October 1, 2011

# Test teams' stay at the top: a complete re-look

A graphical analysis to measure the series records of top Test teams over the years
132

A great fall-out of my Test Series analysis has been that it has provided me an alternate and very effective way of looking at the various teams' stay at the top. This has been triggered by a suggestion provided by Raghav Bihani.

I have approached this analysis with the following points in mind.

1. 4-0 wins should carry more weight than 2-1 wins.
2. Big wins (Inns/10-wkts etc) should carry more weight than narrow wins (1-wkt/20 runs etc).
3. Away results should carry more weight than home results.
4. Deciding Tests should carry more weight.
5. If a 4-Test series is pegged at 1.00, 3 & 2 Test series should carry lower weight than this and 5 & 6 Test series should carry more weight.
6. 1-Test contests are not series and have been ignored in this analysis as also the three Triangular tournaments. The reason for not including 1-Test contests is because inclusion would have a significant adverse impact on the calculations. As can be seen later the averaging across multiple series pre-supposes the need to have series performance as the base. Taking single Test performances as series performances, especially as the strength differentials are quite substantial, distorts the numbers. Anyhow there have been about 100 1-Test series and most of these involve teams in their early stages.
7. Win indices should be adjusted by relative team strengths. Stronger teams should get lower weight and weaker teams should get higher weight.

The first four of these were built into the Team analysis for Series and the last two have been rationalized with multiplying factors, suitably limited. Just to recap the series team analysis, the winning of a match gets a SIN (Series Index) value of just above 60 (for a 1-run win), upto a maximum of around 97 (for the innings and 579 run win). The losing team gets the balance, out of 100. The draws get either side of 50, depending on the nature of draw. Assigning 60+ for a win, as against, say, 55+ is to recognize Test wins in a sharp and definable manner. At the same time the team which draws the match but has been in command throughout, will get nearly 60.

In order to evaluate the results of the teams, I also have considered 10 consecutive Test Series, including the series being considered and averaged the SIN values to work out a TSIN (Ten Series Index) value. This means that for any evaluation a minimum of 10 Test series (easily 3 years) is considered. This value is determined for each series for each country and rolling values arrived at. These TSIN figures are then plotted on a graph similar to the one I had done couple of years back on batting and bowling streaks. Some of these points may not be clear now but will get clarified as we move on to the graphs.

Readers should understand that it is quite tough to get a TSIN value of 60.0 for the next 10 series for a team. 60 represents a reasonably comfortable series wins and every loss/draw/narrow-win has to be compensated within the 10 series period. Also the stronger teams are already pegged back because they are stronger and expected to do well. All this means that only four teams, viz., Australia, England, West Indies and South Africa have ever crossed 60.0 as a rolling average. The other 6 teams have never crossed 60.0 once in their history. That should put these values in perspective.

First a summary table of Series information by country.

```Team        # of Series   SIN >70   TSIN>60    Mean SIN   High TSIN

Australia        178         22        41        55.74       66.52
England          221         21        12        52.25       62.33
West Indies      121          8        10        50.50       66.06
South Africa      98          6         4        51.32       60.67
Pakistan         116          4         0        48.31       54.56
India            126          6         0        47.88       55.67
Sri Lanka         73          5         0        46.83       53.92
New Zealand      126          4         0        43.03       52.94
```

First let us look at the graph for the Australian team. Let me repeat that these are not series performances but plotted using the TSIN values. As such the stay at the top or bottom would be clearly visible. There would not be abrupt moves up and down and the trends would be obvious.

Australia's Test-series record over the years

What does one say. If you forget the initial few series, Australia have had only one really bad period, between 1982 and 1986. Hughes took over after the Packer era and did not move the world. Greg Chappell could not do much and Border took over a weakened side. The wholly unexpected World Cup 1987 triumph changed everything. Otherwise their TSIN values have almost always been above 50. But the real strength of Australians over the years has been the fact that out of the 178 series being considered in which they have TSIN values of 60 or above in 41 of the series.. They have had two real peaks, one between 1930 and 1951 and the other mind-blowing one between 1998 and 2007. Both these are expanded separately later.

West Indies' Test-series record over the years © Anantha Narayanan

West Indies have had a spectacular Manhattan structure until 2000 and then the poorer shanty towns take over. During the past 10 years, they have barely crossed 40. However their heyday was during the 1980s-90s when they had a run of 27 consecutive unbeaten series. Many teams went into Test series against West Indies during these years, considering a series draw as success. Wins were almost out of question. Maybe this defensive attitude also meant the fair number of draws. The later 25 series of this almost unparalleled period of domination is covered separately.

