Fantasy Football

Fantasy Premier League (FPL) 2017/18 Strategy: Three-Way Budget Defender Rotation

About this time last season I wrote a blog post about the best three-way rotating budget defenders for the forthcoming season based on the fixture favourability. The basic premise was to plan to have two premium defenders and then the final defensive slot in a 3-4-3 / 3-5-2 formation taken by a rotating cast of three budget (£4.5m or lower) defenders from mid-lower table sides. The post proved to be popular and the analysis did help me with player selection at the start of the season, so I have decided to do the analysis again for the 2017/18 season.

Methodology

My task is to find the best combination of fixtures for three clubs from whom I can pick three regular starting defenders for a low price: £4.5m each, or £13.5m in total. A favourable fixture will be determined as:

  1. A home game
  2. Against a weak opposition to increase the probability of a clean sheet

Obviously at this stage the definition of ‘weak’ teams is somewhat subjective and subject to change so apologies if this list doesn’t conform to the readers’ thoughts, but I have determined the most favourable opposition from a defender’s perspective to be:

  1. Burnley (BRN)
  2. Brighton (BRT)
  3. Huddersfield (HUD)
  4. Stoke (STO)
  5. Swansea (SWA)
  6. Watford (WAT)
  7. West Brom (WBA)
  8. West Ham (WHM)

We are looking for cheap defenders who are almost guaranteed starters for their clubs, and I would expect the following 12 clubs to have at least one starter at £4.5m or less (please note that this article is being published before the release of the player price list for 2017/18):

  1. Bournemouth (BOU)
  2. Brighton (BRT)
  3. Burnley (BRN)
  4. Crystal Palace (CPL)
  5. Huddersfield (HUD)
  6. Leicester (LEI)
  7. Newcastle (NEW)
  8. Stoke (STO)
  9. Swansea (SWA)
  10. Watford (WAT)
  11. West Brom (WBA)
  12. West Ham (WHM)

As with last season, I will look for a combination of fixtures from these teams that are favourable for the first 33 weeks of the season. I have only looked at the first 33 of 38 weeks because in FPL, this is when the double gameweeks kick in – players playing twice in a single week, which also means there are blank gameweeks elsewhere in the calendar – and I am more likely to use my second unlimited transfers wildcard. At this stage, in the final run-in when long-term thinking goes out the window, I will likely adopt a very different strategy.

Update: I have had comments asking me to look at the first few weeks only and for an extended period in the middle of the season between wildcards. Personally, because I do not know when I will play the first wildcard I like to plan for the bulk of the season in the squad positions, which is why this analysis focuses on the first 33 gameweeks. However, in the appendix there are now the best combinations for gameweeks 1-6 (pre-wildcard) and gameweeks 7-33 (between wildcards) for those readers interested.

Home Fixtures Only

There is one combination of teams that yield 20 favourable home fixtures in the first 33 gameweeks, which is:

  • BRT / NEW / SWA

Following this there are 19 combinations with 19 favourable home fixtures (see appendix for the list of top combinations).

The most favourable combination here features two of the promoted clubs and last season’s strugglers Swansea. Picking a defender from each of these clubs and rotating them will result in a decent home fixture in 20 of the first 33 gameweeks, however there are occasions when none of these three players will face a favourable opponent, such as gameweeks 20 and 24 when the three clubs face Chelsea, Man City and Liverpool home and away, gameweek 17 (tot, EVE, MCI).

Therefore, it would be more beneficial to ensure that if our third defender, if he does not face a weak team at home, is at least facing a weak team away from home. As a result, it is necessary to look at the most favourable combinations when our budget defenders are playing weak teams home or away.

Home and Away Fixtures

The amount of favourable combinations under these conditions increases substantially. When considering home fixtures only the best combination has 20 favourable fixtures, but when we include favourable away fixtures every combination has at least 21. The best combination has a perfect rotation of 33 favourable fixtures in 33 gameweeks which is:

  • CPL / LEI / WAT

This is followed by two combinations with 32 favourable fixtures (CPL / BOU / SWA, CPL / LEI / SWA) and seven with 31. The green bubble in Figure 1 below shows the CPL / LEI / WAT combination (the orange is the BRT / NEW / SWA mentioned above) alongside the rest of the 220 combinations. Of particular interest are the ones that have a good amount of favourable fixtures with a high proportion of those at home. There are six combinations I am interested in analysing, isolated in the dotted box.

