Anantha Narayanan

The tenth edition of the World Cup is not far away. In my last article I looked at the previous 9 World Cups from an alternate analysis point of view. In this article I will look at the ensuing World Cup, similarly from an analytical point of view.

For this article I have used some of the results of the proprietary work on recent form and match simulation I have done for a company which has a web presence exclusively for Cricket. As such I will not be presenting any detailed tables but refer to the conclusions extracted from those. I request readers to kindly bear with me. There are contractual restrictions to be observed.

The cornerstone of this analysis is the Team Strength index (TSI) of the participating teams. This index has been extracted using a complex process which involves the following.

- An estimated Final XI.
- The career figures of selected players.
- The recent form of players, bowling and batting.

The base TSI is determined using the following data.

- Career Runs per innings, after excluding single-digit not outs.
- Career Strike rate.
- Career Bowling strike rate- Balls per wicket.
- Career bowling accuracy - Runs per over.
- Recent Form Runs per innings, after excluding single-digit not outs.
- Recent Form Strike rate.
- Recent Form Bowling strike rate- Balls per wicket.
- Recent Form bowling accuracy - Runs per over.

The teams have been selected by me based on the assumption that ALL the players would be available and would be injury-free. Key players such as Tendulkar, Sehwag, Gambhir, Ponting, Collingwood, Bresnan, Kallis, Vettori and a few others have injury concerns. If these players do not play, the numbers would change as also the predictions. Mike Hussey and Eoin Morgan are already out. The teams concerned have already gone down in strength.

The recent form encompasses the last 10 innings played or last 10 spells bowled, provided these have been played on or after 1 January 2009. Most players have done this during 2010 itself. However someone like Tendulkar has played 2 innings in 2011 and 2 in 2010 and I had to go back one more year. 2009 performances are weighted slightly lower. Runs per Innings makes more sense than Batting average especially since recent form has to be considered. 3 not outs out of 10 distorts the recent form a lot. The weight for recent form numbers as against career numbers is decided based on the number of career matches played. The career figures have the highest weight of 75% if the player has played more than 100 matches. The recent form numbers become more significant if the player has played fewer matches.

After the base TSI is determined it is further adjusted based on the following three factors.

- The recent form of the teams - results, margins, venues et al.
- The venue of the matches (to decide on home/familiarity measures).
- The team performances in the two recent World Cups (2003 and 2007). This will enable us to assign due weight to the manner in which big-matches and big-stage have been handled. Going back beyond 2003 is not right since most of those players would have retired.

Recent form of teams: The recent form of teams considers the last 20 matches played by the team, provided these have been played on or after 1 January 2009. Most teams have done this during 2010 itself. The results are analyzed from results (wins/ties/losses) point of view and the match venue (home/neutral/away) point of view. In addition the margins of wins is incorporated. Finally, unlike the batsmen/bowler recent form calculations, the team numbers are determined based on the recent matches only. Needless to say, the West Indian successes of the 1970s should have no bearing on the chances of today's West Indies. The form related changes work out to around 3% on either side of 100.

Match venue factor: India, Bangladesh and Sri Lanka have been assigned a 2.5% benefit value for playing at home. Pakistan has a 1.25% benefit for playing in familiar sub-continent conditions.

Recent World Cup form: Upto 2% is allotted for this factor. A simple one based on the performances of teams in the recent two World Cups. Australia gets 2% (2 wins), Sri Lanka gets 0.75% (One final and one semi-final), India gets 0.50% (one final), New Zealand gets 0.35% (one semi-final and one super-six) and so on.

These numbers might seem arbitrary. However these have been arrived at after lot of trial runs. Moreover the benefit cannot be made greater than these since these numbers are used for simulation which is very sensitive to these numbers.

The recent form tables are listed below. It should be noted that the actual match team strength of India in its inaugural match against Bangladesh will be slightly lower since they would be playing "away" and Bangladesh would be playing at "home". Similarly actual match team strength of Sri Lanka in its match against New Zealand will be very slightly lower since both of them would be playing at "neutral" venue.

Team         BtIdx BwIdx TmIdx LocAdj RF_Adj Wc_Adj BtIdx BwIdx TmIdx

The tenth edition of the World Cup is not far away. Over the next 10 days, I will first do a review of the nine World Cups so far and then a preview of the World Cup 2011. I will try to do something different to what is already available easily in public domain.

