don't see any valid reason how you can presume something like that. Can we get an analysis on that? afterall this is statistical analysis :-) You can probably compare the positions of modern stalwarts like Lara, Sachin, Richards and Ponting after considering the not outs.

  • riverlime on September 2, 2008, 19:18 GMT

    Dear "John", Your spelling and grammar both suggest that English is not your first language, so stop pretending. Btw, no comments on why Sehwag is ranked higher than Tendulkar?

  • john on September 2, 2008, 15:29 GMT

    Scoring most of your runs at home carries less weight than scoring runs away from home tendulkar averages more than all the modern players away from home. Hussey? what class bowlers has he scored against? Half an english attack? Pitch conditions have to be taken into account in swing conducing cnditions tendulkar averages more than lara. Who has a higher average against the worlds best team? What does tendulkar scoring hundreds have to do with what the rest of the team does? If anything it should carry more weight. Lara is great no doubt but as an english man ive seen lara do nothing against the top teams until the series is lost then he pulls out a great pointless innings.Lara was no way near as consistant as tendulkar barring the last couple of seasons, he had some awesome series and alot of really poor ones. And alot of those so called winning hundreds came when he had ambrose walsh and co tendulkar had srinath and prasad. And read the stats of sachin anyone negative

  • David Barry on September 1, 2008, 23:24 GMT

    Alim, no no, the title's not misleading. My bullet point near the end of the main post is misleading. I should probably work out how to edit it. In the meantime, there's only us three still here, so the damage is done now.

    If a batsman is left not out, I treat it as a not-out. That is, I calculate the average by runs / dismissals, just as is usually done.

    Now, a lot of people think that averages are inflated by not-outs (ie, a batsman with a lot of not-outs has an average that is too high). In Test cricket, that is generally false. A batsman with a lot of not-outs has an average about what it should be, just like a batsman with few not-outs.

    Not-outs *do* have some effect on averages, since we don't know how many more runs they would have scored if they'd been able to keep batting until dismissed. But this effect is small.

    I have ignored this effect of not-outs on averages.

  • riverlime on September 1, 2008, 20:24 GMT

    David, old chap, you might be on to something here. Why don't you try it out to see what individual players' highest scores REALLY would be, against middle-of-the-road bowling? (Sobers didn't have much of a change, but perhaps other batsmen may be humbled.)

  • Alim on September 1, 2008, 13:57 GMT

    Dear David, Regret the umpteenth post. But I think you should perhaps change the title of the blog from "...averages..." to "...rpis (or some such)...".the current title is extremely misleading. Talk about reading the fine print.

    Also things seem to be getting a bit complex for simple cricket fans like me. Just a few points: 1) So "ignoring not outs" apparently means that if a batsman is actually not out you have taken it as out?! 2) Again this is effectively penalizing the batsman. So the good old fashioned attritional test match innings wherein staying at the crease is paramount, say in saving a test match or avoiding a collapse may not count for as much as a little cameo or blitz. 3) These hard test match innings may involve scoring say just the few runs inspite of spending literally hours at the crease. So the batsman would fare better off in the above table if he instead just came in and simply wacked it around for a short while. What?

  • David Barry on September 1, 2008, 9:49 GMT

    Riverlime, Hayden's 380 got adjusted down to a 267 (using the 31.5 scale). The highest individual adjusted scores: Sobers 365* --> 353* Hanif Mohammad 337 --> 339 Lara 375 --> 337 Sehwag 319 --> 335 Foster 287 --> 323 Jayawardene 374 --> 315 Bradman 334 --> 306 Rowe 302 --> 299 Ponsford 266 --> 297 Gooch 333 --> 294

  • David Barry on September 1, 2008, 6:44 GMT

    Alim (and Anonymous), I apologise for the poor wording of my post. I meant that I ignored any effect of not-outs - I didn't throw those innings away. I just calculated the average in the usual way, runs (or adjusted runs) / dismissals.

    If batsmen were able to complete their not-out innings (ie, keep batting until dismissed), their averages would, on average, go up a bit. But the difference is small, and not worth bothering about.

  • Alim on September 1, 2008, 6:30 GMT

    @david, the anonymous person above has a point.never mind the "superior batting attacks",if you "ignore" not outs the batsmen with greater not outs naturally suffer more.regardless of superior,or inferior or whatever bowling attacks . Basic Math. @riverlime. thank you for the applause.but you seem to forget that a "counterpoint" was being offered to the "point" raised firstly by you folks. such as the incessant harping about rpis. naturally the batsmen with the lower not outs real avg will be closer to his rpi. a batsman with no not outs will have the same avg as the rpi.as mentioned earlier...elementary math.

  • Mohanlal on September 8, 2008, 15:38 GMT

    Lara scores some 25 more runs on the average than Sachin in all 99+ scores if * is excluded.He scores at 10 more str: rate on avg: in these inns too.All agreed.But I have 1 point which by instincts prompts me to place Lara a touch below Sachin.Lara didn't have a 100 to show vs Akr-Wqr,Don-Pol,Kum-'x'in India.And these 5 bowlers along with Mc-gra-Warne,Amb-Wal,Murali-Vas to some extent formed the bowling force in the era when these 2 bats played.Others such as Lee in his start days, Gillespie,Srinath,Whitney,Zoysa,Saqlain,Raju etc can only be classified as good bowlers to the utmost extent.So to me it is a very vital factor to have as much as big knocks against this bowling force.Here Sachin far outweighs Lara.Only when Akram,Waqar& Donald retired did Lara start scoring heavily against their respective nations.Sachin due to his better technique had more varied 100s against this bowling force.Lara used to give up when going was tough at both ends.Sachin found ways to overcome .

  • Satyajit on September 3, 2008, 6:10 GMT

    @Barry "not-outs don't affect averages much in Test cricket." Hmm that's a very subjective statment. I don't see any valid reason how you can presume something like that. Can we get an analysis on that? afterall this is statistical analysis :-) You can probably compare the positions of modern stalwarts like Lara, Sachin, Richards and Ponting after considering the not outs.

  • riverlime on September 2, 2008, 19:18 GMT

    Dear "John", Your spelling and grammar both suggest that English is not your first language, so stop pretending. Btw, no comments on why Sehwag is ranked higher than Tendulkar?

  • john on September 2, 2008, 15:29 GMT

    Scoring most of your runs at home carries less weight than scoring runs away from home tendulkar averages more than all the modern players away from home. Hussey? what class bowlers has he scored against? Half an english attack? Pitch conditions have to be taken into account in swing conducing cnditions tendulkar averages more than lara. Who has a higher average against the worlds best team? What does tendulkar scoring hundreds have to do with what the rest of the team does? If anything it should carry more weight. Lara is great no doubt but as an english man ive seen lara do nothing against the top teams until the series is lost then he pulls out a great pointless innings.Lara was no way near as consistant as tendulkar barring the last couple of seasons, he had some awesome series and alot of really poor ones. And alot of those so called winning hundreds came when he had ambrose walsh and co tendulkar had srinath and prasad. And read the stats of sachin anyone negative

  • David Barry on September 1, 2008, 23:24 GMT

    Alim, no no, the title's not misleading. My bullet point near the end of the main post is misleading. I should probably work out how to edit it. In the meantime, there's only us three still here, so the damage is done now.

    If a batsman is left not out, I treat it as a not-out. That is, I calculate the average by runs / dismissals, just as is usually done.

    Now, a lot of people think that averages are inflated by not-outs (ie, a batsman with a lot of not-outs has an average that is too high). In Test cricket, that is generally false. A batsman with a lot of not-outs has an average about what it should be, just like a batsman with few not-outs.

    Not-outs *do* have some effect on averages, since we don't know how many more runs they would have scored if they'd been able to keep batting until dismissed. But this effect is small.

    I have ignored this effect of not-outs on averages.

  • riverlime on September 1, 2008, 20:24 GMT

    David, old chap, you might be on to something here. Why don't you try it out to see what individual players' highest scores REALLY would be, against middle-of-the-road bowling? (Sobers didn't have much of a change, but perhaps other batsmen may be humbled.)

  • Alim on September 1, 2008, 13:57 GMT

    Dear David, Regret the umpteenth post. But I think you should perhaps change the title of the blog from "...averages..." to "...rpis (or some such)...".the current title is extremely misleading. Talk about reading the fine print.

    Also things seem to be getting a bit complex for simple cricket fans like me. Just a few points: 1) So "ignoring not outs" apparently means that if a batsman is actually not out you have taken it as out?! 2) Again this is effectively penalizing the batsman. So the good old fashioned attritional test match innings wherein staying at the crease is paramount, say in saving a test match or avoiding a collapse may not count for as much as a little cameo or blitz. 3) These hard test match innings may involve scoring say just the few runs inspite of spending literally hours at the crease. So the batsman would fare better off in the above table if he instead just came in and simply wacked it around for a short while. What?

  • David Barry on September 1, 2008, 9:49 GMT

    Riverlime, Hayden's 380 got adjusted down to a 267 (using the 31.5 scale). The highest individual adjusted scores: Sobers 365* --> 353* Hanif Mohammad 337 --> 339 Lara 375 --> 337 Sehwag 319 --> 335 Foster 287 --> 323 Jayawardene 374 --> 315 Bradman 334 --> 306 Rowe 302 --> 299 Ponsford 266 --> 297 Gooch 333 --> 294

  • David Barry on September 1, 2008, 6:44 GMT

    Alim (and Anonymous), I apologise for the poor wording of my post. I meant that I ignored any effect of not-outs - I didn't throw those innings away. I just calculated the average in the usual way, runs (or adjusted runs) / dismissals.

