Showing results for 
Search instead for 
Did you mean: 

Engine Works Blog

Under the hood of Alteryx: tips, tricks and how-tos.

Wow what a few weeks!! Straight to the point, England made it out of the group stage and Germany are nowhere to be seen. The bookies, the pundits, CavMis ... we did not see a shock German exit.

However, whilst we were wrong about Germany, our model on a number of occasions came through. The teams to make it out of the group stage we predicted 13 correct. We always knew there would be upsets and wildcards. So Denmark, Sweden and Japan well done you really earnt those spots.

Thinking back to our event with Paul Merson, we predicted Senegal will do well and had a high chance to make it through to the round of 16. Senegal nearly went through only just missing out. We said that group would be hard to call and it really was.

Elsewhere it looks like the model has worked well.

As we said at our event, both Nick and I will factor in new data once the group stage is over and create our CavMis v2 model.

What's new?

Now that we have the results, we thought it best to add additional weights when winning above a certain goal margin per match. Our weightings for winning from behind returns, as does the base Pythagorean formula. But now we can use the formula on known outcomes for the current tournament to add to our CavMis v1 model. Let's now put this revised model in action. This is our final prediction for how we see the knockout phase playing out and most importantly who is going to be taking home the cup! (Green highlight for teams to progress through.)
Round of 16.jpgRound of 16
Quarter Finals.jpgQuarterfinals
 Semi Finals.jpgSemifinals. 3rd place: Uruguay.
Final.jpgFifa World Cup 2018 Finals
Looks like its Brazil to win based on our Final model... Paul Merson may have been right all along 🙂

Both @NickC and I had great fun working on this and hopefully it has inspired you too.