If you wanted to know the number of runs scored at what average or wickets captured with attendant details, you only have to go to the excellent Cricinfo World Cup section or peruse Madhu Ramakrishnan's excellent article in this blog. One click will let you know that Tendulkar has scored 1796 runs at 57.93 or that Muralitharan, with 53 wickets, needs 19 wickets to overtake McGrath and so on. What I have attempted to do is to add some weight to these runs and wickets. Also I wanted to do a different type of Team performance analysis.

The 1996 edition had a format exactly identical to the forthcoming world cup. 2 groups, quarter finals, semi finals and final. This is also a very good format since it requires teams to have three consecutive wins to win the world cup, not two. The only difference is that the 1996 edition had 12 teams and the 2011 edition will have 14 teams. The 1999 and 2003 editions had preliminary league, super-six group, semi finals and finals. Both were similar except that more teams participated in the 2003 event. There was yet another change for the unwieldy 2007 event. There was a preliminary 4-group league, a super eights, semi finals and final.

The reason I have taken the trouble of describing the formats is to clarify my weighting process. Despite the format variations, there has been a common feature at the start and end of the tournaments. There have been preliminary leagues at the start and semi finals and final at the end. In between, we have had super sixes, super eights and quarter finals. So I have decided to weight the matches in the following groups to do a weighted runs/wickets/team performances analysis.

The weighting of performances is done based on the WHEN factor since the matches become more significant as we come to the later stages. As far as opposition quality is concerned, it is my firm opinion that this becomes irrelevant in a knock-out match. There is no way I am going to treat Tendulkar's 83 or Ganguly's 111 scored in the 2003 semi final at a lower level, since their innings helped India reach the final. Else India might have lost to Kenya.

As far as preliminary matches are concerned, I am not going to lower the weight for matches against weaker teams. Every match in a World Cup is important. India and Pakistan in 2007 WC are living examples of being derailed by losing the preliminary matches against weaker teams. So all preliminary matches will have a weight of 100%.

Preliminary league matches: 100%
Super six matches:          110%
Super eight matches:        110%
Quarter-Finals:             112.5%
Semi Finals:                125%
Finals:                     150%
Note: The Quarter Finals have a slightly higher weight because of the knock-out 
nature of the concerned match.

I am sure readers would point out that there have been important matches in the preliminary leagues and deserve a higher weight. However I am not going to take that route. Then other questions will come in, especially for early matches. How important was Kapil Dev's 175 or Tendulkar's 98 ? What would have happened if India had lost ? Did they still have a chance ? So many other match, group and tournament related conditions would have to be considered. There is also no way to automate these factors. Each match has to be considered manually. Hence I have taken a reasonably sound weight pattern. Maybe at the end of the 2011 World Cup I would probably do a more in-depth analysis of the 10 world cups, incorporating a few more relevant factors also, including match status, support, bowling quality and importance of matches.

The team performance analysis is done in two ways. The first is a straight-forward analysis of the matches played, wins achieved, no-result and lost matches and does a simple % performance achievement. Let us see this table first.

Team         Mats  W  NR   L  <-Points->    %
Max   Base

This analysis is based on a request by Alex who wanted me to do an analysis of the Test performances by innings. It is a straight-forward analysis based on raw numbers. Please take this as a break up of the Career performances into lower levels with no adjustment whatsoever. The innings status at entry, match conditions, match location, quality of bowling, quality of opposition team et al are relevant factors but have not been incorporated. Once I open one door, the draught will open all the other doors and I do not want to do that. There are a number of tables shown. These tables are provided with minimal comments. The top-20/10/5 entries are shown in the main article and the complete tables are made available for viewing/downloading.

First the Team innings tables. For the Team innings, the cut-off is 3000 career runs. In addition to the batting average and runs scored tables, I have one on the comparison ratio to the career batting average. This table will indicate how close or away from their career averages have the batsmen performed in different innings and will give an insight into whether the batsman has excelled in setting up or finish the matches. Both are important but we need this insight to get a proper handle on batsmen appreciation.

1.1. Team First innings (Match inns 1/2) analysis: Table ordered by Batting average

Batsman             Inns No  Runs   Avge

This is an important classification since it removes the distinction between first/second and third/fourth innings. Bradman just about misses the 100 mark. Note Sehwag's near-70 average. Also how the two modern greats, Tendulkar and Lara are separated only in the second decimal point. Jayawardene is the other modern batsman to appear in the top-10.

