Metrics – Appraising The Chip Strategies

With another season now passed I want to reflect on it and analyse and in particular the variety of strategies used when activating the Fantasy Premier League chips.

We had two major Double Gameweeks this season, 34 and 37, and we also had two Gameweeks that featured a significant number of postponements, namely 31 and 35.

There was much debate during the season about how to approach these Gameweeks with the chips in mind.

One of the areas of contention was whether to use the Free Hit for the blanks in Gameweek 31 or 35 (or even in one of the Double Gameweeks).

There was also plenty of debate on whether to use the second Wildcard early in the second part of the season or perhaps just before one of the Double Gameweeks.

Finally, there was doubt about which Double Gamweeek was best suited to the Bench Boost and the Triple Captain.

I hope to give some insight on all of these questions.

The Methodology

I believe that each of these queries cannot be answered in isolation.

For example, to know whether the Free Hit in Gameweek 31 was better or worse than in Gameweek 35 it is not enough to look only at the results from those weeks, because the deployment of the Free Hit chip is a part of a bigger plan affected by the Gameweeks around it.

Therefore, I decided to look at how managers performed in the last 10 Gameweeks of the season collectively.

These start with Gameweek 29, which was around the time the chip debates kicked-off, with many preparing for the Gameweek 31 blanks either with transfers or with the activation of the Free Hit.

To avoid making the data biased with only success stories I gathered data for a random sample of 25,000 active managers in the top one million, recording when they used their chips and what total score they achieved in the last 10 Gameweeks.

The following are therefore not only statistics for top managers but for everyone who kept playing until Gameweek 38.

Hopefully, with such a big sample the results will be representative of the effect of different chip strategies on overall performance.

I will first present the statistics of each chip used, separately. Then I will show the results for the most used and most successful combined chip strategies.

The Free Hit

30.0% of Free Hits were played in Gameweek 31 while 12.61% were activated in Gameweek 35.

We see that the average scores achieved by both set of managers are very close, with a difference of only two points in favour of the Gameweek 35 Free Hits.

So, it is possible that playing the Free Hit in either Gameweek was equally effective.

We also see that managers who activated the chip in any other week suffered some damage. Even the 4.74% who used it in Double Gameweek 34 lost around 20 points.

The results suggest that the Free Hit was best deployed in a Gameweek with postponements, at least based on this first season.

The Wildcard

The majority of second Wildcards were either activated as soon as the chip became available in Gameweek 22 or right before the Double Gameweeks.

We see that the most successful use of the Wildcard chip was in Gameweek 32 (599.5 points), followed by Gameweek 36 (596.6) and Gameweek 33 (592.3).

All these successful users were right before the Double Gameweeks.

There were some arguments before Double Gameweek 34 about whether to Wildcard directly after the blanks, so in Gameweek 32, or wait one week and Wildcard in 33.

The numbers suggest that waiting that extra week was not the best move and caused managers to lose around seven points on the Gameweek 32 Wildcard, while the managers who Wildcarded late in Gameweek 36 suffered only a three-point loss on average to that group.

The Bench Boost

The most popular Gameweek for Bench Boosts was Double Gameweek 37 (28.72%), not 34 (22.18%). However, the scores suggest that managers who activated Bench Boost in Gameweek 34 are the ones who earned more points over the final push.

They got 7.4 more points than Gameweek 37 Bench Boosters, and at least 20 more points more than managers who Bench Boosted at any other window, outside of a Double Gameweek.

The Triple Captain

Double Gameweek 37 was managers’ favourite window for activating the Triple Captain chip (31.24%), followed by Double Gameweek 34 (17.85%).

The scores, however, suggest that it was the 7.73% managers who activated it in the final Gameweek who fared best.

But the difference between that and the Gameweek 37 case is negligible; generally, managers who used the Triple Captain in Gameweeks 37 and 38 achieved comparable results.

In summary, individual chip results suggest that the best Gameweeks to use the chips this season have been:

Free Hit: 31 and 35.
Wildcard: 32 and 36.
Bench Boost: 34.
Triple Captain: 37 and 38.

But this doesn’t tell the whole story. There might be a combination of the chips used together that resulted in even stronger results.

The Collective Chip Strategy

Now let’s look at the more accurate picture with all the chip strategies in play.

I will present the most used chip strategies for active managers this season and their results. I only included strategies that a reasonable amount of managers used so that the statistics remain meaningful.

As an example, many odd strategies were implemented by a maximum of ten managers in my sample, those I omitted because their average would not be a good indicator of strategy effectiveness. This means that there could be other potential strategies, which almost no one actually used, that would have fared better

The most used strategies are those much discussed on Fantasy Football Scout which consisted of Wildcarding in either of Gameweeks 32 or 33, then Bench Boosting in 34, Free Hitting in 35 and Triple Captaining in 37. However, they were not the most successful strategies.

Instead, the strategy that won the award for the most effective is:

“Strategy A” which was to Free Hit in 31 to cover the postponements, followed by the use of transfers to load up on Gameweek 34 Double Gameweek players, then Bench Boost in Gameweek 34, Wildcard in 35 to acquire Man City and Arsenal assets, then Triple Captain in Gameweek 37.

I didn’t see this tactic used too much, but managers who did use it scored an average of 10 more points than the more popular strategies.

In hindsight, it does seem like a very efficient plan, particularly when other popular strategies missed out Man City and Arsenal assets in Gameweek 36 and even 37, a problem that “Strategy A” solved.

The second best strategy was even less popular and was deployed by only 0.2% of active managers.

It is “Strategy B”, which is very similar to the most popular route or “Strategy D”, but with the use of the Triple Captain chip in Gameweek 38 instead of 37.

Again, in hindsight, this makes sense as most of the Triple Captain candidates in Double Gameweek 37 failed, while Gameweek 38 was, as usual, high scoring.

Next come two very close strategies, C and D, which are very similar, with the exception of the use of the Free Hit in Gameweek 31 vs 35.

This highlights that there is no significant difference between Free Hitting in Gameweek 31 or 35 if the popular strategy was used when activating the other chips.

While Gameweek 35 Free Hit managers enjoyed a strong return, they seem to have lost some ground around it – particularly in Gameweek 36 when they got their default squad back; this put them on equal footing with the Gameweek 31 Free Hitters.

It is worth noting that most of the best strategies involved using the Bench Boost in 34 and Triple Captains in 37 or 38.

We should also acknowledge that the popular strategies didn’t fail, “Strategy D” was 12.5 points short of the top “Strategy A”. Of course, these 12.5 points could have a significant impact on rank, but still “Strategy D” was better than so many strategies not listed in the table and considered here.

Meanwhile, popular strategy G (which is the same as D but with delaying the Wildcard one more week to Gameweek 33) works out as 25 points lower than the optimum strategy.

Waiting this extra week with the mediocre Gameweek 31 players in the team and missing on so many price rises (and falls) really damaged this approach.

Personally, depending on how the fixtures fall, I will consider using my Triple Captain chip outside of the Double Gameweeks, maybe over the final weekend to mimic “Strategy B”.

Myself and so many managers have been burnt by the failings of the Triple Captain in Double Gameweeks over the last three seasons (remember Andy Carroll?). While Gameweek 38, with its plethora of goals, could be far more effective, particularly if there are for teams and players with motivation on the final day.