Dream XI for every formation after GW29

How would have the best set and forget XI done by this point (at the end of Game week 29) ? Or how much points difference there could have been between dream XIs (where only max 3 players from a team is allowed) of every formation ?

I wrote an algorithm to find that out and the results were interesting, specially when I looked upon to find out the current #1 has total 1878 points.

Please note, the algorithm:

1. Does not consider the total budget, meaning it might not have been possible to put 4 other bench players within the initial FPL budget of 100 in some cases. Eg. Best 343 would have cost 87 & Best 352 would have costs 89 for 11 players alone.
2. Does not consider players who joined the league late (eg. Alonso) - hence it would have been impossible to have this team starting from game week 1. However, it also means that someone else in that position could have played in those missing game weeks and the total would have been even more.
3. Does not consider the games missed though injury, suspensions or any other reasons - again someone from bench would have replaced such players and the total would have been even more.
4. Allows max 3 players from any team - FPL rule.

Now here are the teams:

Best 541

Total pts = 1809

Initial cost = 79

Current cost = 86

                                       

Best 343

Total pts = 1867

Initial cost = 87

Current cost = 93.5

Best 352

Total pts = 1873 Initial cost = 89 Current cost = 95.3

Best 433

Total pts = 1840 Initial cost = 84 Current cost = 91.1

Best 442

Total pts = 1841 Initial cost = 81.5 Current cost = 88.6

Best 451

Total pts = 1847 Initial cost = 83.5 Current cost = 90.5

Best  532

Total pts = 1808 Initial cost = 81.5 Current cost = 88.5

till next post :)

Read More

New Table: Player's Points Breakdown

The table breaks down the Total Fantasy Premier League points obtained by every player who has made an appearance during the season.

•  Here is how it works: Every column lists the total points obtained by the player for that event, so far this season. Eg: a midfielder obtains 5 points for scoring a goal, and if he has scored 10 goals during the season so far - the value shown under column 'Goal' would be 10 * 5 = 50. 
• The 'Rest' column includes the sum of points obtained by the player for 'Making appearances', 'Playing for 60 minutes or more in a game' and in case of Goalkeepers points obtained from 'Shots Saved'. 

For ease of finding later, this table is directly placed under the menu 'Season 16/17' 

Read More

Introducing: Transfer Anomaly Detection Viz.

The chart represents the  rate at which any player is transferred in or out in current game week. EG: if the score of a player is +50, it means that he is being transferred in at the rate of 50 for every single transfer out. And vice-versa for -ve scores.

25 players each of highest scores from +ve and -ve values are plotted. Rest are ignored.

If the FPL cost of player is changed during the week (price rise or drop), the color of the marker representing the player is changed accordingly.

As always, this table will be directly placed on under the menu 'Season 16/17' to make it always easily accessible.  Hope everybody finds it useful. Till next post.

Read More

Introducing: Player's Upcoming fixtures difficulty rating table

This was a popular request - a table where you can see all the players, their status, total points, points per game and their upcoming fixtures. This was also a very logical request as this would enable anyone to make the work less tedious when comparing players to pick every time (the official FPL site only allows to check the upcoming fixtures of one player at a time).

So here it is. I added only next 5 fixtures just to make the width of the table comparable to the rests we have in this website. Adding many more was also possible but then the users would be moving left and right all the time using the sliders and that may not be ideal experience. The colors used are exactly same as that one used on official FPL site to rate the FDR (Fixture Difficulty Rating) and vary from Dark Green to Dark Red representing the difficulty ratings on the scale of 1 to 5.

As always, this table will be directly placed on under the menu 'Season 16/17' to make it always easily accessible.  Hope everybody finds it useful. Till next post 

Read More

Reality Check: Most Selected XI Vs Dream Team (Game Week 4)

This post is continuation of series where we check and analyze how good the FPL managers predict as a sum by comparing the 'Most Selected XI' of the game week to the 'Final Dream Team'.

The most selected XI is based on the percentage ownership of the players among all FPL managers before the kick off of the Game Week.


The results for Game week 4 are out.

Most Selected XI (Game Week 4)

Most Selected XI

Formation    : 4-3-3
Total Points  :  18 (yes 18,  )

Final Dream Team
Formation    : 3-5-2
Total Points  : 140

Dream Team (Game Week 4)

18 out of possible 140 - i.e. less than 13%. And not one player out  of 11 from MostSelected XI could make it to the Dream Team. 

Till next post

Read More

Season Points distribution Treemap

Which team is leading the overall #FPL points table ? Which is at the bottom ? Are there some  players scoring chunks of their team's FPL points for particular teams ? Can they be good bet for your squad ?

I have tried to make a simple interactive Treemap visualization to answer all such questions. The size of the boxes represent their percentage share of overall points. It is multilayered interactive visualization and you can mouse click on anywhere on it to reveal the details (mouse click again will take you back to main chart again).

Oh yea, this Viz will be automatically updated with #FPL scores, thus you can come back again later during the season to make use of it. For later convenience, this will also be put under VISUALIZATION menu, so you do not have to go looking around for this particular post.

Have fun with it. Till next post.

Read More

Value Check: Visualizing Points per Million Cost of Players

Another day, another post. Two quick Data Visualizations of actual value of your FPL players. 

1. Points per million cost
2. Points per game per million cost

I was thinking of making these charts once the season little matured (at least 5 game weeks) - but then, there I was this morning, night after the deadline day -  staring at my monitor thinking how to get through the day. Thus, here they are. Anyway, these charts will be updated automatically as the data changes every game week, so why not ? 

Please note, these charts will be placed under the menu 'Visualizations' - so next time when you are here, you wont have to find this particular post. I believe, these charts will reveal more after some more game weeks - so please come back.

Also note that the charts are interactive (as shown by the GIF above) and you are encouraged to try stuffs.

Now , when you look at them, both charts are somewhat identical,

• On Y-axis is the total minutes played by any given player. 
• On X-axis, first chart has total points while the second has points per game of the player
• The size of the bubble represents the overall value of the player i.e. in case of 1st chart it is Points per million Cost (points/price of the player)  while in case of 2nd chart it is Points per Game Per million Cost. Thus the bigger the bubble size of any particular player, the more value he represents.
• Rest you can figure out - fantasy football mangers are smart bunch. 

Now first check how the chart 'Total Points per Million Cost' looks. We have had only 3 game weeks and hence many bubbles look close to each other, as the season progress - we can expect the chart to expand more and those bubbles to pull apart from each other more.

Often, 'Points per Game per Million Cost' can be better chart of these two make good judgement, even though it is more prone to outliers (a player who played only one game the whole season but scored 10 points can lead to very wrong assumption - hence it was important to include minutes in the chart. Always check the minutes - more the minutes played, better the consistency of the player,  when all other variables are same).



Have fun with above and if you have any request for any type of Data Tables or Visualizations let me know. You can connect me on twitter @fplAnalytics or email at crisscrossdata@gmail.com. 

Till next post.

Read More