What's a reasonable winning score in ODIs?

I did an analysis on a winning target score in T20s and many subsequent matches showed how close the results of my analysis were

Anantha Narayanan

25-Feb-2013

Sachin Tendulkar jumps to cut, India v Australia, 5th ODI, Hyderabad, November 5, 2009

I did an analysis on a winning target score in T20s and many subsequent matches showed how close the results of my analysis were. So I have embarked on doing a similar analysis for ODI matches. For ODIs there are a lot more matches available for analysis.

First some exclusions. For obvious reasons, I am going to exclude "Abandoned" matches, "No-result" matches (100 in all), matches which were decided on previous "revised score" rules (56 matches ), the more recent "Duckworth-Lewis" rules (101 matches) and a few incomplete innings. The reason is that the D/L and similar situations distort the scores quite a bit. If a team scores 300 and loses to another team which scores 150 in 20 overs, nothing can be inferred from the match. That leaves us 2659 matches for analysis.

I have taken the first innings scores, grouped these into run ranges and tabulated the results. Then I have derived some conclusions on winning target scores by inspecting and interpreting the results.

Let me say that this is a macro analysis. I would appreciate readers understanding this and avoid making comments such as target winning score depending on bowler quality, toss, day-night, team strength et al. All these have been considered in the past and will be considered in future. Let us give a break to these in this article.

The analysis has been done for the following sets of matches.

1. All matches.
2. Starting period matches.
3. Middle period matches.
4. Modern period matches.
5. Matches in Asian sub-continent.
6. Matches outside Asian sub-continent.

I tried analysing this for the countries, but did not get far since the number of matches played comes down and the number of matches in each run group becomes so small that it is impossible to derive any conclusions. In fact for a country such as New Zealand the % of wins for 240-249 is 81.2% and for 250-259 is 60.0%. Such inconsistencies make a country-level analysis a non-starter. Only for Australia, with 472 matches, could this be done with some level of confidence.

How does one define what is a winning score? I have worked on the basis that a score which gives the team a winning possibility of around 60% can be considered a winning target score. Anything lower will not give the team any edge in the long run and aiming for much higher than 60% might backfire on the team in that they might aim for 300 and end up with 220.

1. All matches

FBatScore  Matches   Wins  % wins AvgeWinMargin

Below 125 108 4 3.7 12.8 125 - 149 140 13 9.3 25.9 150 - 174 221 36 16.3 29.6 175 - 199 334 82 24.6 34.0 200 - 219 339 134 39.5 46.0 220 - 229 198 94 47.5 42.4 230 - 239 196 104 53.1 45.8 240 - 249 191 110 57.6 55.4 250 - 259 166 100 60.2 59.3 260 - 279 294 217 73.8 62.3 280 - 299 204 157 77.0 80.2 Above 300 268 243 90.7 101.5

Total 2659 1294 48.7 63.3 From a perusal of the above, it is a reasonable conclusion that a winning target score, based on the criteria already set, is around 250.

2. First period matches (1971-1989)

FBatScore  Matches   Wins  % wins AvgeWinMargin

Below 125 26 2 7.7 7.0 125 - 149 32 4 12.5 15.5 150 - 174 65 11 16.9 25.1 175 - 199 98 29 29.6 36.2 200 - 219 91 39 42.9 45.7 220 - 229 42 23 54.8 30.9 230 - 239 56 35 62.5 48.5 240 - 249 41 25 61.0 60.4 250 - 259 23 16 69.6 57.6 260 - 279 53 40 75.5 60.1 280 - 299 21 19 90.5 82.1 Above 300 16 16 100.0 122.7

Total 564 259 45.9 53.9 Things were tough for the batsmen during these early bowler-friendly times. Lower totals were defended more often than not. Hence the winning target score for this period was 235. Even this has been reached with the higher scores during late 1980s.

No team which scored 300+ runs finished on the losing side. The highest score successfully chased during this period was by New Zealand who overhauled England's score of 296 during 1983. India defended a total of 125 against Pakistan quite comfortably while Pakistan defended a total of 87 in 16 overs against India.