1.2. Team First innings analysis: Table ordered by Runs scored

Batsman             Inns No  Runs   Avge

Abhi had a very unusual request. He wanted me to find a way of determining the number of hours Tendulkar was on field during his international career. This was such an unconventional request that it intrigued me a lot. I set about the methodology, which turned out to be quite a bit tricky. Then, as I am wont to do, I decided to expand the scope of the exercise to all the players. That complicated the exercise further. But the results are fascinating.

I decided to use the balls faced, bowled, kept wicket or fielded as the base for all calculations. This is the most consistent measure and is available completely for most recent matches. I had to do some extrapolating for earlier matches, especially for the balls faced by the batsmen. The bowler balls and the fielder balls are, however, always available.

I decided to split the information into broadly two categories. One is the Active Balls, during which time the player has to be 100% aware and involved all the time and the other is the Passive Balls, during which the player is not at the centre of action. The bowler, batsman, batsman at the non-striker end and the wicket-keeper are always at the centre of action.

Active Balls refer to the following types of activities and can be added with reasonable amount of conviction..

- Batsman facing the bowling directly,
- Batsman at the non-striker end,
- Bowler delivering the ball and
- Wicket-keeper engaged in keeping.

Passive Balls refer to the following type of activities.

- Player fielding.

A few footnotes.

1. It is accepted that the two Batsman-related activities are not necessarily equivalent. In one, the batsman faces the ball, plays the same and then runs, if required. In the other, the player only runs, if required. However these are quite integrated activities and I did not want to affix an ad-hoc % of involvement level. Instead consider this as two sides of one activity, one requiring 120% effort and the other, 80% effort. Since the two numbers are almost the same for most players, this works out very well.

2. The bowler effort is split further into the fast bowler effort, requiring more effort and the spinner deliveries, requiring lesser effort. This takes care of the additional run-up required.

3. The wicket-keeper effort is valued as equivalent to the batsman-at-crease effort in view of the intense effort that is required before, during and after the delivery.

4. The Fielder effort will be taken in terms of time the same as wicket-keeper effort but will be given in a separate table since this represents much less taxing and lighter work-load. No way can we split this as close to the stumps or outfield because of the non-availability of data. Anyhow the close fielders might have to concentrate more but the outfielders would be running more. So this would about cancel out. Let us also ignore the few minutes players take off to get to the pavilion to have one or more breaks. There is no data on that. In all these cases I have avoided any ad-hoc weights.

So it is clear that there is present a certain amount of guesswork and approximation. However the idea is not to strive for unnecessary accuracy but move forward on the uncharted seas.

I will convert each delivery into time by using 42.5 seconds per delivery. This works out well across the years. Maybe they bowled more overs during the early years. But this works to nearly six and half playing hours per day which is the norm. I am not going to sit here and worry about the West Indian pace bowlers, or for that matter, the recent Indian teams, bowling at 12 overs per hour. They stay longer on the field, by choice.

There will be a slight tweaking for bowlers. The pace bowlers' deliveries would be clocked at 45 seconds and the spinners at 40 seconds. This will even out to the magical 42.5 seconds over the millions of deliveries bowled. It is possible that the fast bowlers expend more effort during their bowling efforts than the spinners. However I have decided that the additional seconds assigned will compensate for this partly. I do not want to assign any arbitrary effort related weight. The keepers remain at 42.5 seconds which evens out to their keeping for both types of bowlers.

I will do three tables. The first will be one on the batting and bowling times of players, other than wicket-keepers (who have kept wicket for a minimum of 10000 balls). The second will be on the fielding hours spent by these players. The third will be a special one for wicket-keepers on their keeping and batting duties. All the tables will cover players who have played in 25 or more Test matches or 40 ODI matches. These are long lists with around 500 players. Hence only the top-20 will be shown in the article.

Tests: Non-keeper Player - Batting/Bowling hours table

Player (Bat/Bow) Team Mats  <--Total--> <-Bat:Act-> <-Bat:Pas-> <-Bowling->
Hours Hrs/M Balls Hours Balls Hours Balls Hours

The top-3 players based on the hours/match value are given below.

