Crunching the numbers on PGLT for Gameweeks 1 – 22 of the 2021/22 Season

With more than half a season worth of data now available, we’ll now begin using these data to analyse the best use of transfers. First, we’ll present the data through the end of Gameweek 22. In later articles, we’ll concentrate on why using one transfer is underrated, and why using two or more transfers at a time is overrated. Since that’s so controversial, it’s imperative to present the data first.

I’ve been tracking the Points Gained over the Lifetime of a Transfer (PGLT) for myself and four well-known managers for the season so far. As explained in the linked article, unlike the popular metric of immediate points gained from transfers, PGLT takes into account how well a transfer performs over its entire lifetime: until the next wildcard or until the players brought in are transferred out. Today we seek to answer the following questions:

1.     How many points has each manager gained from transfers: a) in total; b) per transfer made; c) per free transfer made?

2.     Breaking down the PGLT by single moves vs multiple moves, which type have gained the manager more points on average?

3.     Across the entire group of managers, what are the best transfers of each type that were made? (For this, I’ll use PGLT per free transfer as the main metric, which is more generous to moves made for hits than PGLT per transfer. For example, a move producing 15 PGLT for three transfers for a -4 produces 5 PGLT per transfer, but 7.5 PGLT per free transfer.) The three types of transfers assessed are single moves, moves for two free transfers, and multiple moves for hit(s).

Besides myself (MLS), the other managers I’ve been tracking are Mark Sutherns, Andy LTFPL, Joe Lepper, and Magnus Carlsen. I chose these partly because they are top managers and aside from Magnus, very prominent content creators, but also to get an idea of how different styles might influence PGLT. Both Mark and (to a lesser extent) Andy are managers who seem to prefer rolling a transfer every other week as long as they don’t have a key injury or similar crisis to deal with. After rolling their transfer, they will typically use two transfers, or sometimes three with a -4, the next week, rather than using just one and rolling the other like Fábio Borges. As we’ve seen in our Styles of Play series, Mark is the prototypical Orange manager, whereas Andy is less of a purist to any particular colour, but leans Yellow. Joe is the prototypical Yellow (Traditional FPL) manager. Magnus is a Purple manager that I like to model myself after.

Overall Results

So far, the scoreboard in order by total PGLT is as follows:

1.     Joe, 226 PGLT from 28 transfers (19 free transfers); 8.1 PGLT per transfer; 11.9 PGLT per free transfer

2.     Magnus, 200 PGLT from 25 transfers (19 free transfers); 8.0 per transfer; 10.5 per free transfer

3.     Andy, 195 PGLT from 30 transfers (20 free transfers); 6.5 per transfer; 9.8 per free transfer

4.     MLS, 173 PGLT from 24 transfers (20 free transfers); 7.2 per transfer; 8.7 per free transfer

5.     Mark, 102 PGLT from 28 transfers (20 free transfers); 3.6 per transfer; 5.1 per free transfer

The managers are in the same order evaluated by total PGLT or PGLT per free transfer. I rank higher on PGLT per transfer at third; however, by taking fewer hits, I’ve probably missed some opportunities to gain more total points.

Single vs Multiple Moves

This time, I’ll order the managers by the difference in average PGLT per free transfer (PGLT/FT) for single moves minus average PGLT/FT for multiple moves. Thus a positive number indicates that single moves have performed better than multiple moves for that manager, and a negative number indicates the reverse.

1.     Magnus, 12.9 (18.0 PGLT/FT for 8 single transfers; 5.1 PGLT/FT for 17 transfers (11 FT) part of multiple moves)

2.     Joe, 9.9 (19.2 PGLT/FT for 5 single transfers; 9.3 PGLT/FT for 23 transfers (14 FT) part of multiple moves)

3.     Andy, 8.5 (15.7 PGLT/FT for 6 single transfers; 7.2 PGLT/FT for 24 transfers (14 FT) part of multiple moves)

4.     MLS, -3.9 (6.9 PGLT/FT for 11 single transfers; 10.8 PGLT/FT for 13 transfers (9 FT) part of multiple moves)

5.     Mark, -6.0 (1.8 PGLT/FT for 9 single transfers; 7.8 PGLT/FT for 19 transfers (11 FT) part of multiple moves)

Only Mark and I have performed better when using multiple moves than using single moves. My numbers were almost even before Double Gameweeks 21 and 22, but success with West Ham assets in DGW22 raised my average from multiple moves. Magnus, Joe, and Andy have performed better using single moves by large margins. Magnus has gained 72% of his points from single moves, which comprise only 32% of his transfers. Similarly, Joe has gained 42% of his points from the 18% of his transfers that are single moves, and Andy has gained 48% of his points from the 20% of his transfers that are single moves. Meanwhile, my moves have been the most balanced, gaining 44% of my points from the 46% of my transfers that are single moves.

Best Single Moves

Top five moves so far:

1.     (tie) Luke Shaw to Reece James (Magnus, Gameweek 8, +62 PGLT) and Brandon Williams to Joao Cancelo (Joe, Gameweek 7, +62 PGLT): These could be considered a bit of a statistical oddity, since some of the other managers also brought in a Chelsea wing-back and most bought Cancelo, but did it on a wildcard. Still, such an incredible return really is the stuff of dreams, and Magnus acted earlier than the rest of the managers to capture James’ best returns.

