Today, I finish my post-season series of research articles with a detailed look at wildcards (the main ones). In this article, you’ll find a gazillion of different charts, notes, observations, as well as a hall of fame and a hall of shame with the best and the worst wildcarders of the season. This article is pretty huge, so be careful. Opinions and critical remarks are welcome ;)
Wildcard Timing in FPL Overall
When do FPL teams normally play their main wildcards?
A brief look at the times when main wildcards were played in FPL during the last two seasons suggests that wildcard timing doesn’t change much from season to season. Which particular gameweeks are most popular depends on season fixtures, but the general picture remains the same:
Almost every second FPL team doesn’t play its wildcard at all (47.5% of teams in both seasons);
Most wildcards are played during the starting spell of a season;
Out of those who play their wildcards, the vast majority do so during the first season half (88% in season 14/15 and 86% in season 13/14).
Theoretically, if wildcards were completely irrelevant to success in FPL, this picture wouldn’t change significantly across the overall rankings. Wildcards are not irrelevant though and are correlated with success in FPL for two primary reasons:
they directly contribute to the final point tally by letting a FPL manager save on the points that he might otherwise spend on extra transfers to change a bulk of his team;
they reflect to a certain extent the level of skills and diligence shown by a FPL manager as more skilled and diligent managers are less likely to waste their wildcards by not playing them at all, playing them badly, or at a bad moment.
Wildcard Timing of Top Teams
So, which moments were particularly good for wildcarding in season 14/15?
For season 14/15, the answer is obvious. 1.1% of all FPL teams, 11% of top 100K, 27% of top 10K, 42% of top 1K, 53% of top 100, 70% of top 50, 100% of top 1 :D played their wildcards in GW34. Saving your wildcard during the first half of the season and unleashing it for DGW34 was the best strategy in hindsight. Of course, such a strategy implied a successful choice of the starting line up and lack of bad luck with injuries.
The declining percentages of wildcards played before GW7 for higher ranked teams indicate that these gameweeks were not good in hindsight.
Does it mean that playing a wildcard early is a bad strategy or was it just this season’s thing?
Season 13/14 suggests that it’s rather the latter. The chart above lacks top 100K data, but the picture is still clear. Most top teams played their wildcards before GW7. The percentages of wildcards played early don’t differ much between top 100, top 1K, or top 10K teams. GW2 was so popular because GW1 was a double gameweek for Chelsea. Playing wildcard late was still a good idea, but the number of top teams that managed/decided to save their wildcards up to that point and still finished in top 10K was much smaller.
The chart below shows what percentage of top teams played their wildcards in first season half (1 – blue bars) and what percentage of top teams played their wildcards in second season half (2 – red bars) for the last three seasons.
The share of wildcards played late is always higher among higher ranked teams, but the amounts of such teams significantly differ from season to season. I think it means that late wildcards are an important advantage for top teams that manage to get up top during the first season half without playing a wildcard; but the level of difficulty of this task (to get up top without playing a wildcard during the first season half) varies from season to season.
How did first season half and second season half wildcarders perform throughout the season compared to each other?
As the chart above shows, the second half wildcarders had a slightly better start but as the season went on and the wildcards were played by the first half wildcarders they started falling behind. The only exception was the top 100K sample where the second season half wildcarders matched the pace of those who wildcarded in the first half of the season. When wildcards were unleashed in GW34, the second season half wildcarders more than covered the accumulated differences.
How was season 13/14 different?
First, it was more difficult to match the pace of the most successful early wildcarders – the top 100 line bends much lower.
Second, there was no DGW in the second half of the season that would have been so favourable to late wildcarders. DGW31 only brought 5-8 extra points for late wildcarders compared to their competitors who wildcardered early.
Did keeping the wildcard during the first half mean for the second half wildcarders that they had to take more point hits to keep their teams in shape?
For the first 10 gameweeks the answer is ‘yes': second half wildcarders spent on average 2 more points on extra transfers. Afterwards, however, early wildcarders started to take more hits (except for the top 100 teams) and had to spend 2-4 more points on transfers in DGW34.
What about season 13/14?
The picture was similar. Late wildcarders ended up taking less point hits than early wildcarders.
To sum up, successful late wildcarders manage not to fall too much behind of successful early wildcarders by taking on average less than one more hit during the starting spell and tend to be more point hit unfriendly afterwards. Points saved on point hits during the final spell are an important source of extra points for them. In season 14/15, GW34 was especially successful because late wildcarders not only enjoyed the points saved by not taking multiple point hits during the double gameweeks, but also because cheap players from Hull and Leicester (who were popular among wildcarders and considered not worth point hits by most non-wildcarders) did so well in that double gameweek.
Average Points Around Wildcard Time
a. Season in General
How important were wildcards and how many points did they normally bring?
Let’s take a look at the points that were gained on average around the wildcard time. The charts below show the difference in average points (net of point hits) scored by wildcarders and the rest of the teams from the respective sample (‘non-wildcarders’):
in the gameweek when the wildcard is played,
for 3 gameweeks before playing the wildcard,
for 9 gameweeks after playing the wildcard.
Keep in mind, that in these charts, wildcarders are compared to non-wildcarders from the same samples, i.e. the average points of non-wildcarders used for comparison were usually higher for higher ranked teams (in an average gameweek of the season, an average top 100 team outscored an average top 100K team by 6.8 points).
Well, these charts look rather similar:
Wildcards were usually played after a relatively bad week. Top 100K wildcarders scored 3.7 points less than non-wildcarders in the gameweek before they utilised their WC. 63% of wildcards of top 100K were played after a GW score lower than the average score of the final top 100K.
The positive effect of playing a wildcard was maximised immediately – in the gameweek of wildcarding – and could be felt for 2-3 more gameweeks. After that, average scores of wildcarders and non-wildcarders converged.