Figure 1: Favourable Home Fixtures Only vs. Favourable Home and Away Fixtures Combined (n=220, bubbles donate frequency of occurrence)

Capture X

Notable combinations

CPL / LEI / WAT

  • 33 favourable games (no weak fixtures)
  • 18 favourable games at home (15 away)

CPL / BOU / SWA

  • 32 favourable games (one weak fixture)
  • 19 favourable games at home (13 away)
  • Notable weakness: gameweek 18 (LIV, lei, eve)

CPL / LEI / SWA

  • 32 favourable games (one weak fixture)
  • 18 favourable games at home (14 away)
  • Notable weakness: gameweek 18 (lei, CPL, eve)

BOU / NEW / SWA

  • 31 favourable games (two weak fixtures)
  • 19 favourable games at home (12 away)
  • Notable weakness: gameweeks 17 (mun, EVE, MCI) and 18 (LIV, ars, eve)

BOU / HUD / STO

  • 31 favourable games (two weak fixtures)
  • 18 favourable games at home (13 away)
  • Notable weakness: gameweeks 7 (LEI, TOT, SOT) and 18 (lei, tot, sot)

BOU / STO / SWA

  • 31 favourable games (two weak fixtures)
  • 18 favourable games at home (13 away)
  • Notable weakness: gameweeks 4 (ars, MUN, NEW) and 23 (ARS, mun, new)

The best combination is CPL / LEI / WAT in terms of fixtures because they have no weaknesses; every week one of them will play against one of the eight weakest sides I’ve chosen at the start of this article. However, this option may not be ideal from a players’ perspective. Crystal Palace will have starting defenders which are proven attacking threats in FPL, such as Patrick Van Aanholt and Scott Dann who are likely to come in at £5.0m+. Similarly the attacking instincts of Leicester’s Christian Fuchs and newly acquired Harry Maguire will probably push them into a higher price bracket. This leaves fewer starting spots for the cheaper players, so options may be limited (we saw this last season with Palace, as Papa Souare was the only likely starter at £4.5m).

From a strategic perspective it is worth considering Bournemouth, Palace and Swansea as a combination because that opens up the prospect of a single transfer at some point in the season (should injury strike) making another of the top six combinations (BOU > LEI to make CPL / LEI / SWA; CPL > NEW to make BOU / NEW / SWA; CPL > STO to make BOU / STO / SWA). Bournemouth and Swansea both had woeful defensive records last season but Eddie Howe is likely to focus on reinforcements over the summer and Swansea did improve somewhat under Paul Clement in the second half of the season. Additionally, with Charlie Daniels no longer on penalties it is likely that a lot of the players from those two sides (Alfie Mawson aside) will be priced at £4.5m.

Favoured Combination

Ultimately a lot is going to depend on the prices and squeezing as much value into the defence as possible (if that is your strategy). Right now the obvious move would be CPL / LEI / WAT but this will depend on whether any of the £4.5m players are guaranteed starters; you can hope for a bargain in one of the teams, but both Leicester and Palace having decent budget starters remains to be seen. Of course the other angle to consider is that both Watford and Palace will be going into the season with new managers, and Craig Shakespeare will be having his first experience of pre-season management. There is perhaps a little too much uncertainty here for comfort.

The combination that feels most comfortable from the above instinctively is BOU / STO / SWA. All are established Premier League clubs (although only just, in Swansea’s case), and there is likely to be at least a couple of £4.5m starters in each team. Additionally, with the returning Jack Butland and the upward trajectory of Swansea since Clement arrived, there are reasons to feel cautiously optimistic about the prospects of each in the coming season from a defensive perspective. Furthermore, their weak gameweeks (4 and 23) are a mirror of each other, during which Swansea play Newcastle in both. Whilst Newcastle’s size and manager mean they are likely to be a solid mid-table size next season and not one of the ‘easier’ fixtures of the season, it is likely far from the most difficult of games to navigate for Swansea. If you fancy their chances in those two matches of keeping a clean sheet, then the combination of these three teams suddenly has 33 out of 33 favourable fixtures. Tentatively, this is the combination I am looking at for next season’s 3rd, 4th and 5th defenders.

