To see less ads Register or Login ----- Daily Fantasy Sports games 18+

4-4-2 and Team Structure

A Fantasy Football forum for news on fantasy football games run by the Premierleague (FPL).
User avatar
Ruth_NZ
Grumpy Old Gorilla
Posts: 9156
Joined: 25 May 2015, 22:46

4-4-2 and Team Structure

Post by Ruth_NZ »

I'd like to initiate a discussion about team structure and will try to keep this as concise as I can although it will still be lengthy.

I guess that most of us will be familiar with the traditional 3-4-3 structure. It has many variations but it typically features a cheap bench midfielder and 4 cheaper (usually 4.5m or less) defenders rotating around 1 premium defender. With a cheap 2nd GK that means usually not much more than 17.5m or so on the bench meaning that the maximum budget is on the pitch.

I once asked Triggerlips why he was so sure that 3-4-3 was the optimum structure. He said that it was "best by test". Which is fair enough, apart from the fact that other systems haven't been equivalently tested. Most experienced managers structure their squad for 3-4-3. That could mean that 3-4-3 is indeed the best solution; alternatively it could mean that it is generally assumed to be best because that's what nearly all the experienced managers do. Either way, fewer competent managers have ever used other systems (even 3-5-2) on a long-term basis. They haven't been equivalently tested.

The benefit of 3-4-3 is that the highest-scoring players in FPL are usually midfielders and strikers. Currently the top defender (Cahill) has 58 points, the top midfielder (Hazard) 86 points and the top striker (Costa) 82 points. So it stands to reason that if the attacking players score bigger then you want 7 of them, right? That means 3-4-3 or 3-5-2. But these points totals are misleading. Costa costs 10.5m, Hazard 10.2m, Cahill only 6.1m. That has to be taken into account.

I don't entirely agree with the way FPL calculates "value" because it's too simplistic (it doesn't factor in VORP theory for example). But let's use it as a reference point. So far this season there are 12 defenders who have delivered a return of 8 points per £1m cost or better. 5 of those 12 (41%) are premium defenders - Cahill, Azpi, Walker, Koscielny, Blind. There are 16 midfielders who have delivered that value but only 3 of them (19%) are premiums - Hazard, Coutinho, Firmino. And there is only one striker that has achieved that value level - Defoe (though Costa is close on 7.8 points per £1m).

Sometimes it is said that what FPL managers are after is points. Well that's self-evident but also simplistic. Because who would have expected Vokes to be the 3rd-highest forward on value and the 8th highest on points after 12 games? He has only 15 points less than Aguero! We aren't ever playing the game with the benefit of hindsight. So I would argue that what we are really after is predictable points most of the time. Players we can rely on.

This is where the value tables start to tell a different story. We can forget the forwards, they aren't delivering value and they seldom do in any season. Among the midfielders, well you could have predicted that Hazard, Coutinho, Firmino had a good chance of doing well. But most of the others? Not a chance in hell. Capoue maybe but most had him as a 5th mid and missed most or all of his points anyway. Defenders - another story. Much more predictable performance from premiums (the ones you might have a fair chance of selecting) with a lot more (VVD, Fonte, Bellerin, Vertonghen, Alonso) in the 7-8 points per £1m range.

So, to summarise so far. We have a better chance of predicting value in terms of points per £1m with defenders than with attackers. And amongst the attackers, the midfielders are more predictable for value than forwards.

Now, on what basis do we select squads? I reckon there are 3 types of player we want: (a) Players we select for pure points (these will usually include our preferred captain options); (b) Players we select for value (these need primarily to deliver a good points per £1m ratio); (c) Players we select to provide a defensive rotation and/or for the bench (the main requirement being only that they start for their clubs and sometimes not even that).

I would therefore suggest that the optimum structure for a squad will vary according to player pricing and performance in any season. Last season value scarcely mattered. TV scarcely mattered. Mahrez and Vardy, Payet and Alli were all reliable, predictable points-scorers. And they were cheap. 2 or 3 of them and you could have pretty much whoever else you wanted. Meanwhile it was hard to find many premium players that were performing. Chelsea and United assets were pretty much dead in the water. The top 4 teams were Leicester, Spurs, City, Arsenal. Well, an attack featuring their best players - Mahrez, Vardy, Alli, Kane, Aguero, Sanchez, KDB - was pretty easily done, those 7 could be had for around £60-62m (depending on when you bought them). If you needed to spend a bit less you could put Payet in for KDB and save another £2m or so. Easy peasy.

Try that with the current top 4 teams. You'd want Hazard, Costa, Coutinho, Firmino, Aguero, KDB, Sanchez. That would be around £72m or so. Even if you put Kane in for Aguero it's still around £70m. It can't be done. We don't have the budget.

This is where 4-4-2 comes into its own. In the current situation you can compromise with a couple of cheaper attackers - a Benteke or a Redmond - but you lose predictability. You get lucky or you don't. And at the same time you will be paring the defensive investment down as hard as you can. I hate to think what the PPG we have achieved from our defenders is this season. I doubt it bears scrutiny. [In actual fact I believe that delivering a better defender PPG than others is the most surefire way to gain ground, especially in a season like this. I have looked at top-10 teams in previous seasons and one thing that has marked them out has often been their high defender PPG.]

Or instead of compromising you can decide to have only 6 attackers. Do that and you can have 2 expensive (predictable) forwards and 4 expensive midfielders much more easily. And you can use some of the cash saved on premium defenders who are much more likely to be secure on value. 6 attackers mainly for predictable points. 4 defenders mainly for predictable value. That's the killer advantage of 4-4-2. By reducing your number of attackers to 6 you can actually shoehorn more of the predictable high-scorers into your team.

