Why humans make any prediction ?
- To feel in control.
- To give the impression they know what they are doing.
- Or just to get some attention - because others are listening.
Predicting comes naturally us. Our brain has to work really fast - that is its nature and it does so by predicting the next move analyzing our actions and the surroundings. This is why most of our day to day work are done in some automatic cycle efficiently without much effort.
Thus may be we can not stop predicting - even when we dont want to do it. It is the natural tendency, even in most complex and extremely unpredictable circumstances. Most of these predictions fail but that does not stop us from making another prediction - it only makes us to predict the next similar situation differently - adds a new variable to consider the next time.
As of now, there are 3,630,948 FPL managers trying to predict the best possible fantasy team for next game week. When the results are out, the best possible team of any particular game week will be officially called the 'Dream Team'. Thus in a way, the goal of every FPL manager is to get as close as possible to this 'Dream Team'.
We can have the data of percentage ownership of every player any given time - and hence we are able to make an eleven of most selected players each week before the game week kick off. Lets call this 'XI' - the 'Most Selected XI'.
Now, by the theory discussed previously, this 'Most Selected XI' is actually cumulative social prediction of the 'Dream Team'. The total of 3,630,948 FPL managers are trying to predict the best players in each position and hence this 'Most Selected XI' is the prediction of best possible players of each position for the game week - the ultimate 'Dream Team'.
Now lets see how well the humans actually predict
Game Week 1
Most Selected XI
 |
| Game Week 1 - Most Selected XI |
Formation : 4-3-3
Total Points : 35 (LOL
)Official Dream Team
 |
| Game Week 1 - Official Dream Team |
Formation : 3-5-2
Total Points : 109
35 out of possible 109 - i.e. 32.11%. Stop predicting now and making fool of yourselves humans.
Lets compare the both XIs side by side with an interactive visualization below:
The biggest difference here is the purple bars of both XIs, which represent the total points from Midfielders.
Game Week 2
Most Selected XI
 |
| Game Week 2 - Most Selected XI |
Formation : 3-4-3
Total Points : 57
Official Dream Team
 |
| Game Week 2 - Official Dream Team |
Formation : 3-4-3
Total Points : 127
Little better % this time 57 out of possible 127 is 44.88%.
Again the comparison of both XIs side by side with an interactive visualization is below:The biggest difference in game week 2 too remains the Midfield. For sure, Midfield is the most confused position yet and there are possibilities of massive improvements for all FPL managers in that position.
However, overall the Most Selected XI of game week 2 is much better (44.88 %) than that of game week 1 (32.11 %) and may be as more week passes the gap will be better (assuming, FPL managers get better at predicting). May be, may be not - we will see.
Till next post.