This is the story of tackles won (as ratio, so not game number sensitive) and average goals conceded per game. Size of the points given by the average number of fouls committed. Added two normal regression lines (on without the outlier, North Korea) one for the whole set of teams. Plus smoothed regression lines for those who went beyond the group stage or did not - to see whether that would be a possible explanation for this weird relationship. It does not. Looking at the data, WC 10 was an event where the more tackles you’ve won the more goals you’ve conceded. This is given by the fact that high numbers of tackles could indicate the dominance of the opposing team. Just guessing though! (data via R-Bloggers)