3. Antonio Rudiger (injured) to Trent Alexander-Arnold (Andy, Gameweek 8, +48 PGLT): This was an interesting one, because Andy’s original plan following his wildcard was to get Alexander-Arnold back in at the expense of Aymeric Laporte, but Rudiger got injured, so he became the fall guy instead. Rudiger was only out for one week, but Alexander-Arnold’s outperformance still made this Andy’s best move.

4.     David Raya (injured) to Aaron Ramsdale (MLS, Gameweek 11, +44 PGLT): Admittedly Raya’s long-term injury inflates the results for this transfer, but still, bringing in Ramsdale before a good run of form was highly profitable.

5.     Raheem Sterling to Bruno Fernandes (c) (Magnus, Gameweek 22, +40 PGLT): Amazingly, this move was only one gameweek old at the time of writing, but had already produced a 40-point gain after Fernandes’ massive Double Gameweek haul.

Best Moves for Two Free Transfers

1.     Ben Chilwell and Ismaila Sarr to Reece James and Conor Gallagher (Joe, Gameweek 13, +48 PGLT, 24 PGLT/FT): By far the most successful move for two free transfers. Both of the players transferred out in this move scored no points over its lifetime, while Gallagher provided 33 points and James provided 15.

2.     Andrew Omobamidele and Raúl Jiménez to Marcos Alonso and Joshua King (Andy, Gameweek 15, +26 PGLT, 13 PGLT/FT): Andy successfully capitalised on the end of a good run of fixtures for Jiménez to move him to King, freeing up money to upgrade a fodder defender to Alonso.

3.     Son Heung-min and Callum Wilson (both injured) to Cristiano Ronaldo (c) and Demarai Gray (MLS, Gameweek 4, +25 PGLT from 2 transfers, 12.5 PGLT/FT): This move showed the power of having two free transfers to remove two injured players after an international break. I predicted in advance that most managers, after rolling their free transfer in Gameweek 2, would spend both in Gameweek 3, leaving them without a second transfer in Gameweek 4 when they would need it more. I made sure to use only one of my transfers in Gameweek 3 and roll the other, and it paid off.

4.     Mohamed Salah and Tino Livramento to Son Heung-min and Diogo Dalot (MLS, Gameweek 19, +18 PGLT from 2 transfers, 9 PGLT/FT): Removing Salah for Son before a Liverpool cancellation paid off, thanks to the heroic penalty-saving efforts of Kasper Schmeichel the next gameweek!

5.     Harry Kane and Raphinha to Cristiano Ronaldo and Jarrod Bowen (Joe, Gameweek 16, +17 PGLT from 2 transfers, 8.5 PGLT/FT): Removing Kane for a Blank Gameweek in favour of Ronaldo paid off, despite removing Ronaldo the next week.

Best Multiple Moves for Hit(s)

1.     Conor Gallagher, Ollie Watkins, and Reece James to Jarrod Bowen (c), Michail Antonio, and Craig Dawson for a -8 (MLS, Gameweek 21, +42 PGLT from 3 transfers, 42 PGLT/FT): Not surprisingly, the best move for hits was moving for West Ham players for their Double Gameweek. My PGLT numbers were better than others due to just bringing in Bowen that week and captaining him, while removing three players who did not play.

2.     Phil Foden and Ben Chilwell to Diogo Jota and Reece James for a -4 (Andy, Gameweek 13, +25 PGLT from 2 transfers, 25 PGLT/FT): This is the best multiple moves in a single Gameweek. Both players who were sold were injured, allowing for significant gains even though James had gone off the boil by this point.

3.     Son Heung-min and Ivan Toney to Jarrod Bowen and Emmanuel Dennis for a -4 (Mark, Gameweek 16, +24 PGLT from 2 transfers, 24 PGLT/FT): A common theme in these moves, bringing in Bowen, was also present here. Son had a blank gameweek and Toney had Covid, so there was less risk than normal in removing them.

4.    Mohamed Salah and Joshua King to Cristiano Ronaldo and Jarrod Bowen for a -4 (Andy, Gameweek 19, +23 PGLT from 2 transfers, 23 PGLT/FT): Both Salah and King had blank gameweeks, making a move to Ronaldo and Bowen profitable even though captain Ronaldo greatly underperformed vice-captain Bukayo Saka (16 points).

5.     Adama Traore and Patrick Bamford to Adam Armstrong and Ismaila Sarr for a -4 (Joe, Gameweek 6, +20 PGLT from 2 transfers, 20 PGLT/FT): This was a deliciously understated move with a long lifetime. Replacing the injured Bamford and struggling Traore was a good use of a hit, even for two budget attackers.

Brief comment

Despite the widespread prejudice in the FPL community against single moves (and especially “sideways” single moves), three of the five managers have averaged significantly more PGLT with single moves than with multiple moves. These three have each averaged between 15 and 20 PGLT per single move, but only between 5 and 10 PGLT per free transfer involved in multiple moves. Mark’s results are more what the advocates of multiple moves would expect, but mainly because his results from single moves have been worse. The best single moves often had very long lifetimes where they continued to pay dividends over many months. Interestingly, none of the top multiple moves for hit(s) involved two free transfers plus hit(s); four of them involved one FT and one hit, and one involved one FT and two hits. The best move for two free transfers plus hit(s) was Magnus’s quadruple move for a -8 in Gameweek 16, producing 16.5 PGLT/FT.

In future articles we’ll try to explain these results and explore what we can learn from the data.