If an index is performing comfortably inside the second decile – indeed, is knocking on the door of the first – you might reasonably conclude there is a fundamental problem with its construction. In the case of MeanReversionMachine, those responsible for that construction must ruefully concede you would be right to do so – though not for the reasons you may initially think.
With three gameweeks left of the 38-gameweek season – and off the back of six consecutive green arrows – the MeanReversionMachine portfolio sits 1,688,216th out of a total of 13,089,564 players. And it is in that last word that we come to the crux of our tracking problem. For one thing, a growing proportion of that 13 million will be disgruntled, disillusioned and, either way, disappeared players who no longer want to play the game.
At the same time, as we discussed a few months back in I see dead teams, a further, smaller chunk of single-use teams are created in an attempt to win one of a multitude of ‘highest weekly score’ leagues and then discarded like so many wet-wipes to clog up the FPL pipes. Given there were just 11.9m teams at the end of Gameweek 5, one might reasonably conclude there are some 1.2m such interlopers.
Guestimating the number of FPL teams whose managers have just given up is a good deal more fraught but, at this late stage of the season, it would not be outlandish to suggest only a third to two-fifths of those who signed up in the first month remain actively engaged. If so, MeanReversionMachine sitting at 1.69m out of a range of 3.97m to 4.76m is rather more respectable – from an index perspective, that is.
One possible solution next season – he types with uncharacteristic optimism – is we look to base an index on the ownership statistics of those managers ranked in the top one hundred thousand or million – although that would involve a lot more work mining a data site, such as the excellent LiveFPL. It would presumably also introduce a whole new set of biases of its own.
What it would not do, however, is address a significant problem now being faced by the portfolio adviser – a complete lack of liquidity. Actively seeking to reflect the squad of the most-owned players in FPL is inherently retrospective, which means we are rarely ahead of the game on price rises and, worse still, lose value with every transfer.
Thus, while the typical squad value of a still-active player is around the £104m mark, the pool of cash the MeanReversionMachine portfolio has to play with is actually a shade under its original £100m. That extra £4m would be enough for a wholly-accurate first 11, with midfielder Wilson becoming Rogers and defender Senesi in for Gudmundsson, but instead we must make do with the status quo (albeit with another banked free transfer):
“After five successive blanks, it will be interesting to see if Semenyo reverts to his early-season mean – or if this poor run is his reversion.
Source: Fantasy Premier League
Not that our skewed team did us too badly last week – if we only care about performance. Once again, Arsenal pair Raya and Gabriel delivered clean sheets while our forward trio of Joao Pedro, Kroupi Jr and perma-captain Haaland combined for 28 points. Not much to report from midfield although, after five successive blanks, it will be interesting to see if Semenyo reverts to his early-season mean – or if this poor run is his reversion.
Finishing up, as ever, with a regular look at our Herdwatch tables of the most bought and sold FPL assets of recent gameweeks, the latest ‘buy’ list shows a continuing appetite for Manchester City, ahead of their double gameweek – and especially for Cherki. He tops the charts for a second successive week as part of a preferred City triple-up that also includes Haaland and O’Reilly. Doku is seeing some love too – unlike Semenyo.
A similar degree of frostiness is being show to Crystal Palace assets – which would be unusual for a doubling team were they not evidently focusing more on their upcoming Europa Conference final than their remaining domestic games. To be fair, interest in the more appealing Palace defenders will be split according to how much budget managers have to spare – and so it is only the sporadically-playing Sarr who makes the latest ‘buy’ table.
The double gameweek is also likely to be the last big opportunity for FPL managers to play their second triple-captain chip of the season. Depending on your risk appetite and/or current rank, the big bets are Haaland on one side and Cherki or O’Reilly on the other. Here on MeanReversionMachine, we do not have the luxury – or curse – of FPL choice and Haaland, as the game’s most-owned player, is again our triple captain. It would be lovely if he could generate rather more than the six-point total he managed for the portfolio back in Gameweek 13.
Source: Fantasy Premier League