    If batsmen were able to complete their not-out innings (ie, keep batting until dismissed), their averages would, on average, go up a bit. But the difference is small, and not worth bothering about.

  • Alim on September 1, 2008, 6:30 GMT

    @david, the anonymous person above has a point.never mind the "superior batting attacks",if you "ignore" not outs the batsmen with greater not outs naturally suffer more.regardless of superior,or inferior or whatever bowling attacks . Basic Math. @riverlime. thank you for the applause.but you seem to forget that a "counterpoint" was being offered to the "point" raised firstly by you folks. such as the incessant harping about rpis. naturally the batsmen with the lower not outs real avg will be closer to his rpi. a batsman with no not outs will have the same avg as the rpi.as mentioned earlier...elementary math.

  • Riverlime on August 31, 2008, 12:58 GMT

    Alim, I applaud your passion but I regret that it is so clouded by your nationalistic fervour. Please re-read my previous post and see that I am not a participant in the "who was better" Lara vs Tendulkar debate. I think that this analysis is a useful attempt at improving the misleading nature of "Averages" in predicting a batsman's potential, and if you look at batsmen other than Lara/Tendulkar, you may see reason. By the way, David, do you think you could apply your adjusted average to specific INNINGS, to see what, say Hayden's 380 against Zimbabwe, would have been against better bowlers?

  • Anonymous on August 31, 2008, 9:19 GMT

    You have made such an obvious error. You say "I've ignored not outs as they don't matter much". Well, use not outs and see how Dravid, Tendulkar, Ponting etc jump over Pietersen, Lara, etc.

  • Alim on August 31, 2008, 6:37 GMT

    @Barry, Am afraid that’s a poor cop out. Batsmen the class of sobers, gavaskar and tendulkar have played in “world xi’s” at some point…and have made their mark. Sobers best innings ever could well have been the one he played against Australia for the world xi.

    The super series was extremely hyped, the players knew it would be counted as an official test, and to top it all it was against Australia. More than enough to get the competitive juices of any top player going. After a pathetic show and getting whipped it is but natural to come up with lame excuses. Rest assured that if they had scored big hundreds it would forever have been enshrined in folk lore and swooned about for years( not to mention included in any stats).

    @river, knight Am afraid the stats show tendulkar is the better “match winning” batsman. After all you are the stats man, so its there in black and white. Since you insist of using stats to suit some predetermined end, I decided to go through some myself. Lara has scored some 1000 runs more (roughly the amount you are referring to) in 16 innings more against England itself. (His favourite whipping boys) These runs obviously include the 775 runs in 2 innings (375 and 400) in the incredibly difficult circumstances which we are all aware of. i.e. dead rubbers, dead disinterested bowlers, tiny featherbeds. So, now apparently these extra runs are placed under a column denoting “ALL TOP TEAMS except Ban and Zim”. Do you yourself see the way stats can be warped to show just about anything? In any case if we are talking about the best batsmen, we should take into account their performance against the best team. Tendulkar has done better than Lara against Australia. Not only in Australia (i.e. away) but also for a longer period of time. Class tells.

  • Elayaraja Muthuswamy on August 31, 2008, 5:31 GMT

    I am writing to respond about a more technical batsman remianing not out at the end of the innings. (follow up for Alim & Knight comments)

    In case of Lara, I have watched some innings where once the tail is exposed he just goes for the shots to maximize the scores as the tail fo WI team is the weakest among all teams. He could have stayed not out if he wanted But I guess, he wanted to score as much as possible as tail enders will not leave him much time to bat.

  • David Barry on August 31, 2008, 2:50 GMT

    Alim, just on the Supertest. It doesn't matter what the individuals' performances were, I'm not going to count it.

    The World XI players didn't care about that match like they would have playing for their countries. Andrew Flintoff admitted that afterwards: "I guess it just goes to show that you don't just play for money."

  • Riverlime on August 30, 2008, 16:33 GMT

    Alim and Knight, please leave the ceaseless Lara/Tendulkar comparisons out of this blog. The aim was to get an accurate figure to compare batsmen of different eras, instead of using their "Averages", which everyone agrees is flawed. Alim, just because Tendulkar figures poorly on this list doesn't mean he is rubbish. IT'S JUST AN ANALYSIS, FOR CRYING OUT LOUD!! And your shortsightedness to claim that Tendulkar is second only to the Don disparages Headley, Hobbs, Sobers, etc, all who could reasonably claim that title. Further reason for Mr. Barry to have attempted such an analysis.

  • Alim on August 30, 2008, 14:45 GMT

    @knight. 1) You seem to have ignored the 2 lara innings in the super series. so that’s 226. same intentional "error" as David? I am sure if lara had got a big hundred there you and David would have surely included it in all your stats. stressing that it was against Australia. this is what I mean by picking and choosing stats. 2) I would also like to stress on an important point. You and others like riverlime seem to seriously lack basic knowledge of the game. the whole point in having good technique, hand eye coordination, reflexes...what have you is not only to play strokes and score runs but NOT TO GET OUT..! That is the very basic of batting. do you really want to see some flashy square drive and then the batsman getting out on the next ball? 3) so by your convoluted logic you are actually PENALISING a batsman who is good enough not to get out!! You mention staying not out "on purpose"! obviously! is there any other means of staying not out?! so the superior batsman is actually penalized! bizarre.

  • kngiht on August 30, 2008, 13:43 GMT

    Well, I have not bashed Sachin. If you are sachin fan you might see it as bashing. I was just making point since both have played 224 innings against good team it is more easier to compare both.

    But I will ask you as you have made point, if in few years pitch has flattened so that every body is scoring then Why isnt Sachin scoring now? Secondly, if its injuries that affected Sachin making runs then why isnt it affects against team like Bangladesh which has been most profitable to him during these times.

    Regarding playing matchwinning innings nobody have to go and check who is the better match winner. Its innings in adversity that matters. To win or to lose mostly depends on team but great batsman always scores in adversity like Lara has done it more often.

    I don't blame other but even talking about consistency (in one of the previous cricinfo blog or this one about the average against better bowling attack)tendulkar has not been the premiere batsman.

  • Alim on August 30, 2008, 10:25 GMT

    @Knight Just reading your comments. 1)lara 11558 runs,tendulkar 10403. runs (using your own stats.i frankly couldn’t be bothered).a difference of 1155 runs.( where did u get 1500 from i’m not sure) 2)against the same teams tendulkar not out some 20 times more than lara. so taking an avg.of say 56(as david barry has mentioned if a batsman is not out ,his eye is in,and in all probability he will then score more than his avg.) ,but say 56....20*56=1120.3)so the diff is then 35runs. 4)this is a classic case of ppl who have never actually played the game sitting on statsguru and filtering out stats to suit some predetermined “targets”. 5)similar to the theory that if you sit a monkey on a typewriter for a long enough time he will churn out the works of shakespeare. 6)any one who’s followed cricket for the last 20 yrs will know that in the last few years,predominantly due to injuries tendulkar’s scoring has fallen considerably. 6)this is doubly sad because in the last few years with the substandard bowling attacks and dead pitches many many other batsmen besides lara have outscored tendulkar...ponting,shiv,dravid,kallis,pietersen…u name it ,not just lara.so i guess they can all be termed as better batsmen than tendulkar by statisticians like you. 7)if u like laras style ,fine. but whats with the childish tendulkar bashing is beyond me. 8)Tendulkar in innings involving wins :3913 runs in 75 inn @ 62, Lara 2929 runs in 52 inn @61. So actually not only has Tendulkar played more innings which resulted in won matches but has averaged higher in the matchwinning innings.see? now that’s a case of me being a stats ass like you. 9)all cricket fans absolutely love laras batting if not necessarily his behaviour. But you will never see any of them ,including tendulkar fans,trying to run down laras batting ability. But tendulkar bashing is absolutely relished. Why?simple. because deep down everyone knows that tendulkar was the better batsman. In terms of pure batting ability probably second only to the don. Cheers.

  • David Barry on August 29, 2008, 23:29 GMT

    Riverlime, I would say that the best estimator I know of of a batsman's scoring potential is his adjusted average (you could improve it with various factors), which you'd then tweak by considering the average of the bowlers that he'll be facing.

    If you don't know what the bowlers will be like, I'd just use the regular average.

    But remember that *any* of these measures (avg, adj avg, rpi) are inherently uncertain, even for a batsman after a long career. Whatever you use, you're not going to be able to avoid a "plus or minus" of a couple or more runs.

    Knight, innings in which a batsman is left out out typically form a sample where the batsman did better than usual (after all, he wasn't dismissed). If you ignore them, you're going to get a low-biased estimator, not a proper indication of how good the batsman is.

    I agree that some batsman occasionally play for the not-out, but this is at most a very minor factor.

  • knight on August 29, 2008, 14:56 GMT

    @David, If not out is not considered in batsman average then the run he made during the time he was not out should also not be considered. That will better indicate what truely batsman makes when he gets out.

