FOUR types of selection decision in FPL

 

It occurred to me the other day, when I was writing about Phil Foden's recent improbably hot run of form, that we may discern 4 main types of selection decision in FPL, differentiated by their timing.

Timing of a transfer: the FOUR TYPES

1)  Anticipatory/Speculative/Precognitive

If you go in for a player before he's started to show any clear signs of form - just following a 'hunch', or because you have some sentimental attachment to him, or because he's a big-name player who's often done well in the past (ahem, Mo Salah...) - you may sometimes get lucky with that. But you have to appreciate that it's a very risky play. Going in so early for someone is inevitably a big gamble, even if you may have some good reasons - though very subtle and uncertain ones - for the choice. Those reasons might be founded on past history (e.g., the player usually recovers strongly from a brief injury absence, regularly displays a particularly keen hunger to return to the fray after missing a few games for any reason, always bounces back strongly after just a few poor performances, often has a great game against this particular opponent...), a positive shift in team form or lineup or tactics, a favourable run of upcoming fixtures, some background information which may suggest a likely psychological lift (a spat with the manager or a legal problem resolved, a family problem improving, a bereavement receding into the past), or subtle indications on the pitch of a possible turnaround in fitness or confidence that might herald a sudden, imminent uptick in points-scoring form. If you weighed up such factors, and subsequent performances by the player seemed to vindicate your assessment, you may congratulate yourself on a perceptive early transfer decision. But more often, FPL managers who go in early for someone are just taking a wild punt, based on no substantive rationale at all; and if that player suddenly hits a hot streak of form out of nowhere, they've just been incredibly LUCKY - but they'll never admit that. [Disgruntled rivals may complain of them relying on a crystal ball or a time-machine, because there really seems to be no other way to explain how such a bizarre decision worked out so well for them....]

2)  Hasty/Hopeful/The Calculated Gamble

If you go in for a player after 1 or 2 good hauls, that will sometimes pay off for you. But again, it's very risky - you might be falling victim to the classically over-optimistic vice of chasing last week's points. 'Form' is not always accurately reflected by points returns; 'form' is often transient or inconsistent. You need to look for evidence of a general upturn in form that is likely to persist - and improve further - over a number of games. You can only find that kind of evidence by watching full games with close attention; stats are no substitute for that; and a highlights roundup is not good enough either - you need to have seen the whole game. 

3)  The 'Goldilocks Zone'

This is the 'ideal' time to make a transfer, the time when most 'smart' managers recognise that a player is probably now worth having. Now, as I just said above, because 'form' is not always reflected fully - or, sometimes, at all - in the FPL points returns, this 'sweet spot' might come when a player has started playing better, but has not yet produced any worthwhile points; or perhaps, they've only produced a few modest hauls, but (not yet) a really big one. And, although it might sometimes be possible to discern indications of a decisive upturn in performance from just one game, usually it takes a little longer to be confident of that. As I wrote in this essay at the end of last season attempting to define the concept of 'form': One good game might be a freak; two on the bounce is very encouraging, but it still might mean nothing; three.... is form.  Of course, with the benefit of hindsight, you can pinpoint the individual gameweek in which a player first began a streak of good points-scoring - even if those returns were at first quite modest and/or quite intermittent (and perhaps not really reflective of any strong positive shift in the player's or his team's performance at that point!). But as the season is unfolding, it is almost always impossible to recognise the exact moment of such a shift: it doesn't generally become visible until a little later - one, or two, or three gameweeks further on. And that is when you should be jumping on the player. Don't beat yourself up that you might have missed a few good hauls from him; be satisfied that you're getting good points from him now, points that a lot of rival managers are still missing out on.

