Im bored so I though I would look to standardise bowling Test bowling averages of different players from any time period. This would hopefully make comparisons easier, though never exact.
Im sure it will have been done before but it is an interesting exercise.
Im sure there may be points made about the methodology but it is the fairest way I can see possible.
Break the history of Test cricket into smaller groups and work out the 'global' Test bowling average for that period.
19th C- 22.38 Global bowling average
Pre WW1- 26.15 Global bowling average
Inter War Years- 33.33 Global bowling average
1945-75- 31.80 Global bowling average
75-90- 32.56 Global bowling average
1990s- 31.87 Global bowling average
2000s- 34.38 Global bowling average
It is important to see the differences as each time period will have had different conditions that may have favoured the batsman or the bowler and affected the balance between the two. eg uncovered tracks, bat technology, LBW laws etc
One of the figures needs to be chosen as the base number. I chose the 1990s as any one would do but it is simply the period I watched the most cricket in.
So the 1990s number of 31.87 is the base number of 1. Then this number of 31.87 is divided by all the other 'global averages' to give the weighting for that period of time.
19th C- 1.42 Weighting
Pre WW1- 1.22 Weighting
Inter War Years- 0.96 Weighting
1945-75- 1 Weighting
75-90- 0.98 Weighting
1990s- 1 Weighting
2000s- 0.93 Weighting
Now averages from those periods would be multiplied by the weighting and a new and more standardised average would be the result.
However, to further ensure that the numbers are more comparable and that we are comparing apples with apples, only the wickets taken in games won or lost will be used in the new average.
This is to get a stable base where all games are played in an environment where 20 wickets are capable of being taken and doesnt penalise players who played on dead tracks or favour players that spent most of their times on bowling heavens (eg Lohmann never played in a drawn test match).
So the stats from non-drawn tests are given their weightings and the new standardised averaged are the product.
Many players have played across different periods so the games they played are weighted differently.
Actual Test Av= 20.94
Standardised Test Av= 17.78
41 games in 75-90 that were either won or lost taking 238 wickets @ 17.24
11 games in 1990s that were either won or lost taking 43 wickets @ 21.09
((41/52)*0.98*17.24) + ((11/52)*1*21.09) = 17.78
Actual Test Av= 21.57
Standardised Test Av= 20.26
45 games in 45-75 that were either won or lost taking 218 wickets @ 20.26
20.26*1 = 20.26
Actual Test Av= 16.43
Standardised Test Av= 19.58
In Pre-WWI he took 165 wickets in games either won or lost @ 16.05
16.05*1.22 = 19.58
Actual Test Av= 21.73
Standardised Test Av= 19.79
28 games in 1990s that were either won or lost taking 161 wickets @ 24.08
50 games in 2000s that were either won or lost taking 375 wickets @ 18.7
((28/78)*1*24.08) + ((50/78)*0.93*18.7 = 19.79
So with the few I have done.
Standardised Test Bowling Av Leaderboard
Marshall = 17.78
McGrath = 19.29
Barnes = 19.58
Murali = 19.79
Hadlee = 20.12
Trueman = 20.26
Donald = 21.33
Warne = 22.91
Kumble = 23.97
Harmison = 26.45
Sobers = 30.22
Would be willing to work out some other players if people want to request them