    Just look at the record of Sachin and Lara. Despite playing 224 innings against good opposition ( Not counting Zimbabwe and Bangladesh), Lara has scored over 1500 runs than Sachin purely because of the fact that he scores more runs per innings in test match and quickly than sachin. Sachin over the year have been more keen on being not out like Jaques Kallis. A prime example is his 248* against bangladesh where during a partnership number 11 batsman easily outscored him as he was more keen on not getting not out.

  • Riverlime on August 29, 2008, 13:45 GMT

    "But remember that averages are not perfect indicators of a batsman's 'true talent' – there's some inherent uncertainty with them." (your own words). This jars with your last post. You're not being very clear, Mr Barry. What then are you proposing is a clear indication of a batsman's scoring potential? Let's use Ponting as an example, leaving out the inflammatory Lara/Tendulkar comparisons. Which is best, his average of 58.3, his RPI of 50.7, or his adjusted average of 50.3?

  • David Barry on August 29, 2008, 9:56 GMT

    Riverlime, that's just an artefact of the scale. The regular average correlates better with the adjusted average than the RPI does. The regular average is also a better indication of batting ability than RPI. Treating not-outs as outs is biased against batsmen good enough not to be dismissed.

  • Riverlime on August 29, 2008, 8:54 GMT

    Just a thought. Isn't it curious how closely the adjusted average of this analysis mirrors the players' RPI? As I've been saying for a long time, the "Average" that is quoted as an indication of a player's scoring potential is misleading. A truer representation would be their Runs Per Inning (RPI).

  • David Barry on August 29, 2008, 0:27 GMT

    Vidhya, the problem of career-to-date averages is a difficult one. The main problem is that averages can fluctuate wildly early in a career, just due to random luck rather than changes in the player's true talent. Using whole-career averages is much more stable.

    The best predictor we have (at the moment) of future performances in cricket are career averages. It may be possible to do better by weighting recent performances more heavily than older ones, and incorporating some sort of correction factor for age or experience. But I'm not sure that we have enough data to do that properly, and certainly some bowlers are able to maintain a high level of performance much longer than others.

  • Vimalan on August 28, 2008, 13:41 GMT

    @Mr. Indian

    So according to you, if a person rates Sachin as the best he is doing it for marketing and monetary purposes and if the same person rates Lara as the best than thats the utmost important statement. what a hypocrite.

    So according to you, all people who rated Sachin as the best, including Bradman, Donald, Akram, Warne, Waugh, Ponting, Richards, Roebuck and so many players, officials and others just did it only for marketing purpose??? what a stupid logic

  • Vimalan on August 28, 2008, 13:35 GMT

    with all the flaws in this analysis, i find Sachin has one good company in this list. that is Sir Viv Richards who has a adjusted average of 47.6 compared to Sachin's 47.4, just a difference of 0.2

    So according to this comparison, Richards is not as great as a Youhana, Border, Sehwag, Kallis, Pieterson, and many unknown names..funny isn't it??

    and I can see why the author deliberately avoided to show Richard's statistics in the main page and gave his statistics only in the full list page. Otherwise, people would have questioned this analysis straightaway and put it off. Good marketing guys. Bringing Sachin's name is a good way to popularize the blog and that is what is happening now

  • Vidhya on August 28, 2008, 12:35 GMT

    Some more thoughts on polishing this further, though I could very well be wasting your time !

    We have two major strategies to improve the accuracy - career-to-date and adjusted bowling averages, plus a few assorted ones (exclude the runs against the batsman from the bowler's averages etc).

    How could we possibly handle the reverse case of the "career-to-date" where the bowler's abilities decay over the years. Wilfred Rhodes played Tests some twenty five years after his best days. Kapil Dev's bowling career has two halves split at c.1984. Botham did practically nothing in Test cricket from 1987. Rhodes was nowhere near a the 26-run-average bowler in 1930. Ditto for Kapil and Bothamof later years. Any ideas on how we could possibly make an adjustment for this ?

  • Ravi on August 28, 2008, 5:26 GMT

    A brave attempt to quantify ability. However, this is the second time that I am seeing such and article that leaves out Not-Outs. Why? In statistics data are left out if they are deemed erroneous. To come out NOt-OUT is very deliberate. In fact, strickly speaking, from a mathematical point of view, ducks should be left out before not-outs. It is easier to be bowled for a duck by a silly mistake than to be batting on not-out by a silly mistake. NOT_OUTs are deliberate and attest to the batsman abililty more than ducks do. So, their best and deliberate performances are left out. \not good.

  • Ray on August 28, 2008, 3:58 GMT

    Call me cynical but I find it hard to believe that the level of competition 70-100 years ago was remotely close to the level of competition that the top tier teams of today (Australia,India, Sri Lanka, England) have to offer. I would've loved to see the likes of Bradman and Herbert Sutcliffe face bowlers like Kumble,Murlitharan,Hadlee,Imran or the pace battery of the 70's to mid90's Windies. I guess we can never tell. However I tend to believe that as great as Bradman was he never would've had a batting averaqe that high playing against the competition of the 80's to the late 90's. Even nowadays the advent of minnows Zimbabwe,Bangladesh)and an immense saturation of bowling attacks (too much competition) today's batsmen have padded their statistics for the past decade or so.Just look at the number of triple centuries that get scored now. I believe that throughout the mid70's to the late 80's,not even one was scored - obviously speaking volumes of the quality of bowling during that time

  • Sharath on August 28, 2008, 3:18 GMT

    If this shows anything, it is that if you crunch enough numbers for long enough, you can make a "model" that can prove anything (doesn't apply to Bradman, of course). A model is only as strong as its assumptions, and in my opinion, the assumptions behind this one are a little too tardy.

  • Prasad on August 27, 2008, 22:46 GMT

    Apologies, something went wrong with the way I did the statsguru. Sachin's combined figure for Zimbawe and Bangladesh is (918 + 556) 1,474. Still only about 12.4% of the total 11,877.

  • Prasad on August 27, 2008, 22:23 GMT

    Someone in the business world once told me that you can twist statistics whichever way you want it to go. It is not the statistics that decides it, its the intent that decides which way you go. If people want to dis out Tendulkar, Gavaskar, Richards, etc. they will find a way. If they want to praise them they will find another way. To me ask the players that played against them and have played the game at the highest level. Listen to them carefully. In the meantime, just enjoy them, they are here today and gone tomorrow.

  • Prasad on August 27, 2008, 22:07 GMT

    People keep saying that Tendulkar has scored heavily in Tests against the weaker teams (I am assuming that it is Bangladesh and Zimbawe). It really confused me so I decided to look up how much is "heavily". So in real figures he has scored 499 runs against Zimbawe and 471 runs against Bangladesh. Thats only 970 runs combined out of the 11,877. I am still scratching my head. Can someone tell me how did we come to the conclusion that he scores heavily against the minnows when he has scored 1,384 runs against Australia alone? Or are you guys talking about his ODIs?

  • Praveen on August 27, 2008, 21:09 GMT

    Steps to make your blog popular: * Manipulate statistics such that there will be an artificial controversy with Sachin. * Post it. * Sit and enjoy the mayhem.

    Otherwise why the author would have removed the not-outs.

  • priceless on August 27, 2008, 20:37 GMT

    Value to the team. divide runs and wickets by the number of matches played. Runs = 1 point Wickets = 20 points.(reason 5w per inns is like a 100)

  • Chamara on August 27, 2008, 13:59 GMT

    I think when comparing BC Lara and Sachin, these statistics are valuable. Also I rate Lara much higher than Tendulkar, because in Lara era West Indies team is very weak one with no world class batsmen. He scored his runs almost single handedly. On other side India has more world class batsmen than any other team. Even batsmen rated more than Tendulkar, like Dravids. The other fact is Lara has only 5 not-outs comparing Tendulkar's 25+ not-outs. Which certainly helps Tendulkar to keep his average. And I think that excludes the idea that Lara played only for himself.

  • knight on August 27, 2008, 13:03 GMT

    I don't know complex mathematics but what I know is if we remove record against Bangladesh and Zimbabwe from a players profile we generally know how good batsman is against good bowling.

    Looking at the record of Sachin and Lara surprisingly both have played same number of innings (224) against all teams if Zimbabwe and Bangaldesh is removed from their record. Lara has made 11558 im 224 innings where as sachin has only made 10403. That is difference of staggering 1500 runs. Then if you take not out Sachin average drops to 46 where as lara still has average over 50.

  • Marcus on August 27, 2008, 11:22 GMT

    I'm not surprised to see how far Ponting comes down, as I've long said that he's benefitted from playing in an era of weak bowling attacks (although an adjusted average of 50 is still mighty impressive). But I am genuinely surprised that Mark Waugh and other '90s players have slipped, because I regard the '90s as being a decade of good bowling, with especially the Australians, South Africans, Pakistanis and West Indians having excellent attacks. Just out of curiousity, I'd be interested to see what the strongest decade of bowling is, to see which batsmen had it toughest.

    But I am glad to see an adjusted average for W.G. Grace, because an average of 32 does him no sort of justice, and I also like the fact that Murray Goodwin's adjusted average is much of a muchness with his career one, which marks him out as a quality player who could have achieved a lot if he had the opportunity.

    All in all, a very interesting analysis which provides a good context for batting stats.

