[Imtiaz]

The analysis is excellent. It certainly looks good and matches our gut instincts.

However, consider this scenario...

Bangladesh scores 177 in 50 overs and India goes into bat. As the Required Run Rate is not that high, India plays more cautiously than it normally would if they were chasing 257, say. The result, their [our opponent's] runs/over is lower than it could be otherwise and their runs/wicket is higher than might be the case normally.

You may have seen results like this:

Team "A" 202 in 50 overs

Team "B" 203/4 in 49.1 overs.

I have heard some people in these scenarios remark on how "close" the match was. Only, 5 balls !, they exclaim. Anyone, with some knowledge of the game knows it was not close at all.

It could be argued that this happens with all countries. True. But since Bangladesh has been on the losing end of a game 85 out of 92 times, a statistical bias exists. This bias , theoretically can be reduced by some adjustment using ICC rankings of the day, for example. But, for simplicity, a certain k factor can be introduced to take care of it.

Am I just being difficult ? Or, do I want to see more pretty pictures ?

Note: I can see why you took the sample size to be 21 matches as it corresponds with the Whatmore era. However, it might be better to take the last series [even a one-off match ] against each country for all countries. The only limitation I would place would be to disregard any series/match more than 3 years ago. This way the sample of matches selected would be broadly representative. Of course, if some kind of adjustment could be made regards strength of opposition, then any fixed period would be acceptable as each game would then be "normalized" in Arnab's words.