T.W Goddard  (Saf) : 6.99
G.S.A Sobers (Win) : 6.87
V.M,Mankad (Ind)   : 6.84

Tests: Wicket keepers - Keeping/Batting hours table

Player (Keeper)  Team Mats  <--Total-->  <-Keeping->  <-Bat:Act-> <-Bat:Pas->
Hours Hrs/M  Balls Hours  Balls Hours Balls Hours

No one ran up ODI bowling streaks like Waqar Younis did © Photosport

This turned out to be a tough task since I also wanted to utilise this opportunity to build a player-performance database. This is essential since I needed to get the best 1-10 ODI performances for each player and then get the all-time best performances. I also wanted to provide the information on the top players' 1-10 ODIs best performances so that the readers could do their own comparisons. And I was sure that there would be queries on the best performances by specific players after the article was published. I wanted to be able to provide the information quickly. In fact I have also provided the huge table of all qualifying players for downloading.

First let me emphasise that this is only a run aggregate. I myself will clarify that this aggregating of runs in specific sequences of 1-10 ODIs is irrespective of opposing team, home or away, match conditions, period lapsed between matches, not outs et al. That is not the purpose of this article. Readers should appreciate this and not come in with a comment such as "opposition bowling quality is not considered". But that is wishful thinking! Anyhow I will just publish such readers' comments without any response. Also readers who worry about batting average should understand that when someone scores over 500 runs in 10 ODIs, it does not matter about averages. It is going to be quite high.

Let us now look at the tables.

Maximum runs scored in 1 to 10 ODIs

Batsman              Cty StMtId-Year  No Runs
ODIs

Recently Unnikrishnan had suggested a way of measuring individual Test innings in a different manner. His suggestion was that the innings should be evaluated against the average score of the other batsmen who batted in that particular innings. He also wanted the individual innings values summed across all innings for each batsmen and averaged across their career, similar to the way the Batting Averages are calculated. For instance, to compute Don Bradman's career Innings Index value, the Innings Index values for all his 80 innings would be added and divided by 80. These are excellent suggestions in view of the following plus factors.

- This is an out-and-out peer comparison, that too within the same team.
- The comparison is within the same innings: Hence the conditions would be almost identical.
- The bowling quality faced would be almost identical, barring innings-level variations.
- This effectively takes take care of the oft-repeated complaints by readers regarding batsmen playing in weak or strong teams.

In some ways this is similar to the simpler % of Team score measure. However the one major difference is that in the % TS measure the number of batsmen who batted is not taken into account. "For no loss" and "for 7 wkts" will produce the same % TS, as explained in the examples. However the Innings Index takes care of this very well and is a true peer comparison. Team score, as given below, is sans extras.

Team-Score Batsman-Score  % TS   Inns Index

For the bowlers I have aggregated wickets which are the most important acquisitions which any bowler can have. The averages really do not matter too much since wickets are the means to win in Test cricket. I have also made sure that the few-Test wonders like Sivaramakrishnan and Massie find their deserved place in these tables.

This also turned out to be a tough task since I had to create a player-performance database. This is essential since I needed to get the best 1-10 Test performances for each bowler and then get the all-time best performances. I also wanted to provide the information on the top bowlers' 1-10 Tests best performances so that the readers could do their own comparisons. And I was sure that there would be queries on the best performances by specific bowlers after the article was published. I have also provided the table of key bowlers for downloading.

First let me emphasize that this is only a wicket aggregate and will clarify that this aggregating of wickets in specific sequences of 1-10 Tests is irrespective of opposing team, home or away, match conditions, period lapsed between matches, quality of batsmen dismissed et al. That is not the purpose of this article. Readers should appreciate this and not come in with a comment such as "opposition batting quality is not considered". But that is wishful thinking! Also readers who worry about bowling average should understand that when someone captures over 50 wickets in 5-6 Tests, it does not matter about averages. It is going to be quite low. This is consistent with my stand on the comparable Batsmen analysis.

Let us now look at the tables.

Maximum wickets captured in a single Test

Bowler           Cty Wkts  StTest/Year

This completes the quartet of anecdotal articles on the "second" innings batting and bowling performances. The "first" innings does not offer this level of variations and I will refrain from looking at the first/second innings performances.

The third innings bowling performances broadly fall into three categories.

1. In this case the third bowling team is bowling with a substantial deficit and only a truly great performance can help them win. About 4 such performances find their place in this list.

2. The second category is one in which the teams have finished within about 10% of each other's score and a powerful bowling performance helps the team to win. About 7 bowling performances belong to this category. There are obviously many more such performances but only the outstanding ones are selected.