  • WeakUP on August 27, 2008, 6:50 GMT

    Smell the roses guy. Sachin is a poster child, India keeps playing him for records. Look what has happened in the last 4 yrs. He plays weaker team only. He recently did good in australia, but without MG or SW. BL and ST are not good with old ball. In the SL series, he used injury to bail out. Simple drama and the Indian are mesmerized. If people dont want to believe in analysis like this, then they should believe in scores and runs he has selfishly scored

  • Indian on August 27, 2008, 3:21 GMT

    Hi Vimalan, The youtube link where Benaud says that T'kar is the best after Bradman MAY also be Richie Benaud's marketing strategy to be liked by a BILLION people. He knows that the sales for his DVD (Benaud's Greatest XI) will increase if he included T'kar than if he excluded him.

    So, one cannot take the words of these people very seriously these days as there may be a lot of different motives (money etc) involved.

  • David Barry on August 26, 2008, 23:35 GMT

    That's a good point, Jeff. Simply using the bowlers' (or batsmen's) averages works best when there are a large number of teams who all play each other an equal amount of times.

    It seems to work 'pretty well' though, even for pre-1900 cricketers.

  • Srikanth on August 26, 2008, 20:21 GMT

    With the Centenary of Sir Donald's birth approaching, its insulting to talk of insignificant gnats (comparatively) like tendulkar. Would all those singing tendulkar's praises take a moment to check the stats on the greatest performer in the history of ALL sports. Statistically there is NO ONE who towers over the rest the way the Don does in cricket. He did it in the pre helmet era with uncovered wickets. Incredible.

  • Vimalan on August 26, 2008, 17:26 GMT

    Anand, very good reply

    may be the so called true-fans think they know better than Bradman, Benaud, Donald, Akram, Warne, Ponting, Waugh and so many who rated Sachin as the best they have seen..just for starters, guys please watch this video.

    http://in.youtube.com/watch?v=vthfSPF_LCw

    this is from a man who has seen it all for the past 50-60 years

  • Jeff on August 26, 2008, 9:10 GMT

    David,

    Thanks for the interesting analysis.

    While a lot of the comments are depressingly predictable, some have truly helped the discussion. I particularly liked the one about Barnes from early in the list, and also the one relating to Simpson's Paradox.

    I have my own regarding the adjustment for Bradman... is he not being penalised for his own success?

    In his prime, there were only 3 teams playing a significant amount of matches (with the Windies just getting started) - therefore, bowlers would have played a high proportion of matches against Australia and Bradman. Given his high scoring, this would have increased their bowling averages and therefore brought down his own adjusted average.

    I guess in an ideal world, you would adjust for bowling averages excluding runs made by the batsmen in question. But practically this is probably impossible.

  • David Barry on August 25, 2008, 23:22 GMT

    Vidhya, suppose that a batsman has scores r1, r2, ... against bowling attacks b1, b2, .... Then his adjusted total runs is r1*31.5/b1 + r2*31.5/b2 + .... You can factor out the 31.5: (r1/b1 + r2/b2 + ...)*31.5

    So if you want to change the 31.5, you just need to do a simple overall scaling.

    Marcus, I think I screwed up in saying that I ignored not-outs. I meant that I ignored any effect of not-outs, ie, I just take runs/dismissals as the average (and adjusted runs / dismissals as the adjusted average).

  • marcus on August 25, 2008, 14:57 GMT

    David, a flaw in your presentation is that you are listing their average and their adjusted average, rather than their runs/innings and their adjusted average. I believe this is confusing many readers.

  • Vidhya on August 25, 2008, 13:59 GMT

    Jim Burke is a curious entry. His adjusted RR is up by almost 5 (34 -> 39). The figures of no other post first-WW cricketer shows a comparable increase. Burke played more than half his matches against England against whom he averaged only 29. A single series in South Africa where he averaged 64 seems to have made all the difference. Or perhaps this is an example of the thing that Prasanth pointed out.

    Barry, as for my previous question, it may not be as straightforward as multiplying by (33/31) because you are treating each innings *individually* for the "adjusted runs".

  • srini on August 25, 2008, 13:41 GMT

    Anand, another thing about hype. shane warne is a great bowler and was named wisden's cricketer of the century (something which i strongly disagree). when he was named cricketer of the century in 2000, he had an absolutely horrible record against india and sri lanka (best players of spin) and an even worse record in the sub continent. the equivalent of mcgrath or donald having their worst records in aus/sa. just a year before he was dropped in favor of colin miller in west indies. he had absolutely no business in being considered amongst the cricketers of the century. why was he named cricketer of the century then? y is that so?

    pure hype thats what. his 8 years werent even his best. that is what hype can do.

  • srini on August 25, 2008, 13:29 GMT

    anand, i said his record IN aus is overblown. similar to gavaskar's against windies. i am no way belittling his achievements. even tendulkar has not such a great record against allan donald so that really puts things out when u compare him with lara. and lara had a pretty successful first tour in australia when he scored that magical 277.

    if anybody claims tendulkar has no quality is ignorant and probably considers pele to be brazil's best wicketkeeper.

  • David Barry on August 25, 2008, 11:28 GMT

    Ananth: 1. Walcott did much better than Weekes against Australia (57 to 40), which I think accounts for most of the reason why Walcott goes up and Weekes down.

    2. 31.5 is the overall bowling average over Test history. It doesn't matter what it is - it just sets a scale.

    3. Not-outs don't affect averages much in Test cricket. If batsmen were able to bat until dismissed, their averages would be very similar to what they are now (ie, runs / dismissals). Abhijeet's suggestion assumes that a batsman left not out above his average would have been dismissed before he scored another run. This is plainly wrong in general - he's often likely to score even more runs than his average, since he's got his eye in.

  • Ananth Narayanan on August 25, 2008, 11:21 GMT

    Since I was busy with two other articles, have not had the time to look at the comments. Hence any repeats may be condoned. 1. Headley's adjusted average is up by 3%. It is possible that he faced much better quality bowling. Similarly Walcott's figure is up by over 2% but Weekes' figure is down by 3%. That to me is intriguing since both had almost parallel careers and must have faced similar bowlers during the 10 years or so they have played. Any explanations or insights into this. 2. How was the figure 31.5 arrived at. Is it the all-time Bowling Average for bowlers. Then what about the non-bowling dismissals. 3. The complete exclusion of Not Outs leaves a big void. Maybe incorporation of Abhijeet Dongre's comment on my post, to consider any not-out innings below the batsman average as not out and considering RPI, may, I repeat, may lead to a better result. To sum, a thought provoking analysis. Ignore the brickbats. Tendulkar surely is great enough not to need this type of support.

  • Anand on August 25, 2008, 8:52 GMT

    Phew. Guess I was right after all. The venom spewers are out in full force. Let’s see: @srini: Tendulkar has a better record against Aus in Aus than any other batsman, that too for a period of almost two decades. Perhaps it may have been a bit worse with Mcgrath (but then the more he played him he would probably have got the “hang’ of him even more)but then so would have laras if you take out his first tour etc. Using the same logic if you take laras batting prowess against a class fast bowler(not medium pacer like McGrath)like A.Donald it would make even his most ardent fans blush. @ V.S.MUTHUSWAMY : A "mediocre" batsman like tendulkar has got the greatest players that ever played the game drooling incl.recent "matchwinners"like lara,ponting.Not to mention the fans.If you seriously cannot see the sheer class of Tendulkar perhaps you are watching the wrong sport.. So correct me if I am wrong but either: 1)Your knowledge of batting is far greater than practically all the games greats or 2)there is some sinister secret in your possession which ppl like bradman,benaud down to lara and ponting are unaware of. Further,as ppl have mentioned above, firstly what exactly does a "matchwinning innings" constitute? Scoring runs which contribute to a win or being there till the end of the innings? In either case Lara's "matchwinning innings" would then unfortunately be perhaps the 1 or maybe 2( inspite of having the mighty walsh/ambrose in the team). I’ve seen some stats above which clearly show that tendulkar has scored much above his career avg. during wins than the laras and pontings. Or perhaps you are referring to some dramatic last man standing win like the lara 153 and the loss like the tendulkar 136. perhaps then you will also be aware that 10 other folks are playing in the same team? That there is always an element of luck involved( Healy dropping a lara sitter etc)? That no matter how much you score ,in a test match the bowlers HAVE to get 20 wickets for a win? That the situation is different in ODIs where this requirement does not hold? That perhaps if you get off statsguru and actually hold a bat it may help? @srini Of course I fully agree that things are much more hyped nowadays. But how long can the hype last for if a sportsman does not have the basic quality anyway??? 1 yr?2? 5? But for 20yrs like tendulkar!? Are the millions of fans ,greats,legends etc of the game so very dumb that the media can make asses out of all of us for 20 years!!? Do you think that now at my relatively advanced age I still stay glued to the TV when Tendulkar bats because of the goddamn media?! Can you honestly say you (or any true cricket fan)still don’t get goosebumps when Tendulkar takes on the fast bowlers with his straight drives and back foot punches?? I’ve been watching tendulkar (and of course the laras and pontings) from the beginning.( sunny and co. too)……and to me tendulkar is unquestionable the best batsman ive seen. Id just loved for India to have had a say walsh/ambrose kind of combination in the team. But with the kind of bowling India has had I think it is miracle India didn’t lose practically every test match it played, especially in the late ‘80s,’90s. You know , during Sunnys time ,everyone in India adored him but noone really put him above Viv. Not so with Tendulkar. As a batsman he is non pareil.

    Further the sheer venom spewed out by you folks would conversely immediately put Tendulkar out on a pedestal!! Noone in his right mind would ever even bother to go to such lengths for a truly run of the mill batsman .

