THE CORDON HOME

BLOGS ARCHIVES
SELECT BLOG
April 7, 2010

Test cricket

A Test series for the gods - part 1

Anantha Narayanan

An intriguing title to an article radically different from my normal analytical efforts. I can assure the readers that they would not be disappointed.

During early 1990s we had developed a series of complex and unique Test and ODI simulation systems. We had simulated for Sportstar an ODI World Cup. We had also conducted an inter-school tournament between the top schools, letting the children captain various teams. Also we had done some innovative pre-match simulation of the matches during the 1999 World Cup.

During 2002, I undertook a very different and unusual exercise with Times of London, in conjunction with Wisden Online. This was to simulate a series of 5 Tests between an all-time England XI and all-time World XI. For various logical reasons we restricted ourselves to the post-war players. These matches were to be played at Lord's, SCG, Bridgetown, Cape Town and Calcutta. The two teams were selected by Christopher Martin-Jenkins with inputs from us. The actual simulation was done in Bangalore over a few days.

The results were published in London Times, with comments by Steven Lynch, between 26 July 2002 and 3 August 2002. The published scorecards will be scanned and shown in the next article.

Since most readers might not have seen these articles, I felt I ought to do an article on this unique exercise. In the first part I will talk about the simulation methodology and the teams which were selected. In the second part I will cover the actual "Test" match scores and the original match reports as sent by us to London Times. I am sure the readers would find these worthwhile to peruse.

SIMULATION METHODOLOGY:

1. Player Data:

The actual career data of the player concerned is used. The following figures are part of the Player data. Readers should remember that I have lot more data available now than during 2002.

- No of Tests
- Career years span
- Type of batsman (Opening/EMO/LMO/Tail)
- Runs scored
- Batting average
- Highest score
- No of 100s & overall pattern of 100s
- Catches/Stumpings
- Type of bowler (Fast/FM/M/SLA/OB/LBG/LAC)
- No of balls bowled
- Runs conceded
- Wickets taken
- Bowling average
- Bowling RPO
- Bowling strike rate.
Certain other data is derived from the player career figures and perusal of scorecards. The derived data explained below.

- Innings Size index: The ability of the player to play long innings. Bradman and Zaheer (and now Lara/Sehwag) converted most of their 100s into big scores. Not Lamb, Mark Waugh or M.Amarnath.
- Expected balls per innings: The average number of balls expected to be faced by the player. Highest is Bradman with 175 balls.
- Expected runs per ball: Ranges from .75 for Jessop/Gilchrist to .2 for strokeless wonders. Product of above two generally works out to the Career average.
- Strokeplay index: How far can the batsman be moved into attacking situations. Ranges from 6 for Bradman/Jessop to 1 for C'Shekhar/Malcolm.
- Defensive index: How far can the batsman be moved into match saving situations. Ranges from 5 for Gavaskar/Atherton to 1 for C'Shekhar/Malcolm.
- Bad wicket technique: High for Hobbs/Gavaskar and low for Srikkanth/Smith.
- Adjustment factor: Provide for Trumper et al scoring runs on uncovered wickets.
- Fielding index: Gives an indication of the quality of players' ground fielding. Highest index is 5 for Constantine/Bland/Randall and Rhodes.
- Bowling type: Type of bowler – Attacking, Normal or Defensive.
- Variation index: Ability to vary the deliveries. Hadlee/Grimmett/Warne high.
- Effectivity index: Ability to use atmospheric conditions for Fast/Medium bowlers and ability to flight for spinners. Prasanna/Murali high and Emburey low.
- Fielding: Team's Run-saving based on Fielding Index. Catching ability on average catches per match (Hammond/Solkar/Mark Waugh fairly high).

2. Ground Data :

Around the world, all test playing grounds including Sharjah have been included. Each ground has 6 index values.

- The support to Fast Bowlers (On a scale of 1 to 5)
- The support to Medium Pace Bowlers (On a scale of 1 to 5)
- The support to Spin Bowlers (On a scale of 1 to 5)
- A Run-Index for the ground (On a scale of 5 to 25). These have been built based on an article which appeared in Sportstar during the early 90s and have been updated since.
- Rain Index : Ranges from 5 for Manchester (almost every match will be affected by rain) to Sydney (almost always sunny).
- Close of play Index : Ranges from 4 for UK Grounds (Good chance of 2 Hours after Close of Play) to 1 for Calcutta (virtually no chance of extension of play). This factor has an impact on the number of overs bowled during the day.

3. Current Form :

This is randomised by the Computer for each player. The form index ranges from 4 (In Great form) to 1 (In poor form). Each block of 50 balls safely negotiated by the player will improve his form. The captain has to take into account the form and shepherd his players through poor form phases.

4. Rain

Rain is an integral part of the game, especially if a match is played at Manchester, Galle or Port of Spain. The concept of rain is built into the game depending on the ground rain index. Complex calculations determine the occurrence and duration of rain. It is also possible that matches take place without any occurrence of rain.