England's Test-series record over the years © Anantha Narayanan

England has had a fairly steady performance graph. They peaked for a spell of 12 series during 1950s and this has been covered separately. Hutton, May, Cowdrey, Compton formed an immense batting lineup. Tyson, Statham, Laker, Appleyard and Lock were formidable on any surface. Other than this they had a brief spell of 60+ TSIN values during 2002-03, with the series ending around 2005. The 1980s were the lowest point for them. Note the spike in the last series. This has been caused by their 4-0 whitewash of the Indians, which fetched them a SIN value of 79. This about 25 above their average and has given a lift-up of 2.5 or so in the TSIN value.

South Africa's Test-series record over the years © Anantha Narayanan

South Africa has had quite a few peaks and near-peaks. Look at the period just after 1962. And as soon as they returned to international cricket during 1998 they had a peak of 10 Tests during which they averaged just above 60. Then they dropped off getting to a fairly low period around 2003, probably prompted by the World Cup debacle. They have since then picked up.

India's Test-series record over the years © Anantha Narayanan

India has been just around average for over 70 years until around the turn of the century. Even then they have been averaging only around the 50-55 mark, never once putting in a sequence of 10 good series level performances. Not once have they reached a TSIN value of 60. Note the fall in the last series. This has been caused by their 4-0 loss to the Englishman, which fetched them a SIN value of only 21. This about 30 below their average and has dropped the TSIN value by around 3.0.

Pakistan's Test-series record over the years © Anantha Narayanan

Pakistan has had a similar graph to India. They had reasonably good periods between 1975 and 1995, the Imran Khan years. They were pretty badly off around 1998, then picked up but have fallen off recently. Again no steady streak. No single TSIN figure exceeding 60.0.

New Zealand's Test-series record over the years © Anantha Narayanan

New Zealand have had alternating good and bad periods. Other than for a short while during early-1980s, their best period has been either side of 1990. This period was orchestrated by Hadlee and Martin Crowe. They are badly dropping off recently.

Sri Lanka's Test-series record over the years © Anantha Narayanan

Barring the first 15 years, Sri Lanka have been fairly steady around the 50+ mark. For a fairly young team, this has been a very good level of consistency.

I have given below the four truly outstanding streaks at top of teams. The criteria is that the concerned team should have secured an average TSIN value of over 60.0 in a minimum of 10 consecutive series. I have taken trouble to find as long a streak as possible. I have also not included the 10-series streaks which have only around 60% value. The bar is higher for these minimal streaks. Looks easy and simple to get in. Let me assure you that it is a very tough criteria and only four streaks have qualified. Australia have two such streaks, West Indies has one and England had a wonderful streak during the 1950s. South Africa had 4 series with TSIN over 60, that is all. The other four teams never even had a single TSIN value of 60.0.

These four streaks have been represented in the following graph. This time the graph has been posted on the actual SIN value since we need to look the details of these series streaks.

Test-series record of top four teams © Anantha Narayanan

Australia played 28 series during 1999-2007. They won 24 of these, often by comfortable margins. The four series outside these successful ones are the 2-1 loss to India during 2001, Ashes loss by 2-1 to England during 2005, the 0-0 draw with New Zealand during 2001 and 1-1 home draw with India during 2003. It can be seen that in these four series Australia have ended with value below 50, but above 40. Australia's average SIN value during these 28 series was 64.57, an achievement which can only be understood after understanding the nuances of numbers used in this article.

West Indies' streak is the only unbeaten one in this elite group. However their 25 series average is not very high since they drew 8 of the 25 series. They also had the two white-washes against England during this streak. This 10-0 record also indicates that their other wins have been closer.

The Bradman-led Australian teams between 1930 and 1951, had a streak of 14 series during which they had an amazing average of 64.82. The only loss was the bodyline series to England and then the 1938 draw.

England had a nice 12-series streak during 1950s when they did very well. A single exception being the Ashes series of 1958-59 when they did very poorly.

Now for a numerical summary of these four streaks.

```Team        Streak Period Series Won Drawn Lost SIN avge Tests Won Drawn Lost

Australia    1999-2007      28    24   2     2   64.57     90   69   11   10
West indies  1981-1995      25    17   8     -   60.13    104   56   33   15
Australia    1930-1951      14    12   1     1   64.82     69   45   12   12
England      1954-1960      12    10   1     1   62.26     54   33   15    6
```
Which team's streak was the greatest. We can comfortably leave out the last two ones. There are not enough series and the results are not that great, although the Bradman-led streak is quite impressive. However 28 series is double the number 14 and means a lot more. Let us take the first two streaks. Australia, ever willing to take chances, playing for a win almost always, had an amazingly high winning record in these 28 series, viz., 85.7%. West Indies, had an unbeaten sequence of 27 series, the earlier two not included here since that would have got the SIN avge below 60. However many a drawn series cropped up. A not so great winning record of 68%. You can take your pick. Both teams have their many pluses and a few minuses.