Top teams played their wildcards more efficiently than lower ranked teams: for example, among the top 100 teams wildcarders outscored non-wildcarders by 19.1 points during the first 4 gameweeks after wildcarding, for top 100K this difference was only 8.9 points on average.
In the charts above, wildcarders are compared to non-wildcarders from the same sample. But if you compare, for example, top 1K wildcarders vs. top 100K non-wildcarders, then the wildcards of top 1K look even more successful (for reference: in an average gameweek of the season, average top 1K team outscored average top 100K team by 5.3 points):
b. Average Points Around Wildcard Time by Gameweek
In which gameweeks did playing the wildcard bring most extra points?
The charts below show how the points around wildcard time varied during the season.
In the top 100K, GW34 was the only gameweek when wildcarders outscored non-wildcarders in the pre-WC gameweek. Indeed, at such moments, when a gameweek with blanks (like GW33 where Arsenal and Liverpool didn’t play) is followed by a double gameweek, the potential of a wildcard is often maximised. Those who played their wildcards in GW34 didn’t need to think of future gameweeks while planning for GW33.
Red negative bars are larger and more frequent in the first half of the season. This suggests that most wildcards played in the first half of the season were played rather to ‘fix’ underperforming teams than to maximise the potential of a wildcard.
GW8, GW10, and GW13 were the best moments to unleash the wildcard in the first season half, GW34 and GW36 – in the second season half.
The top 1K and top 100 charts are based on fewer observations, therefore they are more ‘random’, but still worth seeing. A couple of teams in top 1K played their late wildcards very successfully in GW32.
Hall of Fame and Hall of Shame for Wildcarders
The charts above deal with average points and show what extra points were scored by wildcarders compared to non-wildcarders on average. Although in each gameweek there is some kind of a template team that may look ‘optimal’ for wildcarders at that particular moment, every wildcard is different and some people manage to play their wildcards successfully even in the weeks when most wildcarders fail. It works the opposite way too: even in the gameweeks when most wildcarders prosper some people manage to completely screw up their teams. So, let’s take a look at the best and worst examples of wildcarding in the top 100K of season 14/15.
It is not clear though which time horizons should be taken into account when evaluating the power of a wildcard. Some people play a wildcard to fix their teams after a series of bad gameweeks (in this case, pre-WC gameweeks should be excluded from analysis), others plan to play their wildcards at a certain moment and incorporate pre-WC gameweeks into their wildcarding plans. The lasting effect of a wildcard is also varying and it is not clear how many gameweeks after wildcarding should be taken into account. Therefore, I consider two time horizons.
NOTE: Since FPL is undergoing maintenance, to see transfers pages and wildcard histories, for the links below, you need to replace ‘/history/’ with ‘/transfers/history/’ in your address bar. Example: http://fantasy.premierleague.com/entry/7412/transfers/history/
1. Long horizon: 2 gameweeks before, wildcard gameweek, 4 gameweeks after.
Hall of Fame
The champion in this category is well known: it’s ‘Dirty Sanchez‘ by Jon Sumner due to his amazing run at the end of the season which almost brought him the champion’s title. He played his WC in GW34 and outscored the final top 100K average by 138 points since GW32.
Other top wildcards include:
‘Sommerfuglpete‘ by Peter Solgård – GW33, +135 points
‘FC Lenivi babraci‘ by Juraj Fronc – GW34, +133 points
Some successful early wildcards:
‘InBryanWeTrust’ by Bryan McGann – GW7, +133 points (even though he had a terrible score in the gameweek of wildcarding)
‘Blue Ash FC‘ by Ashil Parekh – GW8, +130 points
‘Hiphopopotamus‘ by Peter Chesley – GW4, +132 points
‘Laneni‘ by Milan Mihajlovic – GW4, +123 points (his WC brought him to the top of FPL at one point)
Hall of Shame
For Steve Hartshorne and his team ‘CityAllOverTheWorld‘ the word ‘wildcard’ will be highly offensive after this season. Everything was going smoothly until January: green arrows were frequent, a top 10K spot was around the corner. But everything changed when the January wildcard arrived in GW22. Four red arrows in a row and a fall from 10K to 58K. “Luckily, I still have my main wildcard to stop this slump,” – thought Steve and unleashed another wildcard in GW26. The following five red arrows brought him down to 345K. Fortunately, he didn’t have any more wildcards to unleash and managed to climb back into the top 100K during the final 8 gameweeks of the season so that we could take a laugh at his misfortunes.
Steve lost 128 points to top 100K average during the period from GW24 to GW30. Some other unlucky wildcarders are:
‘Willy Pullbacks‘ by Will Greenhalgh (GW33): -124 points
‘OZAC‘ by Omar Abouchacra (GW34): -118 points
2. Short horizon: wildcard gameweek and 3 gameweeks after.
Hall of Fame
In this case, ‘Blue Ash FC‘ by Ashil Parekh (GW8) tops the rankings: +124 points
‘Aleti‘ by Alberto Paredes (GW8): +122 points
‘kalata‘ by Kaloyan Roidev (GW34): +121 points
Hall of Shame
The champion here actually deserves some kudos. ‘FC NeraZzurri‘ by Server HaYou probably made all the mistakes that a FPL rookie could do in theory. He missed GW1, started the season with what looks like an autocomplete team in GW2, played his wildcard immediately, failed, took a 40 point hit, and was basically the personification of this meme:
The fact that he somehow finished in top 100K with such a start is a great achievement, well done.
His team lost 89 points to top 100K average from GW3 to GW6. Some other examples of extremely unfortunate wildcarding include:
‘The bean flickers‘ by Lee McGavin (GW8): -87 points
‘Mantol‘ by u e (GW34): -82 points
The end