5. Simulation :

The match is, in reality, between two external captains who have at their disposal the players as resources. They are responsible for all actions including team selection, batting orders, strategies and fine tuning of plays.

The simulation is a complex process. Each ball is a mini-match and the complete match consists of x such mini-matches. For each ball, a total of no less than 30 randomising decisions are taken to decide on the outcome of ball. Some of the factors depend on the quality of the players involved, some on the match situation and some on the decisions taken by the two non-playing captains. These are briefly described below.

- The characteristics of the batsman who is batting. 13 factors are used.
- His form at the start of innings.
- How long has he been batting. Form improves as he settles down but he will get tired.
- The characteristics of the bowler who bowls the ball. 11 factors are used.
- What is the bowler type (See note below).
- His form at the start of the innings.
- How long is his current spell, breaks taken care of. Fast bowlers lose some effectiveness after 10, Medium after 15 and Spinners after 20 overs. - What is the condition of the ball, how old is the same.
- What is the time of day. Early mornings will favour seamers.
- Ground characteristics, both in terms of support to bowler type and run getting index.
- What is the average of the fielding index of the team. This will have an impact on the runs taken by the batsman.
- What is the average of the catching index of the team. This will have a slight impact on the wicket falling scenario.
- What is the bowling strategy. The fielding captain will be allowed to select one of 8 bowling strategies (ranging from completely wicket-taking (8 fielders near the bat) all-out defence (all at the boundary)) for both the bowlers. He/she has to take into account the match situation, ball situation, bowler bowling the current over and the specific skills of the batsman. The strategies can vary between the two bowlers. These can be changed at any time.
- What is the batting strategy. The batting captain will be allowed to select one of 5 batting strategies (ranging from all-out attack to all-out defence) for both the batsmen. He/she has to take into account the match situation, ball situation, bowler bowling the current over and the specific skills of the batsman. Nothing will be gained by asking Atherton to attack or Jessop to defend. The strategies can vary between the two batsmen. These can be changed at any time.
- What is the innings status. A number of factors are used in this.
- Is there a Run control option in force. This is mainly to let a senior batsman control the strike when batting with a tail-ender. Runs will be declined during early part of the over and odd number of runs will be attempted during later part.

Note: Only Sobers has bowled at the top level with equal effectiveness as a seam bowler and a spinner. So only for Sobers will the captain be asked at the beginning of each over as to what type of bowling he wants Sobers to do.

The Captain, whose role is a combination of on-field captain and off-field coach, has to use his resources very effectively. He should make his bowling changes with care, give his bowlers required rest, plan his strategies sharply, decide on how to optimise the resources at his disposal, especially outstanding resources such as Bradman et al.

6. Teams selected:

The two teams which were selected are given below. Since each team had two outstanding fast bowlers, two great spinners and a top all-rounder, it was decided that the same team would play in all 5 locations.

All-time Post-war England XI

Hutton
Gooch
Dexter
May
Cowdrey
Botham
Knott (wk)
Laker
Underwood
Trueman
Statham
12th man: Randall
All-time Post-war World XI
Gavaskar
Richards B A
Richards I V A
Lara
Tendulkar
Sobers
Gilchrist (wk)
Warne
Marshall
Muralitharan
Lillee
12th man: Rhodes
It is gratifying to see that the World XI is a time-less one in that if I were to select one today, I may not make a single change. Ponting for Barry Richards perhaps, maybe not, since opening is a specialist position. I would dearly love to have Wasim Akram, but at whose expense ???

As far as the England team is concerned, maybe the same holds good. Possibly Pietersen for Cowdrey.

In addition to the five Test series, a one-off "Test" was played between the team selected by a lucky reader (P.J.Mickleburgh) and an eleven selected by Christopher Martin-Jenkins.

The second part will follow within a few days. This will contain all six scorecards and reports.

Finally let me assure the readers that this is not an attempt to plug any of our company products, as insinuated by couple of readers earlier, possibly when their favourite player was positioned below the top. I have nothing to sell, no products, no services, nothing !!!

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

RSS Feeds: Anantha Narayanan

© ESPN Sports Media Ltd.

Posted by Douglas on (February 15, 2011, 16:28 GMT)

Dear Ananth, as so many readers here are interested in the 'Barry Richards Debate', maybe you could, or get one of your co contributors at cricinfo to write an articles on arguments for and against Richards, and maybe even looking at the statistical value of such small data ranges, and even explore the argument as to whethe county cricket can really predict the success of an international cricketer : I remember Duncan Fletcher thinking the contrary I just find these topics interesting. [[ To do a thorough analysis on Barry Richards, we have to do a lot of extrapolating since he played only 4 tests.. Ananth: ]]

Posted by companionship on (June 19, 2010, 2:17 GMT)

Rather cool blog to spend some time on reading it in my opinion. BTW, why don't you send that post to social bookmarks? That might bring pretty big traffic to this blog.