3. Matches in which the third bowling team enjoys a huge first innings lead and the bowlers bowl with this cushion. It is immaterial whether this is a follow-on innings or not. The bowlers would have to come in with once-in-a-lifetime performances to get selected in this tough list. Only two bowler efforts in this category have been selected.

It is a fact that there could be truly great batting performances which help teams save tests (Hanif, Amiss et al). However normally great bowling performances help their teams win tests. Indeed it is almost impossible to win a test without capturing 20 wickets. However I have managed to locate two wonderful bowling efforts which ended in a draw and loss respectively.

I have the usual second XV. And this time I have earmarked the empty table containing Readers' selections and would fill as comments come in.

These match summaries are presented in the order of the type of matches already outlined. Within the type, the matches are presented in a random order. Do not draw any conclusions from the order of presentation.

First the matches in which the teams came back from a huge deficit in the first innings.

MtId: 1876 Year: 2008 Test# 2 of 3 (0-0) England won by 6 wickets

Nzl 381 all out.
Eng 202 all out.
Nzl 114 all out     (Panesar M.S: 17.0-5-37-6).
Eng 294 for 4 wkts.

England just about avoided the follow-on but were 179 runs in arrears. Anderson dismissed Redmond early but New Zealand recovered to 50 for 1. Then Panesar produced, inarguably, his finest spell for England. He captured the next six wickets and reduced New Zealand to 106 for 7. Sidebottom cleaned up the tail and New Zealand were dismissed for 114, leaving England to get a formidable 294 to win, which they did in style, winning by 6 wickets. One of the best come-back wins ever. Panesar, who has not done much otherwise, engineered this remarkable win.

MtId: 1677 Year: 2003 Test# 2 of 2 (0-0) Pakistan won by 7 wickets

Nzl 366 all out.
Pak 196 all out.
Nzl 103 all out     (Shoaib Akhtar: 18.0-3-30-6).
Pak 277 for 3 wkts.

This match is almost a replica of the first one. Pakistan just about avoided the follow-on but were 170 runs in arrears. Then Shoaib Akhtar produced, almost certainly, his finest spell for Pakistan. He captured six wickets for 30 runs and dismissed New Zealand for 103, leaving Pakistan to get a huge score of 274 to win, which they did in style, winning by 7 wickets. Shoaib Akhtar has done this on other occasions but this was amongst the best ever by this devastating bowler.

MtId: 1503 Year: 2000 Test# 2 of 5 (0-1) England won by 2 wickets

Win 267 all out.
Eng 134 all out.
Win  54 all out     (Caddick A.R: 13.0-8-16-5).
Eng 191 for 8 wkts.

The English bowlers seem to have specialized in these third innings bowling efforts. This time England trailed by 133 runs against an excellent West Indian bowling attack.then Caddick produced a dream spell, capturing 5 for 16, the only 5-wicket spell in this elite collection. West Indies, with Lara, Chanderpaul and Adams were a fair batting side. England struggled to make the 188 needed to win and lost eight wickets. Only Cork's brave 33 carried them to an unlikely win.

MtId: 0692 Year: 1971 Test# 3 of 3 (0-0) India won by 4 wickets

Eng 355 all out.
Ind 284 all out.
Eng 101 all out     (Chandrasekhar B.S: 18.1-3-38-6).
Ind 174 for 6 wkts.

Next the matches in which the teams were more or less equal in the first innings.

England were 71 runs ahead in the first innings and were expected to stream-roller India in this deciding Test match. However, Chandrasekhar produced what is, arguably, the best bowling performance by an Indian spinner abroad. He broke the back of English batting with three top order wickets and then cleaned up the late order, to finish with 6 for 38. England were dismissed for 101, leaving India to score 173 to win. This was not easy, but was achieved by contributions from all batsmen. India had won their first test series in England.

MtId: 1687 Year: 2004 Test# 1 of 4 (0-0) England won by 10 wickets

Win 311 all out.
Eng 339 all out.
Win  47 all out     (Harmison S.J: 12.3-8-12-7).
Eng  20 for 0 wkts.

After two matching first innings, England had a lead of only 28 runs. It was expected that West Indies would wipe out this lead quickly and go for a win. Instead Harmison produced a vintage fast bowling performance, capturing 7 for 12 in 12 overs of devastation. West Indies lost half their side wiping the deficit and were finally dismissed for 47, their lowest total in Test cricket. One of the great modern day bowling efforts.