My personal vote is is for the Australian streak of 28 Test series. Primarily because of the way they changed the approach to Test cricket, their conistent scoring rates well in excess of 3.50, their willingness to lose the odd series/test in their quest for a win, their more balanced bowling attack and Gilchrist. Again, let me emphasize, this is my personal preference. You need not agree, that is your prerogative. But do not criticise my selection in a negative manner. And, if these two teams face off in a 5-Test series, I will get this simulation going within the next 6 months, the result will be 3-2 for Australia on odd days and 3-2 for West Indies on even days !!!.

Mansur Ali Khan Pataudi. RIP. A truly great cricketer and human being, fearless person, attacking captain, secular to the core, fielder extraordinaire, mover of Indian cricket forward in a manner no one has ever done and would have been one of the greatest Indian batsmen ever if he could have seen one ball with two eyes instead of seeing two balls with one eye. All these with no helmets, no chest pads, no arm-guards, no thigh pads and hopefully the box, if the Indian Board could have afforded one. Who can ever forget his 148 at Headingley, one of the bravest back-to-the-wall knocks ever. The images of Pataudi brought to memory the black-and-white era, the period of Guru Dutt, Richie Benaud, Waheeda Rehman, Rod Laver, Dev Anand, Pele, John Wayne, Ramanathan Krishnan, Nutan, Salim Durrani, Alfred Hitchcock, Mohd Rafi & Lata singing, Milkha Singh, Raj Kapoor, Sivaji Ganesan, James Stewart, Paul/John/George/Ringo, Prasanna/Bedi/Chandra bowling together, a collection of magical Singhs with hockey sticks et al. Everything was done for the love of doing it, for a few bags of peanuts.

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

• testli5504537 on October 31, 2011, 6:35 GMT

Wonderful analysis and such terrific debates on the Bradman,O'Reilly era.Psychologically Bradman was a bigger factor than any two other players.He could not have been boring getting runs at the rate he did. I have read Cardus,Arlott and others talk about his footwork and placement and precision-someone who is ruthlessly efficient need not be boring at all!! When Hutton was asked to name his best England team to play Australia's best,he picked Leyland surprisingly-and explained the choice by saying he played O'Reilly best

• testli5504537 on October 21, 2011, 22:35 GMT

I know you said you'd do a match up between the two teams, but without going into the analysis (yet), who would you select for both sides? I was driving to work yesterday & initially thought the WIndies would be hardest to select. Certainly there would be a great pacer missing out. I found the Ozzys were the hardest to select. IMO - WIndies; 1. Grenidge, 2. Haynes, 3. Lara, 4. Richards, 5. Richardson, 6. Lloyd (c), 7. Dujon, (top 7 was the easy bit - Logie, Gomes & Adams missing out. 8. Marshall, 9. Holding, 10. Ambrose, 11. Garner. 12th man: Harper/Logie. The Baggy Greens 1. Taylor (c), 2. Hayden, 3. Ponting, 4. M Waugh, 5. S Waugh, 6. Symonds, 7. Gilchrest, 8. Warne, 9. Lee, 10. Gillespie, 11. McGrath 12th: Bichel as he was just about the greatest 12th man ever! Oz is difficult as M Hussey would of only just came on the scene, I went for Symonds for the allround value. I think Oz would select different sides for different pitches, whereas the WIndies wouldn't. Tubby qualifies??

• testli5504537 on October 19, 2011, 14:47 GMT

Don't know if the idea of Bradman averaging 110 against "weak" test bowling long term is unreasonable.

If Bradman had never played, wouldn't we all say that a batsman averaging 100 over 52 tests was unreasonable? - and wouldn't we all have been wrong?

Looking at Don's record in Aus 1st class cricket, there are only two teams he played against throughout his career (having played for NSW early on, and then moved to SA, he didn't play as much against either of those teams): Victoria and Queensland.

Presumably, Victoria had a better bowling attack than Queensland. Don averages 107 against Victoria and 141 against Queensland. This is from 1927-1948.

Would probably exclude the West Indies from group of "weak" attacks. Francis, Martindale and Constantine have decent records, and both Bradman and Hammond underperformed against them. Weak batting probably was their Achilles heel

• testli5504537 on October 18, 2011, 4:44 GMT

Ananth: Top article again. I think the Oz v WIndies debate is impossible to prove. My thoughts are that if a showdown took place between the two great sides it wouldn't so much be about the rules (although no helmets would definately swing the WIndies way), but WHERE it was played, more importantly what type of PITCH. My bias is towards Oz being the better but the thought of the WIndies on an English greentop, would scare the preverbial out of me & that's despite being on the other side of the world typing on a computer. Forget Nightmare on Elm Street, you could do a horror movie titled Green Top! It would be carnage, that being said, Warney could turn it on a green top & McGrath would take wickets. I think the match up would have to take place at the Gabba (yes Ozzy territory), but I think the pace, bounce and lateral movement would be as favourable to the WIndies as anywhere else in the world, & IF, the Ozzys survived the first couple of days, Warney could turn the match late??? [[ Let us also have metches at SCG and Eden Gardens. Then again the first two days would matter but would be a more even contest. My reading is 3-2, that is all. I agree some batsmen would rather face Freddie than Malcolm/Curtley. Ananth: ]]

