And just like that we have reached the last of the eight articles that look at the analysis of the pre-tournament World Cup polls that were carried out a fortnight or so ago. If you have reached this article with reading any of the others I would recommend going back to the start and working your way through each of the previous seven. If you have been reading them all, then you will know that we finished up with Group G last time around and now we look at Group H.

First up let’s take a look at how the poll to see who will finish top of Group H played out.

Colombia may have left it late but they eventually finished top of Group H, after wins over both Poland and Senegal. 103 votes were logged in this particular poll and of those 63% opted for James Rodriguez to lead Colombia to the top.

The Bayern Munich midfielder’s contribution may have been somewhat fleeting a number of injury issues and he now faces a race against time if he is to be fit to face England in the Round of 16. This means that twitter is now 11 for 15 in terms of correct polls. But how will they fare in the 16th and final poll of this stage.

Well there is nothing to celebrate here as of the 98 votes that were logged, 34% incorrectly chose Poland to finish runner up. While surprisingly Japan finished last in this poll with only 17% of the vote, although that said had they had a couple more bookings to their name it would have been Senegal going through.

Still any of Japan and Senegal advancing was still not Poland and that means that Twitter finished up on 11 polls out of 16 correct. Just under three quarters of all the predictive polls were correct before a ball had even been kicked.

Let’s now take a look at the voting breakdown for Group H:

65 votes were correctly logged for Colombia to finish top of Group H
33 votes were incorrectly logged for Poland to finish runner up in Group H.

This means that 32.3% of all the votes in this group were correct, just under one in three.

In terms of the final title votes, there were 1515. A big thanks to everyone who took the time to vote – of course people may well have voted more than once, so they are not necessarily individual but importantly they all count.

This means that of those 1515, 660 votes were correct – this works out at 43.5% of all votes were correct. Considering that there is a 75% chance it could be wrong, the fact that it is nearer an even split is quite impressive, so well done social media.

If we were to assign a value of 4 points to the poll winner and 1 to the team with the lowest votes, Group H would look like this:

Colombia 7
Poland 7
Senegal 4
Japan 2

Here we can see that there was far too much confidence in Poland but not enough in Japan. Those two teams ultimately need to be switched in order.

But what does that do to complete our Round of 16 list

Croatia (voted second place)
Argentina (voted first place)
Switzerland (voted Serbia in second place)
Sweden (voted Germany in first place)
Japan (voted Poland in second place)

A total of 11 polls were correct, a hit rate of 68.75% while 13 qualifiers were correct a hit rate of 81.25%

A pretty decent return, but not one that you would surprise you either. The real test will be when we analyse the knockout stage polls. Can Twitter get things right when it really matters?


I am always available for any football data projects you may have and/or content creation.
If you have any requirements then please feel free to contact me at with the subject line “Football Data”

I am also available via twitter at @dantracey1983 – again feel free to get in touch to discuss any work. Thanks, Dan