Posted by TequilaGuy on (April 15, 2010, 12:33 GMT)

Hello Ananth, I read your blog regularly and usually end up spending 2-3 days a month understanding and analyzing your posts.(This invariably happens in office)I appreciate the efforts you put in for each blog entry.I understand that this has been thoroughly tested and I am interested in understanding how some of the subjective (or not so objective as OpenSource pointed out)criterion were first arrived at and how they changed based on your testing results.Being a programmer,I have come across similar problems while working on some airline projects.And being a cricket lover, I would love to know Strokeplay index,Defensive index,Bad wicket technique,Fielding index, Variation index,Effectivity index of various players:)Would it be possible for you to publish these for ROW and England?You have mentioned that you will describe the simulation in more detail to a limited audience.On a lighter note,what is the qualification criteria for being part of it?Should i enroll for a PhD at Oxford? ;) [[ TG I will certainly make available a lot of relevant information to interested readers. These are scattered all over my archived data. I have to get these organized. Those who have shown a genuine interest in knowing more will certainly get access. I am off PhDs for the time being. Just be your normal self. No problems. A genuine respect for the work which has been done and a sincere desire to know more are the only qualifications. Ananth: ]]

Posted by Srikanth on (April 13, 2010, 0:51 GMT)

http://www.cricinfo.com/magazine/content/story/455574.html

Alex, nothing against gavaskar...but for real toughness..there was no one to beat Border.

Posted by Youvi on (April 10, 2010, 22:33 GMT)

As I ponder over the discussion on Barry Richards, I am inclined to agree that Barry must be excluded from World XI. Obviously wrt his Test career it is a tragedy that Barry was a victim of circumstances and could play only 4 Tests. Looking at his first class career it is obvious he was outstanding but then aren't we to base our selection upon Test performance over time ? There are several what-if situations - take Gavaskar, for example, 774 runs in first 4 Tests and then took another 7 tests to get 226 runs for a 1000 ! At that point in his career, (from SMG's book Sunny Days)Gavaskar thought he could be on the verge of being dropped ! The rest is history and I don't need to preach to the choir regarding Gavaskar's achievements. Another interesting example is Vinod Kambli who had a precipitous drop in Test performance after his first 7 tests. Last, but not the least, Ananth, this is great stuff. It is fun plus really helps to develop a sense of cricketing history.

Posted by Alex on (April 10, 2010, 12:18 GMT)

Ananth - a philosophical question: what are the rules under which the matches will be played? The rules have changed a lot over the years ... the LBW rule, bouncers/over, front-foot no ball, etc. Hobbs, e.g., it appears, loved to pad away the deliveries for which he would be given out in the modern game. Not to belittle Hobbs - I think a champion in one era will be a champion in any other era. [[ Alex As fas as this simulation is concerned the 1995 rules and strategies in force. Ananth: ]]

While one may lament the SA ban for short changing the Pollock brothers, Proctor, & Barry Richards, what about the inexplicable selection committee policies that killed the careers of great players? Pak has always been notorious on it. Also, the WI on Lara: the travails of the WI brigade at IPL '10 make one wonder just what the selectors were hallucinating about in giving hard time to an incomparable artist who is also one of the top 5 batsmen of all time. For an MI fan like me, watching Bravo & Pollard at crease is so bad that it goes past good and becomes bad again.

Posted by Abhi on (April 10, 2010, 9:34 GMT)

Further (since I am now a bit warmed up)…how come the mentally challenged folks figure that Gavaskar has bribed someone to get into the team when someone like Barry Richards (all of 4 Tests old) stares them in the face!! Talk about rooobish.

Posted by Abhi on (April 10, 2010, 9:29 GMT)

Alex, I remember reading something to that effect too. And you know what?- I think Gavaskar is absolutely right. How on earth can you judge someone on the basis of 4 Tests? This is a completely mockery of the high standards set by the other stalwarts in the team for over a decade in International cricket. And I stand corrected in my remark that you cannot possibly make a change to this World XI. Barry Richards actually must go… He fails the most fundamental test for any sporting standard– The “Time test”…if a record is not “Time tested” it is of little value.

Posted by Abhi on (April 10, 2010, 8:35 GMT)

Underground. Actually , that is a pretty insightful observation. For eg. we have DGPS ( differential GPS) used out at sea (or even ashore) wherein a “fixed” station who’s coordinates are exactly known (Similar to your actual real life match) gives us the “error” between the known fixed coordinates and the GPS readings of the same , so greatly improving accuracy in the area of coverage. Don’t know how far we can use a similar approach for this sort of simulation though.

Posted by Underground on (April 10, 2010, 7:56 GMT)

Have simulated outcomes been verified against actual matches?

Has anybody tried to put in figures to simulate Botham in 1981 and the result of the Ashes? Or, more specifically, has anyone tried to simulate the state of a real match at a particular time, and manipulated parameters to obtain the actual result?

Comments have now been closed for this article

ABOUT THE AUTHOR

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

All articles by this writer