MtId: 1673 Year: 2003 Test# 2 of 4 (0-0) India won by 4 wickets

Aus 556 all out.
Ind 523 all out.
Aus 196 all out     (Agarkar A.B: 16.2-2-41-6).
Ind 233 for 6 wkts.

This was probably the only moment of greatness in Agarkar's otherwise journeyman Test career for India, forgetting the throwing-the-bat 100. Two huge first innings of 500+ left Australia with a lead of 33 runs. Then Agarkar had his few hours in the sun, capturing 6 for 41, including three top order wickets. India managed the chase of 233 reasonably well and India had won a historic test after conceding 556 in the first innings. Agarkar more than made up for his sequence of zeroes.

MtId: 1266 Year: 1994 Test# 3 of 3 (0-1) England won by 8 wickets

Saf 332 all out.
Eng 304 all out.
Saf 175 all out     (Malcolm D.E: 16.3-2-57-9).
Eng 205 for 2 wkts.

Another English bowler !!! After two middling first innings England finished 28 short. Then Malcolm produced one of the greatest ever bowling performances by an English fast bowler, capturing 9 wickets for 57 runs. He captured the first three wickets for no run in 2 overs. Gough captured the other wicket. Malcolm's bowling performance has been bettered by two fast bowlers only, Hadlee and Lohmann.

MtId: 1159 Year: 1990 Test# 2 of 5 (1-0) Australia won by 8 wickets

Eng 352 all out.
Aus 306 all out.
Eng 150 all out     (Reid B.A: 22.0-12-51-7).
Aus 197 for 2 wkts.

After two 300+ scores Australia were 52 runs adrift. England were sitting pretty at 103 for 1. Then Bruce Reid destroyed England capturing 7 for 51 and got them all out for 150, that too on a good MCG batting track. Australia chased the target comfortably by 8 wickets. Bruce Reid was an under-rated left arm pace bowler. His is the only Australian performance featured here, ahead of greats like Warne, Lillee, Spofforth, McGrath et al.

MtId: 1477 Year: 1999 Test# 1 of 2 (0-0) New Zealand won by 9 wickets 1477 1999 Cairns C.L New Zealand Win 28 22.5 10 27 7

Win 365 all out.
Nzl 393 all out.
Win  97 all out     (Cairns C.L   : 22.5-10-27-7).
Nzl   70 for 1 wkt.

New Zealand led by 28 runs. West Indies were nowhere the dominating side they were during the 80s or early 90s. However they were also not the pathetic side of the 2000s. Cairns produce a devastating spell of 7 for 27 which included 4 top wickets. West Indies were blown off for 97 and New Zealand won comfortably. Let us not forget that Cairns also captured 3 wickets in the first innings and scored 72. One of the truly great all-round performances ever.

Finally the matches in which the teams had a huge lead in the first innings.

MtId: 1423 Year: 1998 Test# 1 of 1 (0-0) Sri Lanka won by 10 wickets

Eng 445 all out.
Slk 591 all out.
Eng 181 all out     (Muralitharan M: 54.2-27-65-9).
Slk  37 for 0 wkts.

This time Sri Lanka had a huge first innings lead over England of 146 runs. Then Muralitharan wove his magic and captured 9 wickets for 65 runs in 54 overs. If Stewart had not been run out, Murali might very well have captured all ten wickets. All this, at Oval, London. And on a pitch on which over 1000 runs had been scored in the first two innings. The fact that Murali conjured 9 wickets on a batsman-friendly pitch has made this performance a stand-out contender.

MtId: 0128 Year: 1912 Test# 8 of 9 (0-0) England won by 10 wickets

Saf  95 all out.
Eng 176 all out.
Saf  93 all out     (Barnes S.F: 16.4-4-29-8).
Eng  14 for 0 wkts.

Two low scores meant England were ahead by 81 runs. Then S.F.Barnes, certainly in the top group of all-time best fast medium bowlers, bowled unchanged. He bowled 16 overs and captured 8 for 29. South Africa were dismissed for 93 and England won in a canter. Barnes had another equally great third innings spell of 9 for 103 a few months later in a match where England had a big lead.

MtId: 0428 Year: 1956 Test# 4 of 5 (1-1) England won by an innings and 170 runs

Eng 459 all out.
Aus  84 all out.
Aus 205 all out     (Laker J.C: 51.2-23-53-10).

And the matches which were drawn or lost.

