Tomorrow the Netherlands will play their game during the Round of 16 against Romania. In the Netherlands, there is this vibe that everyone is sort of “happy” that we do not need to play against Spain or England in this round and that we are on the good side of the tournament schedule. There can’t be any room for underestimating Romania, this is a great opponent and it will be an exciting match. This can be potentially the last chance for AI to predict the right outcome, so far it is not doing a great job. When I think about Romania, I think about Gheorge Hagi, one of the greatest Romanian players in history. As always, first I would like to share some interesting facts about the history of the Netherlands against Romania:
- Romania only won once during a tournament in 2007, the winning goal was scored by Dorin Goian
- The last victory from the Netherlands was in 2017, 0-3. The goals were scored by Depay, Babel and De Jong.
Before I go to the prediction, I have a couple of things to share beforehand. The whole configuration of the table, preparation of the data, and building of the custom model can be found in this blog article.
Data
There is not much history when it comes to the Netherlands against Romania, but of course, all matches are included. As mentioned in my previous post, it would be awesome to hear feedback or ideas from everyone to make a better model.
Data Source: Netherlands national football team: record v Romania (11v11.com)
Last chance for AI to predict right?
Now let’s find out what the AI Builder prediction model will predict for the match against Romania, will it be:
- Win, W, 806650000
- Draw, D, 806650001
- Lost, L, 806650002
We run the Power Automate Flow that is described in the previous blog article, with the following input:
- Opponent = Romania
- Where = Away
- Type = Tournament
- Last2Decades = Yes
As you can see, the AI Model crushed the game against Romania by predicting 806650000 which is mapped to W(Win). So this means that according to the AI Builder prediction model, the Netherlands will Win the game. With a Likelihood score of 0.6841, it would mean that the Netherlands will win the game and it is quite sure about it. Perhaps it is good for the Netherlands that it won’t take any sentiment into consideration of the Dutch population, because I bet that the score was then in favor of Romania.
Hopefully, the Netherlands will make it to the quarter-final which means a new opportunity to predict or will this be the last chance for AI to predict right? For now, may the best win and enjoy the game everyone.