4)  'Late to the party'

Often, you'll miss these 'early signs' that a player may be about to go on a tear. And that's forgivable - especially if, like me, you don't have good access to live TV coverage of the games. And the 'Goldilocks Zone' is very unforgiving, usually an extremely narrow window of opportunity: it's often just a single gameweek, and, even with the most generous dispensation, it can rarely be extended to more than two or three. However, after that point, it should be fairly obvious that this player is now a top pick, and if you continue to ignore him - well, then you're really not paying attention; or, worse, you're being stubborn (sticking to a previous selection in defiance of mounting evidence against it), or unreasonably prejudiced against this new option (perhaps he hurt you in the past: the old 'He always starts blanking as soon as I buy him' superstition - we've all fallen victim to that at one time or another). Often, such stubbornness grows out of being in thrall to the ideal of self-consistency, or what is sometimes called a 'committment bias'  - one of the most deep-rooted and damaging of our cognitive biases: we may have convinced ourselves that a player we'd picked ahead of the newly in-form player was always going to remain the superior prospect, or we may have convinced ourselves early on in this emerging trend of form that it was going to prove ephemeral - and nothing will now change our minds about this choice, no matter how much evidence is stacking up that it has now become wrong. But.... better LATE than NEVER. If you can overcome your pride, your stubbornness, your prejudice, and accept that a player is now so hot that you must have him - you might still be able to catch one or two more decent hauls from him before the streak dries up (as they all do, eventually).

In summary, then:...

The first type of transfer is very early: made before there has yet been any evidence of an improvement in a player's performance (or at least, not in his FPL points returns); it is hence a big, and probably unjustified, risk. Such transfers can occasionally pay off very well; but you should ensure that such a pick is based on definite evidence - if only in the form of positive background factors, rather than anything yet in the on-field displays - and sound reasoning.

The second type is also early, and possibly over-hasty: it is made on the basis of an improvement in performance, but perhaps only scant evidence of such, perhaps from just one or two games - which is rarely enough to make you fully confident of a sustained upturn in form. If you're allowing yourself to get over-excited about one or two good returns, you're most often not shrewdly anticipating the next big thing, but simply falling prey to a fatuous 'sheep pick'.

The third type is based on an astute assessment of early evidence of an upturn in performance: not just the FPL points returns, but the overall contribution of the player and the way he and his team are functioning together. Such evidence may occasionally be visible in a single game; but usually it takes two or three games to become persuasive.

The fourth type is at least slightly late - or maybe very late! - only recognising the turn in form some time after it has become evident, at least to the smarter and more perceptive FPL managers. (As a general rule-of-thumb, if The Scout - FPL's vapid, anonymous, in-house pundit - has just recommended a player,... you probably should have bought him at least one or two gameweeks back.)

To apply this template to the recent example of Phil Foden's extraordinary streak of goalscoring form....

If you'd gone in for him in Gameweek 13, that was definitely a Type 1 decision. And it must have been based on having a crystal ball, because there really was no indication that he was suddenly about to come good so strongly, after a long run of 'blanks', and often pretty anonymous performances (he'd only produced 20 points from the previous 8 games, with just a solitary assist!). A fixture against Leeds was not sufficient reason to start fancying a player who'd shown almost no indication of suddenly becoming a major FPL points contributor again over the previous two months. (Neither was a moderately promising fixture-run immediately thereafter. And it was only moderately promising: Fulham away is not a particularly 'easy' fixture; Palace away certainly isn't.)

If you'd bought him for Gameweek 14, that would have been a Type 2 decision. Was there really any evidence in the Leeds game of a decisive change in mentality, confidence, style of play (or in team tactics, that might give him more scoring opportunities...)? Not really. And that good performance was only against Leeds, who are struggling at the bottom of the table. Moreover, there were two further Premier League games within the coming week, and we might reasonably have expected that Foden would get short minutes in these games - or perhaps be rested completely for one of them - and/or might have been inhibited a little by fatigue. I would say, getting Foden in GW14 was essentially just chasing last week's points; there was not yet any convincing rationale for buying him.

The interesting question is whether, after two outstanding displays in quick succession, it would have a Type 3 decision to get Foden in Gameweek 15. I would say NO. Foden had just played twice within a few days; and there was a crucial Champions League tie away to Real Madrid coming up the following midweek - and he did indeed get short minutes in this next match. Also, Sunderland were much the best of the three EPL teams he'd faced in this remarkable week - although they produced an uncommonly poor performance on this occasion, and, despite playing only just over an hour, Foden again managed to come up with a goal.