This is the only performance in this collection by a team leading by million runs in the first innings. There is a reason. The first 10-wicket haul by a bowler. Laker bowled 51 overs and captured 10 wickets for 53. What is amazing is that the other bowlers, Statham and Lock, no mean bowlers themselves, bowled 71 overs and did not pick up a wicket. This was the bowling performance of multiple life-times. Not to forget the small matter of 9 wickets in the first innings.

MtId: 0436 Year: 1957 Test# 3 of 5 (0-2) Match drawn

Eng 218 all out.
Saf 283 all out.
Eng 254 all out     (Tayfield H.J: 37.7-14-69-8).
Saf 142 for 6 wkts.

This is the only drawn Test match performance in this collection. South Africa led by 65 runs. Then Tayfield produced an excellent spell of classical off-spin bowling, capturing 8 for 69 runs. South Africa had about 4 hours to score 190 runs and, readers note, England bowled the equivalent of 77 overs. Today, the Dhoni-led India, with two spinners, bowls 26 overs in two hours. South Africa struggled and finished 48 short to earn a good draw. In the course of Tayfield's spell, he bowled 119 consecutive score-less balls, and 19 in the second innings, a world record.

MtId: 0967 Year: 1983 Test# 3 of 6 (0-1) West Indies won by 138 runs

Win 281 all out.
Ind 241 all out.
Win 201 all out     (Kapil Dev N: 30.3-6-83-9).
Ind 103 all out.

The only losing performance in the third innings is featured here. Again it is the turn of Kapil Dev. This was a forgettable match for India, barring a fighting 90 by Gavaskar and this bowling performance by Kapil Dev. India were only 40 runs behind in the first innings. Then Kapil Dev captured 9 for 83 in what is, undoubtedly, the greatest display of pace bowling by an Indian and the most memorable on Indian grounds. Haynes was dismissed by Sandhu. West Indies were dismissed for 241 leaving India to get a reasonable target of 242. However West Indies, still smarting under the World Cup loss a few months before, blasted India out of the park for 103. But Kapil Dev cannot be blamed for the batting failures.

In this selection, bowlers like Panesar, Cairns, Agarkar, Reid, Harmison and Caddick would not rank amongst the great bowlers. The bowlers in this group other than Cairns would not figure amongst the top-10 of their teams. However when there was a need they produced once-in-lifetime performances and find their well-deserved place in this list.

The second XV of third innings bowling performances

MtId Year Bowler          For          Vs   Deficit   Analysis

This is my selection, 75% objective and 25% subjective. Readers will have their own favourite third innings bowling performance and are welcome to send in their comments referring to these innings. The only requirement is that you have to take the trouble of looking up the concerned scorecard and give some details. Rather than posting comments such as "What about Imran's 8-wkt spell", the comments which are likely to get published are the ones where a better insight into the concerned innings are provided. Do not get upset that one specific performance is not in this list or in the nearly-made-it list. Put up your cases in a nice and emphatic manner, without resorting to abusive language, unfortunately used in a few mails recently.

The Readers' selection third innings bowling performances, in order of comment acceptance.

1906 2009 J.E.Taylor  Win (Eng) 9-4-11-5 (Balaraman)
0991 1984 M.D.Marshall  Win (Eng) 26-9-53-7 (Alex)
0765 1975 AME Roberts  Win (Aus) 14-3-54-7 (Alex)
1110 1988 Hughes M.G  Aus (Win) 37-9-87-8 (Shane) loss
1554 2001 Warne S.K.  Aus (Eng) 18-5-33-6 (Karthik)
0032 1889 Briggs J  Eng (Saf) 14-5-11-8 (Alex)
1444 1999 Srinath J  Ind (Pak) 27-6-86-8 (Adil) loss
1377 1997 Kasprowicz  Aus (Eng) 16-5-36-7 (Karthik) loss
1654 2003 Pollock S  Saf (Eng) 18-4-39-6 (Venkat) loss
1615 2002 Shoaib Akhtar  Pak (Aus) 8-2-21-5 (Yasir) loss
0172 1928 G.F.Bisset  Saf (Eng) 19-5-29-7 (Ad)
0677 1969 Bedi 23-11-37-5 & Prasanna 24-10-42-5 Ind (Aus) (Arjun)
1719 2004 Kaneria  Pak (Slk) 60-20-118-7 (Arjun)
1356 1997 Warne Aus (Saf) 28-15-43-4
0443 1957 Lock GAR Eng (Win) 16-7-20-6 (Abhishek)
1630 2002 Lawson JJC WIn (Bng) 6.5-4-3-6 (Arjun)
0876 1980 Iqbal Qasim Pak (Aus) 42-22-49-7 (Ad)
0427 1957 DS Atkinson 40-21-53-7 WI (Nzl) (Ad)
1747 2005 M Ntini  Saf (Win) 20-7-37-7 (Mayank)

I have really enjoyed doing these series of four anecdotal articles. My editor probably more. As many a reader has pointed out this has opened the door on wonderful efforts of the past. I myself have learnt a lot. The readers' response has been terrific.