  • Natwar Modani on August 25, 2008, 8:08 GMT

    Continued: We may require some adjustments to judge the batsman better by applying some multiplicative factor which would adjust the contribution of the batsman based on the batting position and the expected contribution from that batting position (e.g., you would expect batsman from 1 to 7 to contribute more than the number 8-11). Similarly, one would need to handle not outs and did not bat situations also in some meaningful way. One may come up with more such adjustments, but basic idea can be useful to get a sense of relative contribution of a person.

  • Natwar Modani on August 25, 2008, 8:03 GMT

    Why don't you use performance of a person relative to others in the match? This would be a uniform way of measuring both bowlers and batsman and would implicitly factor in the pitch condition. For example, in the current IND-SL series, Sehwag scored (201*+50) runs in a match where there were 1026 runs scored for 40 wickets, so his relative average is (251/1)/(1026/40) ~ 9.78. On the other hand, Mendis took 10 Wkts for 209 runs, which gives him a relative bowling average of (209/10)/(1026/40) = 0.81. Here, a relative batting average is better as it is higher and relative bowling average is better as it is lower (consistent with current definitions). It would penalize scores against week oppositions (since team mates would also have done well) and against favorable conditions (on a flat pitch, many would score runs; on a tough pitch, many would take wickets). Further, it would reduce the era effect and pitch condition etc as the matches are scaled to themselves.

  • Petr on August 25, 2008, 7:02 GMT

    A few comments based on what ive read above: 1) Bowlers, like batsmen, go through different phases (form), especially when just returning from injury, and finally they just simply age and fade away. So as some have mentioned, facing say Donald, in the early to mid nineties would be vastly different to facing him in the early 2000s. So obviously it is not just which bowler which matters but also at what point of his career he was faced.

    2) A method which may give a better idea of ‘superior” batting skills would be not only how “well” a batsman has faced “superior bowling attacks’…but also how “BADLY” he may have faced them! So say a Ponting, Gilchrist etc would then be penalized for their poor showing in India on some tours against quality spin. Similarly other batsmen against serious pace bowling. Surely, overall skill involves not only excellence against good bowling but also lack of weakness against such bowling. This weakness can ONLY be quantified if “negative points” are given for the same. So a stretch of miserable scores against top class bowling must be give negative points. Other wise on a particular day a single huge score would apparently make it appear that a batsman apparently does not possess any weaknesses.

    3)@anand,matt,mark etc. puerile comments from "mark" only serve to reinforce anands comments. I suppose subconsciously the bloggers know it too. Just put up a pic of tendulkar with a slightly odd comment and you are guaranteed a zillion hits and comments. In any case bradman had sachin in his world xi and benaud too thinks he's the best...so I guess"sachin who” must be a fairly good batsman.

    4) Some have also asked what laras avg. would be if you do take out the 375 and 400* against England which would obviously vastly inflate his average against "superior bowling attacks" in general. similarly what effect such odd "peaks" in batsmen with a penchant for big scores have in general towards overall performance against “superior bowling attacks”? In another blog “anant narayan” has used a “stretch” of 10 innings which is ridiculous to say the least. 5/6 innings should be the max.For e.g. a batsman who tends to score big and who simply cannot play say Donald at his peak on a juicy pitch would then “even” out things given a chance on a featherbed. Would this justify a higher ranking?

  • srini on August 25, 2008, 6:07 GMT

    Anand, tendulkar has got pulses raising because he has been playing in an era where communication gaps have been reduced. its purely a result of advanced technology. for example, in the eighties if u had to compare kapil dev and imran khan, how would u do it? where wud u get ur statistics? where could u see their past performances? how could u find out who has taken his 132nd wicket with which ball and test?

    similarly how wud u compare richards and gavaskar? hell if we never had internet we could have never said gavaskar struggled against windies when the overall opinion has been that gavaskar was legendary against them.

    i am pretty sure if tendulkar and lara were born in the 1930s tendulkar would not be this hyped (rightly or wrongly).

    dude these days crap like IPL gets headlines over real issues like communalism corruption or terrorism so get real pal.

  • V.S.MUTHUSWAMY on August 25, 2008, 6:05 GMT

    anand

    It is not in that, as a fan of cricket most of them dont want a mediocre batsman like Tendulkar to be rated as the best in the business. your question itself has answers. Why other greats were not debated. bcos they have good record as well as match winning performances. But with Tendulkar????????????

  • srini on August 25, 2008, 5:58 GMT

    while i am not denying the achievements of tendulkar. his record in australia is overblown by two series where he did not face mcgrath. shane warne has been an absolute non factor against india. amit pagnis had his number in a tour game. ppl wud say how can mcgrath as a single bowler can influence when u have warne gillespie etc. i would like to point to the ashes in 2005. without mcgrath aus lost a test and with mcgrath not fully fit they lost the series!!

    also most indian fans have gavaskar at this high pedestal by claiming his records against windies. gavaskar for all his talents his average against west indies is overblown by 2 series, his debut when west indies were in a transitional phase(they lost to england away, and could not beat new zealand at home and lost to australia home) and the another series in india when windies were decimated by packer series.

  • David Barry on August 25, 2008, 4:32 GMT

    Vidhya, if you change the 31.5 to something else, you just scale everyone's adjusted averages, so you can never change the order.

    If you want to use (say) 33, you multiply the adjusted averages in the tables above by 33/31.5.

  • Vidhya on August 25, 2008, 4:29 GMT

    Barry, how significant is the effect of the number 31.5 ? I understand that it is the bowling average in Test cricket. Does changing the time span considered for calculating this makes any appreciable difference in the order of batsmen.

    For instance, if you consider only post-war years, the average is 32.08. Does the table of the current batsmen look the same if you use this instead of 31.5 ? Or if you stop at 1960 and construct the table will, for example, Hobbs become "better" than Jackson ?

  • Mark on August 25, 2008, 3:49 GMT

    All these ridiculous made up stats do is show how good Don Bradman was. You reduced his average by 9 (ridiculous) and he is still the best by 28 runs. Sachin?, Sachin who?

  • Anand on August 25, 2008, 2:39 GMT

    One thing there is no doubt about whatsoever... No cricketer in history has ever raised more heated emotions than Tendulkar!! For such an unassuming,modest,non controversial sort of chap that has got to be a unique achievement... There have been lots of international stars(from grace,bradman down) and indian stars(kapil,sunny) stars...but no one,repeat no one,can get pulses racing like tendulkar(either for or against).. Just look at any blog! Many,many batsmen have come and gone(and "lost it" along the way)...but only Tendulkar will get posts either like "Matt" above or like those from his adoring fans(like me) to whom he's simply the greatest ever. Ever noticed the Laras,pontings,pietersens etc failures ever even mentioned? no chance... But with good ol' sachin it's practically the other way around. as roebuck says: It is the fate of great batsmen to be chastised for their occasional lapses while the mediocre ones are forever hailed for some perceived heroic performances. Amen

  • Satheesh Nair on August 25, 2008, 2:38 GMT

    V.B warrier I think you are misinformed about Gavaskar's achivement against WI. Eventhough he played in the same era of mighty West Indies fast bowelrs Roberts Holding Marshall and Garner, he seldom played aginst them except in 1983 WI tour where he was miserable. His high average agianst WI is influenced by his debut series (Hoder and Julien) and the series gkery Packer times when none of these fast bowelrs played.Similarly he never played well against Lille Thompson Pascoe combo.

  • Matt on August 25, 2008, 0:58 GMT

    how has it become impossible for any sensible conversation about cricket history? All it ever turns into is a bunch of Indian supporters whining and bitching about something to do with Tendulkar. No matter which way they have tried to justify it he is still inferior to the other greats of the modern game...Ponting, Kallis. Further, the longer he continues to play the worse his stats are going to get. He is clearly losing it and would hardly figure in the worlds current top ten batsmen. If he does make runs these days they are slow, boring and unproductive as far as the team situation goes. HAS-BEEN.

  • David Barry on August 25, 2008, 0:39 GMT

    Couple of points. The reason why I give Lara's average as 53.2 and not 52.9 is because I ignore the Supertest (I probably should have mentioned this, but I was wondering if anyone would notice...).

    On not-outs: I mean that I calculate the average in the usual way: runs (adjusted or otherwise) divided by dismissals.

    Not-outs don't affect averages much. When a batsman is left not out, he's denied the chance to continue his innings, having (usually) already got his eye in. You can estimate how many more runs, on average, the batsman would score after being left not out. When you apply this to all not-outs, you end up with only small changes to averages, with an average change upward of a small fraction of a run. That is, not-outs usually deflate averages.

  • Vidhya on August 25, 2008, 0:25 GMT

    Riverlime - Blame David ! The table he provided has 53.2 as Lara's average and I copied it without cross-checking. I hadn't used a calculator either (used Excel) :-)

    If you read my post again, you'll see that I am talking about including the not outs.

    Barry has excluded not outs from his calculations (see the comments at the end of his blog entry). Lara's adjusted average is 50.4. If you consider not outs, it will be 50.4 * (230/224) which is 51.75. This is 97.86% of his conventional batting average (or 97.27% of the 53.2 which I used earlier).

    Tendulkar's adjusted average is 47.4 which ignored the not outs. If you consider the not outs, it is 47.4 * (244/219) = 52.81. 52.81 is *97.38%* of his career average of 54.23.