• testli5504537 on October 17, 2011, 23:32 GMT

Thanks Ananth for the responses. Also, what i mean is that Australia lost 2-1 to Indian in 1998 and 2001 but in the former, they won a dead rubber and in the latter, a live test. So, irrespective of the victory margins in individual tests, Australia should get more points for the 2001 series loss than the 1998 one. Same for India vs Aus in 2004 and India vs Aus in 2007-08.

Also when one purely considers series performances, a 3 test series with a narrow win sandwiching 2 bad losses should get greater weight than 2 decent losses without a victory. I wonder if it is possible to assign points in a manner that will favour a live test victory in a series more than decent losses. [[ Yogesh, I feel over-complicating this will answer a few queries but the effort is not worth it. As far as I am concerned there have been no "Dead rubbers" over the past 10-12 years. Was the last Test in England a Dead rubber. It meant so much to both teams. With the Test rankings at stake, no Test is dead. The fact that Australia took off their foot on the pedal in some cases should not deter from this fact. Ananth: ]]

• testli5504537 on October 16, 2011, 22:18 GMT

One of the better comments section in Cricinfo and elsewhere. The nature of Ananth's analysis means that fewer people can comment as it requires serious reading but compensated by very well thought out comments.

A small clarification : By deciding tests you mean live tests and not merely deciders. [[ As of now I have taken only the true deciders since I wanted the Test to mean equally for both teams. I understand what you mean. However a team leading 2-0 at the end of the third Test of a 5-0 series has a different motivation for the fourth Test as compared to the other team. Ananth: ]]

I was wondering if there is a way to measure series competitiveness of teams. Someone pointed out that Ind, Eng and SA had similar W/L records in the last 10 years. But i think that if one compares in how many series of the respective teams, the final test counted for something (towards winning or drawing the series), India would have a higher percentage than Eng or SA. Eng & SA won a great many series comfortably but also lost quite a few easily. India neither won nor lost a series easily. [[ The Series Index is a recent introduction and opens the way for many a critical analysis of Team performances. A single pair of numbers, allotted out of 100, is a true way of measuring the way the series went. I will look at what you have suggested. Ananth: ]]

• testli5504537 on October 16, 2011, 16:29 GMT

Would Bradman have averaged 110 against weaklings? I dont think so

Nor do I. I was just trying to put Bradman's feats in the context of 30s cricket. The fact that he averaged 99 though his "great" contemporaries averaged 40 runs less despite facing much, much weaker attacks!

• testli5504537 on October 16, 2011, 8:00 GMT

Would Bradman have averaged 110 against weaklings? I dont think so. He did flog India,, but i would put that down to the rarity of facing weaklings. Had the weaklings been numerous, he would have lost interest after some time (of course, on an innings wise basis, before losing interest, he may have scored 100 each time). So perhaps average 100, yes, but significantly more than that, unlikely. [[ Gerry, yes above 110 was probably out of question, just as Lohmann taking a wicket at sub-10 average. Boll, my commiserations. Wales were a single kick away from winning the match. The Wallabies were outplayed. Not a great consolation, though. Ananth: ]]

• testli5504537 on October 16, 2011, 3:57 GMT

Hammond was the opposite. He'd butcher the weaklings (not that he fared badly against the good teams), he seemed to make a point of doing so

Yeah. I think Ananth's research some months ago seemed to suggest that the BQI of attacks faced by Hammond and Hutton is far weaker than the BQI of attacks faced by Bradman!

If Bradman had played for England, he would've had many more opportunities to play more tests against Windies, India, SA and NZ! I wonder whether he'd have averaged 110!

• testli5504537 on October 16, 2011, 2:31 GMT

He said as terrible as bodyline was, one benefit of it was that it cured Bradman of his "chanceless" style of batting that was bad for the game. Absolute DRIVEL

Yep. Drivel it is. Especially when you consider that Bradman was by far the quickest run scorer in 30s cricket. The guy scored at a Strike rate in excess of 60 most of the time, in sharp contrast to "flamboyant strokemakers" like Hammond whose SRs seldom crossed 40 in Ashes test matches.

I'm amazed at the number of Englishmen who regard Bradman as "boring". How can a guy who scores 300 runs a day be boring? Batsmanship was quite defensive in the 20s/30s. Nearly as defensive as it was in the 50s/60s. Bradman was an exception and not the rule. One has to go back to the days of Trumper to find someone who scored at a comparable rate.

• No featured comments at the moment.