3. Middle period matches (1990-1999)

FBatScore  Matches   Wins  % wins AvgeWinMargin

Below 125 21 1 4.8 14.0 125 - 149 42 5 11.9 18.4 150 - 174 73 15 20.5 35.8 175 - 199 115 30 26.1 32.1 200 - 219 131 56 42.7 38.5 220 - 229 77 42 54.5 44.4 230 - 239 66 36 54.5 40.2 240 - 249 66 43 65.2 45.8 250 - 259 59 34 57.6 44.6 260 - 279 91 70 76.9 67.4 280 - 299 54 41 75.9 73.6 Above 300 61 57 93.4 91.6

Total 856 430 50.2 54.7 Things improved for batsmen during this period. Consequently the winning target score increased to around 240.

4 300+ totals were chased successfully. Australia defended a total of 101 in 30 overs against West Indies.

4. Modern period matches (2000-2009)

FBatScore  Matches   Wins  % wins AvgeWinMargin

Below 125 61 1 1.6 23.0 125 - 149 66 4 6.1 45.8 150 - 174 83 10 12.0 25.2 175 - 199 121 23 19.0 33.8 200 - 219 117 39 33.3 57.1 220 - 229 79 29 36.7 48.6 230 - 239 74 33 44.6 49.1 240 - 249 84 42 50.0 62.1 250 - 259 84 50 59.5 69.9 260 - 279 150 107 71.3 59.8 280 - 299 129 97 75.2 82.6 Above 300 191 170 89.0 102.8

Total 1239 605 48.8 73.5 In the modern times, many more high totals were chased successfully. This effect percolated down and the winning target score could be pegged at 260.

300+ chases were commonplace with South Africa's overtaking Australian score of 434 being the highlight. West Indies defended a total of 124 in 30 overs against Bangladesh.

5. Asian sub-continent matches

FBatScore  Matches   Wins  % wins AvgeWinMargin

Below 125 31 2 6.5 7.0 125 - 149 52 1 1.9 38.0 150 - 174 81 17 21.0 26.9 175 - 199 121 31 25.6 41.5 200 - 219 123 54 43.9 41.8 220 - 229 74 31 41.9 47.7 230 - 239 80 43 53.8 48.3 240 - 249 72 38 52.8 52.8 250 - 259 58 39 67.2 39.5 260 - 279 118 90 76.3 61.5 280 - 299 91 67 73.6 80.6 Above 300 104 95 91.3 94.1

Total 1005 508 50.5 61.1 The winning target score for the Asian sub-continent is around 255. It is not easy to defend low totals on these batting-friendly pitches.

6. Outside Asian sub-continent matches

FBatScore  Matches   Wins  % wins AvgeWinMargin

Below 125 77 2 2.6 18.5 125 - 149 88 12 13.6 24.9 150 - 174 140 19 13.6 31.9 175 - 199 213 51 23.9 29.5 200 - 219 216 80 37.0 48.9 220 - 229 124 63 50.8 39.8 230 - 239 116 61 52.6 44.1 240 - 249 119 72 60.5 56.7 250 - 259 108 61 56.5 72.0 260 - 279 176 127 72.2 62.9 280 - 299 113 90 79.6 79.9 Above 300 164 148 90.2 106.2

Total 1654 786 47.5 64.8 Surprisingly the winning target score is the same as for Asian sub-continent. This has been caused by the way the New Zealand and English pitches have eased in recent times. The winning target score is around 250. Quite a few sub-150 totals have been defended.

Finally it can be seen that, barring the first period, the winning target score is either side of 250.

I started this article before the Hyderabad ODI between India and Australia, and fibnished it after the match. One more 300+ total (oh! a 350+ total) almost bit the dust. No score is safe, it looks like. However this match does not change this article a bit.

As requested by Khalil, I have done an analysis of the period 2005-09 and presenbted the table here.

7. Recent matches (2005-2009)

FBatScore  Matches   Wins  % wins AvgeWinMargin

Below 125 27 0 0.0 0.0 125 - 149 33 2 6.1 49.0 150 - 174 39 6 15.4 27.3 175 - 199 52 10 19.2 33.8 200 - 219 62 20 32.3 62.3 220 - 229 34 9 26.5 49.9 230 - 239 48 22 45.8 47.9 240 - 249 40 17 42.5 69.1 250 - 259 42 24 57.1 67.0 260 - 279 67 45 67.2 54.2 280 - 299 62 46 74.2 84.2 ABove 300 125 111 88.8 103.2

Total 631 312 49.4 76.6 The winning par score could be pegged at 265, 5 runs above the 2000s value. Otherwise the numbers have stayed similar to the 2000s values.

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