  • Riverlime on August 24, 2008, 20:01 GMT

    Vidhya, where do you get your figures? If you check Cricinfo, you will see that Lara's average is 52.9, while it is 51.5 if you ignore notouts. That's still 97.3 %, as you mentioned. However tendulkar has an average of 54.2, which drops to 48.6 if you ignore not outs. THAT'S 89.8%, NOT THE 97.4% you quoted. Why else do you think he has played 12 more innings than Lara and is still 76 runs behind him? (At his recent rate in SL, that will take him another 6 innings!!) Get a calculator the next time you want to do statistical analysis, and you will save yourself some embarrassment.

  • Mukund on August 24, 2008, 18:34 GMT

    I agree with you Shrik. I came here because there was Sachin's pic there!!

    Nevertheless, I must second a thought by another commentator here "Any statistical method should be tested for true positives". I think that is the undoing of this analysis. If one knows that 'x' is the answer to a particular problem, how useful is that method then, which gives the answer as anything but 'x'?!

  • kumar on August 24, 2008, 18:16 GMT

    well, what is not counted is the so-called great batsmen not played against their "own" best bowlers. Example: Tendulkar's runs against Lee, McGrath, Warne etc vs Ponting's against Zaheer, Munaf, Srinath? How do you rate those runs? Australian bowlers always been one of the best in almost all the times. But, how their best batsmen like bradman, ponting, hayden, chappell etc fared against them? One can imagine their averages?

  • Michalite on August 24, 2008, 18:13 GMT

    We also need to consider that he often doesn't play many shots against the best bowlers in the team because they're more capable of taking wickets. And if sachin falls, usually India fails. So he probably negated the best bowlers spell and tried to score against the weak options

  • Anand on August 24, 2008, 16:08 GMT

    I'd say Tendulkar's career can easily be split into 2 parts... Before and after injuries.

    Till,say the year 2000 he was the best batsman. Since then several batsmen ,including Lara and Ponting have scored much more heavily and consistently..... This basic fact solves the apparent "mystery" when using selectively chosen stats over Tendulkar's career.

  • Alim on August 24, 2008, 15:21 GMT

    Just wondering what Lara's average against "superior bowling strengths" would be if you take out the 375 and 400* against dead,disinterested bowlers,in dead rubbers on featherbeds?

    Further to comments on Tendulkar's "5 matchwinning innings"...firstly what exactly does a "matchwinning innings" constitute? Scoring runs which contribute to a win or being there till the end of the innings?

    In either case Lara's "matchwinning innings" would then unfortunately be perhaps the 1 or maybe 2( inspite of having the mighty walsh/ambrose in the team).

    Also,no batsman in the last 20 years has played the best team in that period (Australia) as well or as consistently as Tendulkar.In fact Tendulkar has the best record against Australia in Australia!! So much for batting against "superior bowling attacks".

  • gautam on August 24, 2008, 15:20 GMT

    why dont we have another list where we adjust the batting averages depending upon the expectations people have from a batsman ? In that case I bet Tendulkar will have an average which is greater than the sum total of the adjusted avaerages of Bradman, Headley, Walcott, Richards, Lara, et all.

  • Ravi Kumar Putcha on August 24, 2008, 14:32 GMT

    Here is an other interesting problem. Because this measure ignores not outs, it penalizes SRT heavily because of his top five test knocks, 4 are not outs!! In other words he gets no credit for almost all of his top 5 performances. That is another major anomaly.

  • Vidhya on August 24, 2008, 14:26 GMT

    David, why did you skip the not outs ? If not outs are considered, Lara's adjusted average will be approximately 51.7 (against his career ave of 53.2) while Tendulkar will have 52.8 and 54.2.

    Lara's adjusted average is 97.3% of his real average, while for Tendulkar it is 97.4% of the real average. So Tendulkar's has done slightly better than Lara when you consider the adjusted average, and most of the discussions above will be pointless if only you include the not outs ! (the numbers are approximations, i haven't bothered to consider the second decimal point)

  • Ravi Kumar Putcha on August 24, 2008, 14:03 GMT

    Here is the problem with this measure. Take, for example, THAT Kolkata test match. India were bowled out for 171 in the first innings in a mere 58 innings. So when you adjust averages, you get the batsmen getting 1.19 times the runs they actually scored. But it gets interesting now. In the second innings, India batted 178 overs for 657. That would not be too bad, BUT 21 of those overs were bowled by Ponting, Afghanistan, Hayden and Langer. It means that the runs they scored now get adjusted downwards! So Laxman's 281 becomes 277 which, without those 21 would be 315!! And if the measure does punish the batsmen for a good innings - and that is as good as it gets - and rewards them for a poor showing, there has to be something wrong with it. Would be interested to know what Barry thinks of that.

  • no_quiero on August 24, 2008, 13:33 GMT

    Tendulkar in recent times have struggled in multiple series against England, South Africa, Sri lanka, New Zealand, to an extent against Pakistan and scroring very heavlily against Banaladesh.

    As far as Australia are concerned, Tendulkar have recently had a good series. But I doubt he have scored against best Australian attack in the past. In 1998 which is supposed to be one of the best Sachin Series again Australia, the lead Australia bowlers were like Adam Dale, Kasprowicz, Gavin Robertson. Harldy a world class attack with just one Shane Warne. I would love to see Sachin's record against Australia when Glenn Mc Grath was playing.

  • Sagar on August 24, 2008, 11:39 GMT

    Doesn't win games- Avg in games won by India is (shockingly!) ~8 runs higher(vs. career avg) compared to Ponting's hike of ~4 runs in games won by Australia and same as Lara's (~8 runs). Away average in winning games (81) compared to his home average in winning games (54)

    Admittedly, Tendulkar didn't do well against SA (38.81), and his second innings average is not as great. But pressure is involved in setting a target to win games as well. No one is perfect and is bound to have some things going against them in a long and distinguished career.

    Accounting for Tendulkar efforts played against best bowlers in his generation (read Akram and Younis, Donald, Ambrose, Walsh, McGrath, Warne, Lee, Pollock in their prime) in different conditions and immense expectations he played under most of his career with a bowling lineup which was very weak in away games and sheer joy of his exhilarant shot making. All this without considering his ODI exploits! *Stats used from Cricinfo's Statguru.

  • Sagar on August 24, 2008, 11:38 GMT

    Doesn't win games- Avg in games won by India is (shockingly!) ~8 runs higher(vs. career avg) compared to Ponting's hike of ~4 runs in games won by Australia and same as Lara's (~8 runs). Away average in winning games (81) compared to his home average in winning games (54)

    Admittedly, Tendulkar didn't do well against SA (38.81), and his second innings average is not as great. But pressure is involved in setting a target to win games as well. No one is perfect and is bound to have some things going against them in a long and distinguished career.

    Accounting for Tendulkar efforts played against best bowlers in his generation (read Akram and Younis, Donald, Ambrose, Walsh, McGrath, Warne, Lee, Pollock in their prime) in different conditions and immense expectations he played under most of his career with a bowling lineup which was very weak in away games and sheer joy of his exhilarant shot making. All this without considering his ODI exploits!

  • Sagar on August 24, 2008, 11:37 GMT

    Appreciating your attempt to calculate a more meaningful average, there are some serious factors which are not accounted for. For example, what is to say that Hussey scoring against Ashley Giles from England in Australian pacy pitches is more valuable than Tendulkar scoring against good spinners from Bangladesh in spin-friendly tracks.

    Also, a close look at stats show nearly everyone on the list has a signficant drop from their career average. Now for most of the comments focusing on Tendulkar deriving from stats on this list- Everyone cashed in against Ban and Zim (read Lara, ponting and Hussey average nearly 100) but played slightly fewer games. Hussey also played most of his games in Aus (not his fault), and his average in recent WI away series was 21.75.

    Now for Tendulkar bashers, Lets use some stats here- Scoring his runs against minnows- played only 14 of his 148 tests, showed his mettle with 56 avg against Aus in 25 matches cont..

  • shik on August 24, 2008, 10:00 GMT

    Whats with all the tendulkar bashing going on, have some decency and just for a second appreciate what he has done. To someone who said he has had only one good series against australia perhaps you are forgetting 1990-91 tour, 99-00 tour, 2003-04 tour or the recently finished tour (all away series, donot get me started on the home series'). I donot want to give stats as they are readily available on the website but please people do some research before accusing a great cricketer. As for the author of the piece, I understand where you are coming from and the tendulkar headline will always get people to read your article which is fair enough, thats your job.

  • A Hams on August 24, 2008, 9:36 GMT

    Ravi, I'd be interested to know how you came to think that Pietersen has a poor record against Australia - I make it three meaty centuries in ten tests against them so far and counting. Also scored heavily against South Africa this summer... and the fact that the Australian batsmen never faced Mcgrath and Warne may explain why their weighted averages fall from their normal ones more than some other players on the list (as they are still fine players so the weighted average remains pretty high).

    Also on the Tendulkar debate - there seems to be a few disgruntled Indians on here bemoaning his lowly standing in the adjusted table. Don't worry guys, he'll go down as one of the finest batsmen of the modern age and rightly so - second only to Lara in his generation in my opinion. He's scored runs against everyone, everywhere in his long career. A mark of a true great, and in such style as well.

  • Abhijit Kulkarni on August 24, 2008, 8:12 GMT

    Good point prashanth !