Figure 2: Fixtures in a GW1-33 Bournemouth / Stoke / Swansea defensive rotation (most favourable fixture highlighted in orange)

Capture X

Combinations to Avoid

The worst combinations are detailed in the appendix, but special mention should go to Burnley, Huddersfield, Swansea which is a horror show featuring 12 weak gameweeks and only 13 favourable home fixtures.

BUR / HUD / SWA

  • 21 favourable games (12 weak fixtures)
  • 13 favourable games at home (8 away)
  • Notable weakness:
    • gameweek 1 (che, cpl, sot)
    • gameweek 3 (tot, SOT, cpl)
    • gameweek 5 (liv, LEI, tot)
    • gameweek 9 (mci, MUN, LEI)
    • gameweek 10 (NEW, liv, ars)
    • gameweek 13 (ARS, MCI, BOU)
    • gameweek 14 (bou, ars, che)
    • gameweek 19 (TOT, sot, CPL)
    • gameweek 22 (LIV, lei, TOT)
    • gameweek 25 (new, LIV, ars)
    • gameweek 26 (MCI, mun, lei)
    • gameweek 31 (CHE, CPL, SOT)

 

Appendix

Figure 3: Top three-way defensive rotation options (GW1-33)

Capture XI

Figure 4: Worst three-way defensive rotation options (GW1-33)

Capture XII

Figure 5: Best three-way defensive rotation options (GW1-6)

Capture

 

Figure 6: Best three-way defensive rotation options (GW7-33)

Capture II

Advertisements

3 thoughts on “Fantasy Premier League (FPL) 2017/18 Strategy: Three-Way Budget Defender Rotation

  1. Thanks for posting this Peter.
    I’m posting this if it helps in your analysis.
    I read your article on this last year, thought it was a decent strategy and used it myself.
    Using this strategy my 3 at the back averaged 4.05 pts per match played each and the total spent on my 5 defenders for gwk1 was 25.5m which freed up a lot of money for the front 8

    the three rotating defenders I started with were WBA, MID, MAN C
    I refined this to WBA,MID,SWA on a gwk6 wildcard. with 13.1m spent on the 3 rotating defenders and 24.1m for the 5 defenders in total.

    My defender points played for the whole season was 486 and played 3 at the back for 34 gameweeks, never captained any of them.
    hope this helps and good luck for next season.

    Buzzy

    Like

    1. Peter. Top notch again! I was highly impressed with last year’s methodology and was hoping you would do an update for this coming season. I just got finished a similar analysis using your idea.

      My bottom 10 differed from that of a few bookie sites, so I ended up using 11 “weak” teams and sorted through all the variations, listing their HOME (strong fixtures) and their weaker away matches against favorable opponents. Obviously the larger pool of teams, my results were a bit different than what you results looked like.

      The best rotating defense I came up with was BOU/SWA/WAT, a total of 25 home games, 8 away, all against the “lower” more favorable teams. Second most favored rotating defense would be BUR/STO/WBA. A total of 24 home games, 9 away, all against the more favorable teams. Third best HUD/WAT/WBA, 23 home and 10 away.

      The teams I used in my calculations were
      BOU
      LEI
      WBA
      STO
      CPL
      SWA
      BUR
      WAT
      NEW
      BRI
      HUD

      The difference between me and the bookies site were BOU and LEI, so I included both. I’d be more than happy to make all the data available. Hopefully it would help others. Appreciate your work on this post!

      Like

  2. Hi Peter. This article is excellent. I will likely be employing two premium defenders from United, Chelsea or Spurs and then three rotating defenders priced at 4.5m each. The conclusions drawn from your analysis will likely inform my choices for the three rotating defenders! Cheers and keep the articles coming!

    Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s