Acquiring him didn't really become a Type 3 decision rather than Type 2 until Gameweek 16; and arguably perhaps not even then. In the wake of the Real Madrid game, there was again a risk that Foden might be rested or subbed off early. And Crystal Palace are one of the best teams City have faced this season, the strongest defence in the league (after Arsenal's); especially away from home, a City win could not have been confidently predicted. Moreover, it is statistically extremely improbable that any player will achieve 4 double-digit hauls in a row (has it ever been done before??) - and while that wouldn't argue against acquiring Foden for this game, it should counsel that it was unreasonable to expect another very big points return from him. And yet - astoundingly - he did produce yet another haul! But that was a once-in-a-blue-moon freak; and if you bet on that, you were lucky rather than brilliant.

This might be a rare case where there was in fact no 'Goldilocks Zone' for acquiring Foden. There have been good reasons to doubt if he would be able to extend this scoring streak in each gameweek that it has continued; and the calculus on this was greatly complicated by the December fixture congestion and the huge importance of City's looming Champions League clash. In each of those four - closely packed-together - gameweeks, there were reasonable arguments for remaining hesitant about acquiring Foden: It was becoming increasingly unlikely in each match that he would produce another successive big haul, and increasingly likely that Pep might rest him. 

Moreover, Doku had been a very attractive pick from City for a while, and Cherki was just becoming so - before Foden suddenly caught fire. And if you already had one of these (or perhaps Nico O'Reilly; albeit that he's in a different position category, that pick would still be eating into your club quota from City), it would not have been unreasonable to hang on to them, in preference to Foden - whose newfound burst of form might well have quickly fizzled out again. And even if you had fancied bringing Foden it, it was still a somewhat speculative pick - so early in the emerging hot spell, after such a long barren streak, after such a disappointing season last year - and it might not have been a leading transfer priority; especially as the mounting toll of injuries and suspensions in the bleak midwinter was using up most of the available stock of Free Transfers for the majority of FPL managers (even with the bonanza of extra transfers we were just given in in GW16).

Perhaps, perhaps... transferring in Foden has jumped straight from a Type 2 to a Type 4 decision!

However, I would suggest that acquiring Foden now, for Gameweek 17, is probably the Type 3, 'Goldilocks Zone' moment. 

Yes, paradoxical as it may seem, I think he's only become a really unassailably strong pick now - when his super-hot streak is (almost certainly?) over. You will have missed 55 points from 4 games in the last fortnight; but a fortnight ago, that was completely unforeseeable; and even as the fortnight played out, it remained hugely improbable. You shouldn't suffer any self-recrimination for missing out on the benefits of such a bizarre freak performance.

But now, we have seen clear evidence that Foden - and his team - are playing extraordinarily well; and this does seem to be a sustained shift in performance. Moreover, they're in a pretty soft run of fixtures still, with only the New Year game against Chelsea offering a significantly formidable opponent in the next month or so (and at least that one's at home). At this point, it is definitely looking foolhardy to resist buying him any longer.

With the benefit of hindsight, we would all realise that Gameweek 13 was the ideal time to have brought him in. With the benefit of a crystal ball, we could have realised that he would become 'essential' in Gameweek 14.... or 15,... or 16. But we do not enjoy those two benefits. And without them, it was reasonable enough to hold off buying Foden until now. Most of the FPL managers I consider 'smart' have not bought him yet (the few that do have him, I think had punted on him earlier in the season - and had probably only stuck with him thus far because they'd had other higher-priority problems to address with their transfers); but I anticpate that the majority of them will be going in for him now.

Alas, it is now statistically very unlikely that Foden will achieve yet another massive points-haul in the next game - even against defensively flakey West Ham. And you can be pretty damn sure that he won't rack up another 50-odd points over the next 4 or 5 games. But he is in sensational form, and has some inviting fixtures coming up.

If you already got him in the last few gameweeks, good for you; but you were taking a gamble on him, and were fortunate that it paid off so extraordinarily well. Now, though, it really is looking as if all of us should join you.