    Even an adjusted average will not even out things on many occasions. Was facing a fiery waqar on debut ( sachin in 1989) easier than facing a past his prime waqar ( ponting in 2002)?

    Was facing allan donald at peak (sachin in 92) easier than facing allan donald near the end of his career ( ponting in 2001-2002) ?

    Home and away conditions do make a significant difference too.

  • P.Satish Kumar on August 24, 2008, 7:30 GMT

    Of course, the Tendulkar bashers are out in full force. However taking their argument forward that Tendulkar is rightly lowly-rankedas he scores his runs against B'desh and Zimbabwe, how does it explain Sehwah not being high on that list? He does not have single century against those two nations.

    So he scored his runs against weak attacks like Hoggard, Harmison, McGrath, Lee, Warne, Bond, Pollock, Bond, Murali, Ntini, Akhtar, etc.?

  • Ravi Kumar Putcha on August 24, 2008, 7:21 GMT

    As kriparam so rightly pointed out, it is strange how Ponting and Hussey figure in this list given they never played McGrath or Warne. That alone sort of invalidates a somewhat charlatanish measurement. Also, how the likes of Yousuf, KP etc, who struggled against Australia score better than SRT, Dravid and VVSL, the last of whom has an awesome record against Australia is puzzling. All in all, yet another poorly constructed measure full of holes.

  • Phat on August 24, 2008, 7:12 GMT

    Ankit - what about the fact that unless they bat for 30 or 40 overs, the openers will generally have to face very little spin from the likes of Warne, Murali, Kumble, Bhaji, Kaneria, etc or any tweaker you care to name?

  • V. B. Varrier on August 24, 2008, 6:34 GMT

    I do not know about the accuracy of the analysis. But, Tendulkar's position is not surprising at all. If you split his career averages against different teams, you will find that he has fared better against weak bowling teams like Zimbabwe, Bangladesh, etc. His average against stronger bowling teams like South Africa, Pakistan dips very low. The only exception is Australia. But, then I wouldn't credit Australia with the best attack in recent times. I mean the last decade or two. Their success has been largely due to their batsmen and overall disciplined performance in the field. McGrath and Warne notwithstanding. In fact, I would put the entire fab four much below our stalwarts of yesteryears like Gavaskar, Mohinder, Vishwanath, etc. who had to put up with real world class bowlers like the famed wEST iNDIANS AND aUSTRAILIANS OF THEIR TIMES.

  • Vinodh on August 24, 2008, 6:16 GMT

    Simply put it dude... He is the most useless batsman of any era... 11877 runs in 244 innings and not more than 5 match winning Innings.. No analysis required.. Just get into the runs scored in the second innings scores by him..

  • Vinodh on August 24, 2008, 6:15 GMT

    Simply put it dude... He is the most useless batsman of any era... 11877 runs in 244 innings and not more than 5 match winning Innings.. No analysis required..

  • Sivakumar A on August 24, 2008, 5:44 GMT

    As usual, an Indian fan would look at Sachin's position in any such list. In this one he is bunched along with richards, border, miandad, gavaskar and hammond, which sounds just about right. it is definitely not bad company.

  • Prasanth Nottath on August 24, 2008, 5:35 GMT

    That's probably a statistical flaw with the adjusted average. I guess Sachin's lower average is a result of him having simply played and scored more runs against weaker teams than say Ponting. To summarise it, one would feel that Sachin is penalised for having a higher percentage of his career runs against weak teams, due to reasons beyond his control (the tour itinerary decided by ICC and the boards) while he has done a very good job against the best in the business, i.e Australia. This is precisely what I have tried to demonstrate thru my example in the earlier post.

    So the fall in Sachin's average does not mean really anything at all to me.

  • hylian lynk on August 24, 2008, 4:23 GMT

    The whole problem with this discussion is that no stat can define a sportmans' value to a team. Lets look at Lara, I saw McGrath reduce him to a quivering mess in the 1996-97 series (Lara was averaging 60 at the start of the series), I saw him coming in to bat at #5 in the order in 2000 after McGrath had crushed his confidence. No stat can quantify the human effect of sport so 50.4 has no value in certain situations. I find it pointless to use any stat other than average to even remotely try to ascertain a players value. We simply cannot quantify "quality " of players or pitches. I have always held the view that batsmen who played on uncovered pitches were more valuable than batsmen on covered wickets. I had the opposite opinion of bowlers for obvious reasons. McGrath vs Lara at Adelaide in the modern era is fine lots of runs but what about McGrath vs Lara at Adelaide after a rain storm on August 24 1928 !

  • kriparam on August 24, 2008, 3:24 GMT

    Well Mike hussey , Ricky ponting have never faced mcgrath, brett lee or shane warne. how is that they figure at the top of the list.

  • David Barry on August 24, 2008, 2:10 GMT

    For those mentioning Viv Richards, remember that his career average was only a bit over 50. He's not going to feature in the top ten in any measure that looks only at averages. His brutal dominance over bowlers is reflected in his strike rate, which is one of the highest of all time in Test cricket (a bit under 70).

  • David Barry on August 24, 2008, 2:08 GMT

    Further on Prasanth's comment:

    It's possible to have the reverse situation.

    Suppose John scores 680 runs in 10 innings against Australia (bowling avg 27), and 450 runs in 10 innings against Bangladesh (bowling avg 40). Then his regular average is 56.5 and his adjusted average 57.7.

    Ram scores 650 runs against Australia in 10 innings, and 200 runs against Bangladesh in 5 innings. His regular average is 56.67 and his adjusted average 54.1.

    John averages more than Ram against each team, but his overall regular average is lower. His adjusted average is always higher, though.

    This is an example of Simpson's paradox.

  • JE on August 24, 2008, 1:44 GMT

    Maybe that explains why Tendulkar's average against Australia (arguably the best consistent bowling attack in the last 10 years) is higher than career average. Any statistical method should be tested for true positives, and any list that doesn't mention Viv Richards flunks this golden standard test. (Mike Hussey, though a great player, hasn't even faced Alan Donald!)

  • David Barry on August 24, 2008, 0:08 GMT

    Vidhya, that is a good point. It would probably be better to do another iteration of this procedure - using the bowlers' adjusted averages (which would fix up Briggs/South Africa problems).

    Prasanth Nottath: Wow. That is really interesting, and I'm going to have to think about it for a bit. It seems a bit like Simpson's paradox.

  • knight on August 23, 2008, 23:52 GMT

    Tendulkar records have been bloated by runs against minnows and this analysis just confirms that. Apart from one good series against Australia he have failed against most of the series in last four years baring Bangladesh. In the last two series against Bangladesh his average is staggering 284 and 127. His two double hundred 248 not out and 201 not out has also come against Bangladesh and Zimbabwe.

  • Tristan Haddow-Allen on August 23, 2008, 21:17 GMT

    Perhaps the statistics for the 'modern greats' are affected by the weight of runs they've all scored against the weak attacks of Zimbabwe and Bangladesh? They can only bat against who's put in front of them, but it makes the runs they've scored against strong attacks an artificially small proportion of their aggregate.

  • The Tooting Trumpet on August 23, 2008, 20:21 GMT

    Great stuff Dave.

    Only Greg Chappell from the 70s and 80s, to me the hardest era to bat against brutal pace with limited protection. Makes him look very good indeed!

    The current list looks about right to me. I expect KP and Punter to swap places next year.

  • Prasanth Nottath on August 23, 2008, 19:19 GMT

    Interesting.

    One curious thing i noticed is that the practise of applying the ratio to runs scored, could lead to sometimes ironical situation. Example- John averages 50 against Australia, whose average bowling average is 27 (and have been maintained throughout), in 10 innings (no notouts). He averages 40 in 5 innings against Bangladesh, whose average bow avg is 40, again no notouts. So ordinary bat average of John is 46.67 and Adjusted average is 49.39.

    Ram has played 10 innings against Australia, and scored 530 runs in 10 innings, and 450 runs against Bangladesh in 10 innings. Again no not outs. So Ram's ordinary average is 49. However his adjusted average is 48.64.

    If we take adjusted averages alone, then we might be tempted to conclude John is a better bat than Ram, which is preposterous, considering that Ram averages higher than John against both teams.

  • Ariz on August 23, 2008, 19:12 GMT

    Dear Barry, "Tendulkar's low position is a bit of a surprise"

    Well, I am surprised at you. I guess you havn't looked at the names of the players in the list above Sachin! May be, as in your own words "perhaps we should pay more attention to Neil Harvey when he compares modern players to those of his day"

    This clearly goes on to show that the cricket history is vast. These stats are in no way a reflection of a players capability, but it nonetheless portrays a better picture of how they fared in tests. The list shows that Ranji was the greatest batter from Asia, I guess I would live with that. Although I have been following cricket for a bit more than 20 years, but have read and seen videos of players since 70's and to me the order is Viv Richards, GS Chappell, B. Lara, Sachin, in that order, next one I am not sure (PS I havent seen much of Barry Richards or G. Pollock, so I am not counting them here, as well as Sobers)

  • Sidharth on August 23, 2008, 18:37 GMT

    While the post is good and insightful , it lacks one major factor in the calculation of adjusted avg's . Now you may say Kumble's weightage is some number , but opne has to factor that kumble at home is much much more lethal than kumble away , similarly scoring runs against Shane Bond in NZ has to be given more weightage than scoring runs against him in dead pakistan pitches , so overall the comparison of a batsman's home and away records excluding Bang and Zim will be a much better comparison. Bowlers like Kumble , Murali , may have a high weightage but look at their records in Australia , England and south Africa , surely you cant say a batsman who scored a 100 against murali in brisbane has equal weightage with a batsman who scored a 100 against murali at Kandy?

