Apr 28, 2017 So here it is as promised, I was doing a photoshoot of my fish so decided to net up my babies and get some good photos for you all:) Sorry for the lowish quality. Had to crop the photos to get them uploaded:( @Polyaddict86 @Hao First up my wild caught female Senegal bichir from Nigeria. World Cup 2018 Poland 1-2 Senegal, RESULT: Relive the action after Robert Lewandowski and Sadio Mane faced one another. A deflected Idrissa Gueye shot put Senegal into the lead in the first-half.
We’ve been gripped by the FIFA World Cup 2018 here at Intelligence Towers. There were high scores and high drama as old favorites crashed out early, leaving underdogs to progress to the latter stages. It was the most uncertain tournament in decades.
Uncertain, perhaps. But apparently not unpredictable. A couple of weeks ago, at the start of the round of 16, we used network theory to make a tongue-in-cheek prediction of the World Cup winner. Mac os software download.
To everyone’s surprise, the predictions turned out to be remarkably accurate. Let’s take a closer look at what happened.
Our prediction was based on this dataset by Paul Campbell. It lists players in the World Cup, their clubs and countries of origin. We had no data about squad or player history, performance, or injuries. But we didn’t need that – we had network science.
Our theory was that teams with the most international connections would perform better than those whose players are mostly signed to domestic clubs:
To find the teams with the most wide-reaching connections, we used a social network analysis measure known as Eigenvector Centrality. Eigenvector Centrality scores each node by how many connections it has, weighting those connections by how many connections the node at the other end of the link has, and so on.
As this calculation iterates through the network, it converges on a score that indicates a node’s relative importance. In this case, the result was a list of the final 16, ranked from most international to least:
Based on our theory, teams with higher centrality scores are more international and therefore should progress further in the competition. How did we do?
Looking back, we can compare each team’s centrality score with how far they progressed. The result was a surprise to us all:
On the left-hand side of the chart, we see the teams that were knocked out at the group stage. On the right-hand side are the teams that progressed further in the competition. Pleasingly, adding a trend line we can see some correlation between the centrality score and team progress in the World Cup – backing the prediction.
The 2018 World Cup was full of surprises. From defending champions Germany failing to make the final 16, to England’s completely unexpected journey to the semi-finals, there were some teams who defied the prediction:
Russia, Sweden and England all progressed further than their centrality score implied they would.
On the other hand, it’s fair to say Germany, Argentina and Brazil expected more from their national teams.
One often-cited reason for this year’s unexpected results was the uneven distribution of quality sides in each half of the draw for the knockout stages.
By the round of 16, it seemed that the majority of teams in one side of the draw consisted of historically strong teams (including France, Argentina, Uruguay, Brazil and Belgium), and the other with historically weaker teams (like Russia, Denmark, Sweden, Switzerland and England).
Based on our scoring, there might be some truth to that:
Here the teams are colored based on which ‘group’ they belonged to: red being the weaker side, green being the stronger side.
There’s a bit of spread, but generally the weaker teams had a lower centrality score and the stronger teams had a higher score. But in the end, the strongest team won.
There are lots of ways we could measure the accuracy of this prediction, but what matters most are the results. If we’d used centrality scores to predict the winner of each match, how often would we have been right?
Bookmakers make billions by predicting the outcomes of sporting events, but we all know animals are the real oracles when it comes to the World Cup. In 2014 we had Paul the Octopus. How do our predicting abilities compare?
From the 16 matches in the knockout phase, we were right 12 times, or 75%. From the quarter finals onwards, all predictions were correct.
Zabiyaka the Goat, on the other hand, seems to have got most of her predictions wrong.
Graph theory 1-0 Goats
This was a fun experiment to see how far social network analysis can take us when exploring a new dataset, but there are real world uses for this approach.
Often threat analysts work with incomplete data and need to extrapolate insight from the information they have at hand. Analyzing connections is often a great way to get deeper insight into the nodes themselves.
Modern day intelligence (probably unlike football punditry – we’ll go back to our day jobs now) is all about quickly prioritizing from large amounts of incomplete connected data – and in that scenario, graph analysis could be invaluable.
If you’d like to kick off your own evaluation, request one here.
image source: https://buff.ly/2me2cFK
Read our white papers
Registered in England and Wales with Company Number 07625370 VAT Number 113 1740 61 6-8 Hills Road, Cambridge, CB2 1JP. All material © Cambridge Intelligence 2021.
Senegal became the third African country to qualify for the 2018 World Cup by defeating South Africa 2-0 on Friday.
The victory gives Senegal an insurmountable five-point lead atop CAF qualifying Group D, ensuring their place in Russia alongside already-qualified Egypt and Nigeria.
The result eliminates South Africa as well as Burkina Faso and Cape Verde, while Senegal qualify for their first finals since 2002 and will appear at the global showpiece tournament for just the second time in their history after an early strike from Diafra Sakho and an unfortunate own goal from Thamsanqa Mkhize.
The match was a replay of the original November 2016 fixture that was won 2-1 by South Africa, a result that was later annulled by FIFA after Ghana referee Joseph Lamptey was found guilty of manipulating the result of the match.
After a bright opening from the hosts, it was an incisive piece of passing that put Senegal in the lead on 12 minutes.
Sadio Mane found some space on the right and his defence-splitting ball was perfect for Sakho to slide the ball past the masked Itumeleng Khune in the Bafana goal.
The home side might have equalised midway through the first half when Themba Zwane's shot was only parried by Senegal keeper Khadim N'Diaye, and Percy Tau turned home the rebound.
Bafana came even closer a few minutes later when Mkhize's low cross into the box was met by Lebogang Manyama, whose shot cannoned off the crossbar with N'Diaye well beaten.
They were applying constant pressure to the Senegal goal through this period, but the visitors doubled their advantage with a sucker-punch.
Mane failed to get a solid connection on his shot from eight yards that was saved by Khune, but the ball deflected off the back of Mkhize and dribbled into the Bafana net.
Manyama had a free-kick that provided a nervous moment for N'Diaye late in the first half, but the keeper made the save in the top corner of his net.
Sibusiso Vilakazi had the first opening of the second period, but his snap-shot was straight at N'Diaye when a yard either side of the keeper would likely have resulted in a goal.
Bafana's second-most capped player Siphiwe Tshabalala was handed an 89th cap for the final 20 minutes almost four years, but Bafana's World Cup dream had long faded by the time Dean Furman flashed a late 35-yard drive wide.
The teams will meet in the final match in the pool in Dakar on Tuesday in what is now a dead rubber.