The idea for this article came when I was discussing Lara's 400 with a friend, arguing that that was not even his fifth best innings. He countered by saying that since this was the maximum runs scored in a test it should be considered great. I had to correct him saying that there were two other batsmen who have scored more runs than Lara in a test. He was quite surprised since he could not think of someone scoring more than 400. He was only thinking of one innings. Then I explained to him about Gooch and Mark Taylor.

A very fascinating set of questions. I decided I would do a complete article on this. I am glad that I did it since it has thrown up quite a few great insights into Test batting. So much so I would do a similar article on Test bowling also.

Indeed this turned out to be a tough task since I also wanted to utilize this opportunity to build a Player-performance Database. This is essential since I needed to get the best 1-10 test performances for each player and then get the all-time best performances. I also wanted to provide the information on the top players' 1-10 tests best performances so that the readers could do their own comparisons. And I was sure that there would be queries on the best performances by specific players after the article was published. I wanted to be able to provide the information quickly. In fact I have also provided the huge table of all players' for downloading.

First let me emphasize that this is only a run aggregate. I myself will clarify that this aggregating of runs in specific sequences of 1-10 tests is irrespective of opposing team, home or away, match conditions, period lapsed between matches, not outs et al. That is not the purpose of this article. Readers should appreciate this and not come in with a comment such as "opposition bowling quality is not considered". But that is wishful thinking !!! Also readers who worry bout batting average should understand that when someone scores over 1500 runs in 10 tests, it does not matter about averages. It is going to be quite high.

Let us now look at the tables.

Maximum runs scored in a single test

Batsman         Runs  StartTest

Little would Abhishek have realized what he unleashed when he made the comment on Jayasuriya's 189 being more than 50% of the combined team scores. A simple statement. However it opened up a chain reaction of multiple analysis of outlying performances.

I decided to first do the work related to what Abhishek suggested. This is really a player's performance compared with the other 21 players. Then I did some analysis of the player against the 11 players of the other team. Finally there is one analysis of the player compared to the other 10 members of his team. A few very interesting facts have come to light.

So much so, there is only one case of a player scoring over 50% of the combined match aggregate. That is Jayasuriya, whose 189 formed 53.5% of the total of 353 (299 + 54). Jayasuriya scored 63.2% of his team score and the very low score of India made sure the overall figure remained above 50%. As I have already said, there is necessity for a specific pattern of scores in this analysis. A batsman dominating his own team's innings AND a very low score by the opposing team.

I kept the cut-off at 33.33% and created the table. The table is given below.

SNo MtId Year For Batsman           Runs   Vs  Total  Score   Share%

Next to Jayasuriya's 189 is the other 189, almost certainly the greatest ODI innings played. By Viv Richards, whose 189 (out of 272+168) formed 42.95% of the aggregate. However the most breath-taking of the chasing innings of all, McCullum's 80 (out of 93+95) formed 42.55% of the aggregate. This bizarre (real-life) scoreline is almost close to the outrageous scoreline I had earlier talked about. For a near-100 target to be chased by a team and one batsman scoring over 84% is unreal. No other innings exceeds 40% of the aggregate. The next chasing innings is Haynes's 85 (out of 116+117) working out to 36.5%.

It is not surprising that 9 West Indian batsmen figure in this list. Gayle is the leading batsman with 3 such dominating performances. In general the top three batsmen dominate the table. It is surprising that there is a single Australian entry (that too from Geoff Marsh) and nothing from England.

The next set of outlier innings are the ones where a batsman has outscored the other team by a wide margin. Note the clear distinction. In the first one we looked at the share of the batsman out of the aggregate. Here we look at the factor by which he outscored the opponents. Let us look at the table.

SNo MtId Year For Batsman            Runs  Vs    Score Ratio