  • marcus on August 23, 2008, 17:26 GMT

    Oh dear, Tendulkar's low position will surely generate many angry comments! Anyway, when I read the beginning of this article, I saw what I think may be a flaw. Weighting a bowler's average by the batsman's average is effectively weighting it by the batsman's value, which I am not convinced is equivalent to weighting them by the difficulty of dismissing them. How about weighting it by the average amount of balls they take to be dismissed (the batsman's balls/outs)? Regarding the main body of the article, I don't see any flaws beyond the obvious or mentioned. Good stuff.

  • riverlime on August 23, 2008, 17:26 GMT

    I have to commend you on your insightful analysis. However, as for not outs, I disagree. They DO affect averages significantly. Considering all inings as completed when they leave the pitch (since they can't continue adding to their score), Lara's Runs per inning (RPI) would be (11953/232)=51.5, instead of 52.9. Ponting would be (10099/199) =50.7, instead of 58.5. Kallis would be (9761/209) = 46.7, instead of 55.4. Tendulkar would be (11877/244) = 48.7, instead of 54.2. Except for Lara, that's an average of 6-8 points drop for each batsman, and reflects their true scoring potentials.

  • Vidhya on August 23, 2008, 16:52 GMT

    Another minor point is what I would call the "Sydney Barnes effect". Barnes' bowling average against Australia was 21, but his average of 9 against South Africa brought down his career average down to 16. So the adjusted averages of Clem Hill, Victor Trumper and others would have received a minor boost because of the two series that Barnes played at the end of his career, which should have really been irrelevant to the calculations. Considering the "career-to-date" would help in the case of Barnes, but not with, say, Johnny Briggs.

    This is much less significant in modern times. But England and Australia play each other far too often than the rest. For bowling averages of the English bowlers would probably have been significantly better if they played Australia less often. My uneducated guess is that the Barnes effect would make a difference of 2-5 points for several batsmen.

  • vin on August 23, 2008, 16:50 GMT

    When you compute the "average average" for an innings, how about basing the calculation on the harmonic mean rather than the arithmetic mean? In your given example this would be (30+30+25+20)/((30/28)+(30/30)+(25/35)+(20/40)). This would change the nature of the biases in the system, and would eliminate the need to round off the worst bowlers to 100.

  • Raghav Bihani on August 23, 2008, 16:11 GMT

    Just a Suggestion. I agree that it is very difficult to account for pitch conditions (Dravid's 81 & 68 on a minefield at Kingston), 4th innings chasing pressure (Lara's 153* in Bridgetown against Australia) or back to wall innings (Laxman's 281 at Eden Gardens).

    But some adjustments may me made for Home/Away matches. Say factors of 1.05/0.95 for away/home innings respectively. Also for which innings of the test the runs are scored in like 0.925 for 1st, 0.975 for 2nd, 1.025 for 3rd and 1.075 for 4th.

  • AG on August 23, 2008, 16:03 GMT

    Hmm.. I dont see Viv Richards anywhere in the listings in the btext of the article ...May it is is just that I am biased but any list without the king must be missing something..esp if the list ranks Hussey at the top ...i wonder how he would have fared facing the 80's and 90's west indian bowlers at test level.

  • Ankit on August 23, 2008, 15:32 GMT

    "Opening batsmen face the opening bowlers disproportionately often, and this isn't taken into account."

    That explains why Tendulkar is rated so down. The best batsman faces the best bowlers and this fact is overlooked. This is a truely incomplete analysis.

  • Raghav Bihani on August 23, 2008, 15:31 GMT

    It would be interesting to note the difference in the adjusted and actual averages. While it is 7 for Tendulkar its only 3 for Lara. Thus the falls in averages are not merely adjustment for the era when one played.

    Tendulkar's low average is a true reflection because a large number of his tests have been against Zimbabwe and Bangladesh compared to Lara. Thus his runs have a lower value than those scored by Lara against quality opposition.

    Only in case of current Australians, I get the feeling that they are being penalised for having the best Bowling Attack and hence always facing relatively weaker bowlers.

  • StatisticsFan on August 23, 2008, 15:24 GMT

    Tendulkar's low position is not a surprise. This is quite evident from India's poor test record overseas and lack of World Cup win during Tendulkar's time.

  • No featured comments at the moment.

  • StatisticsFan on August 23, 2008, 15:24 GMT

    Tendulkar's low position is not a surprise. This is quite evident from India's poor test record overseas and lack of World Cup win during Tendulkar's time.

  • Raghav Bihani on August 23, 2008, 15:31 GMT

    It would be interesting to note the difference in the adjusted and actual averages. While it is 7 for Tendulkar its only 3 for Lara. Thus the falls in averages are not merely adjustment for the era when one played.

    Tendulkar's low average is a true reflection because a large number of his tests have been against Zimbabwe and Bangladesh compared to Lara. Thus his runs have a lower value than those scored by Lara against quality opposition.

    Only in case of current Australians, I get the feeling that they are being penalised for having the best Bowling Attack and hence always facing relatively weaker bowlers.

  • Ankit on August 23, 2008, 15:32 GMT

    "Opening batsmen face the opening bowlers disproportionately often, and this isn't taken into account."

    That explains why Tendulkar is rated so down. The best batsman faces the best bowlers and this fact is overlooked. This is a truely incomplete analysis.

  • AG on August 23, 2008, 16:03 GMT

    Hmm.. I dont see Viv Richards anywhere in the listings in the btext of the article ...May it is is just that I am biased but any list without the king must be missing something..esp if the list ranks Hussey at the top ...i wonder how he would have fared facing the 80's and 90's west indian bowlers at test level.

  • Raghav Bihani on August 23, 2008, 16:11 GMT

    Just a Suggestion. I agree that it is very difficult to account for pitch conditions (Dravid's 81 & 68 on a minefield at Kingston), 4th innings chasing pressure (Lara's 153* in Bridgetown against Australia) or back to wall innings (Laxman's 281 at Eden Gardens).

    But some adjustments may me made for Home/Away matches. Say factors of 1.05/0.95 for away/home innings respectively. Also for which innings of the test the runs are scored in like 0.925 for 1st, 0.975 for 2nd, 1.025 for 3rd and 1.075 for 4th.

  • vin on August 23, 2008, 16:50 GMT

    When you compute the "average average" for an innings, how about basing the calculation on the harmonic mean rather than the arithmetic mean? In your given example this would be (30+30+25+20)/((30/28)+(30/30)+(25/35)+(20/40)). This would change the nature of the biases in the system, and would eliminate the need to round off the worst bowlers to 100.

  • Vidhya on August 23, 2008, 16:52 GMT

    Another minor point is what I would call the "Sydney Barnes effect". Barnes' bowling average against Australia was 21, but his average of 9 against South Africa brought down his career average down to 16. So the adjusted averages of Clem Hill, Victor Trumper and others would have received a minor boost because of the two series that Barnes played at the end of his career, which should have really been irrelevant to the calculations. Considering the "career-to-date" would help in the case of Barnes, but not with, say, Johnny Briggs.

    This is much less significant in modern times. But England and Australia play each other far too often than the rest. For bowling averages of the English bowlers would probably have been significantly better if they played Australia less often. My uneducated guess is that the Barnes effect would make a difference of 2-5 points for several batsmen.

  • riverlime on August 23, 2008, 17:26 GMT

    I have to commend you on your insightful analysis. However, as for not outs, I disagree. They DO affect averages significantly. Considering all inings as completed when they leave the pitch (since they can't continue adding to their score), Lara's Runs per inning (RPI) would be (11953/232)=51.5, instead of 52.9. Ponting would be (10099/199) =50.7, instead of 58.5. Kallis would be (9761/209) = 46.7, instead of 55.4. Tendulkar would be (11877/244) = 48.7, instead of 54.2. Except for Lara, that's an average of 6-8 points drop for each batsman, and reflects their true scoring potentials.

  • marcus on August 23, 2008, 17:26 GMT

    Oh dear, Tendulkar's low position will surely generate many angry comments! Anyway, when I read the beginning of this article, I saw what I think may be a flaw. Weighting a bowler's average by the batsman's average is effectively weighting it by the batsman's value, which I am not convinced is equivalent to weighting them by the difficulty of dismissing them. How about weighting it by the average amount of balls they take to be dismissed (the batsman's balls/outs)? Regarding the main body of the article, I don't see any flaws beyond the obvious or mentioned. Good stuff.

  • Sidharth on August 23, 2008, 18:37 GMT

    While the post is good and insightful , it lacks one major factor in the calculation of adjusted avg's . Now you may say Kumble's weightage is some number , but opne has to factor that kumble at home is much much more lethal than kumble away , similarly scoring runs against Shane Bond in NZ has to be given more weightage than scoring runs against him in dead pakistan pitches , so overall the comparison of a batsman's home and away records excluding Bang and Zim will be a much better comparison. Bowlers like Kumble , Murali , may have a high weightage but look at their records in Australia , England and south Africa , surely you cant say a batsman who scored a 100 against murali in brisbane has equal weightage with a batsman who scored a 100 against murali at Kandy?