I collected Test Match cricket data from cricinfo. Here I'm looking at three plots of first innings runs totals. These data are from 1877 - 2015 and include every game played everywhere in the world. First, we can see that scoring a lot of runs in the first innings doesn't necessarily distinguish a team's likelihood of winning from drawing (at least historically averaged over all teams and locations), but not scoring many runs does increase a team's likelihood of losing.
Next, two plots that show that whilst there is not trend for first innings run totals to be increasing over time, there is a trend very recently for run rates to increase.
More visualizations are available on my GitHub page. For more information please contact me via twitter.