And that's why I would say that as things stand this season, 4-4-2 is the ideal structure. Not in every season. Last season 3-4-3 was certainly better. But last season's price/performance landscape was different to what we have now.

As a last example, here's a sample 4-4-2 wildcard squad that I could afford right now (and my TV isn't huge):

Pickford Mannone
Clyne Alderweireld Evans/Mee Alonso
Firmino Coutinho Özil Hazard Fletcher
Costa Kane Anichebe


It's only a sample and I'm not really talking about which are the best players to have, some may prefer KDB and/or Aguero in there. But in terms of shape I think that squad ticks all the boxes. To transition towards something like that from 3-4-3 takes some time unless you want to take a pile of hits. But I think that will probably be my general direction of travel.

ricardo68
FISOhead
Posts: 644
Joined: 13 Jul 2010, 04:28

Re: 4-4-2 and Team Structure

Post by ricardo68 »

Nice points. Where's Bulgarche

Shevchenko
Wideboy
Posts: 83
Joined: 19 Oct 2016, 23:23

Re: 4-4-2 and Team Structure

Post by Shevchenko »

To be honest ive already thought about this and in the middle of transition between 3-4-3 to 4-4-2. My squad look like this :

Heaton (Pickford)
Alonso Fuchs Holebas (Kingsley) (Lenihan)
KdB Hazard Coutinho Firminho (Fletcher)
Kane Costa (Anichebe)

0.2 ITB

Need about 0.6 M to replace Kingsley and Lenihan to Evans/Mee/Friend/Lowton/any 4.5 that covers well with Holebas. Thinking about switching one of the liverpool midfield to Walcott to cover Arsenal fixtures, so then maybe i can switch Fuchs to Vertonghen cause Leicester look like a trainwreck while Spurs can focus on PL after eliminated from CL.

Unfortunately ive just found out about FPL at GW6, plus ive got so many wrong players at my first selection, so my TV pretty low and need A LOT of tinkering to get here. But pretty much excited when i see my team start to get into shape and hopefully give me superb boost in the next week!!!

User avatar
Ruth_NZ
Grumpy Old Gorilla
Posts: 9156
Joined: 25 May 2015, 22:46

Re: 4-4-2 and Team Structure

Post by Ruth_NZ »

Shevchenko wrote:To be honest I've already thought about this and in the middle of transition between 3-4-3 to 4-4-2.

Thinking about switching one of the Liverpool midfield to Walcott to cover Arsenal fixtures, so then maybe i can switch Fuchs to Vertonghen cause Leicester look like a trainwreck while Spurs can focus on PL after eliminated from CL.
OK, first thing - there are a lot of good FPL managers on FISO and not many will agree with me (I don't think so anyway). Just be aware of that.

If that were my team I'd downgrade KDB to Özil. KDB will get rotated at some stage in my opinion (that's another one that others will likely disagree with). With 1.4m banked you could finish the defence. Lowton Evans Holebas Vertonghen Alonso should be doable. Basically you'd play any 2 from Lowton Evans Holebas and Anichebe in any given week so 4-4-2 with an occasional 3-4-3. That's how I'd do it anyway.

I think you will do well with that team.

User avatar
blahblah
FISO Viscount
Posts: 108503
Joined: 13 Oct 2005, 18:46
Location: .. he thinks that he knows something which he doesn't, whereas I am quite concious of my ignorance.

Re: 4-4-2 and Team Structure

Post by blahblah »

I'll read it through properly when I have the time: but why stop at 4-4-2, what about 4-5-1?

For me the Big Clubs have to be performing defensively ie Jose's first spell at Chelsea...

User avatar
Ruth_NZ
Grumpy Old Gorilla
Posts: 9156
Joined: 25 May 2015, 22:46

Re: 4-4-2 and Team Structure

Post by Ruth_NZ »

blahblah wrote:I'll read it through properly when I have the time: but why stop at 4-4-2, what about 4-5-1?
Because there's only 1 Anichebe. You need decent playing strikers around 4.5m to make this kind of structure work. The emergence of Anichebe is what makes 4-4-2 seriously viable now. Diomande might make 4-5-1 viable too when his suspension is over but he's really a winger and is he likely to be a nailed starter? Anichebe on my bench over Xmas - fine. He has big aerial threat and could feasibly score against anyone. Diomande? Not so sure.

In addition there's captaincy. 2 premium strikers help to provide reliable captaincy options. But there's nothing wrong with 4-5-1 in principle if the right players are available at the right prices. It all depends on that.

User avatar
Aldershot Rejects
Dumbledore
Posts: 9586
Joined: 03 Aug 2011, 17:15
Location: Kent
FS Record: 5th Metro (2010-11); 146 - Sky (2015-16); 218 - Sky (2014-15); 386 - Sky (2020-21); 636 - FPL (2017-18); last 16 Sky Cup (2018-19)

Re: 4-4-2 and Team Structure

Post by Aldershot Rejects »

Only had time for a quick skim but I've been looking at 4-3-3 for most of the reasons you suggest. Think I prefer this structure because of Defoe who is the most reliable sub 7.5 million player in my opinion. I'm also finding this structure is much easier if you go without Kun.

I'll have a longer look later.

User avatar
Tall Paul
Dumbledore
Posts: 7517
Joined: 27 Aug 2008, 12:57

Re: 4-4-2 and Team Structure

Post by Tall Paul »

A friend asked me to run his team for him after GW 3 this season. I wildcarded in GW 4 and went with 4-4-2. His team has outperformed my 3-4-3 since then, only by 10 points, but it seems to be a viable strategy.

User avatar
fred1266
Dumbledore
Posts: 5356
Joined: 16 Mar 2012, 22:56
Location: Trinidad
FS Record: EFCL Fantasy League & Jurassic League Winner 11/12, EFCL Fantasy League Winner 13/14. EFCL, FISO 5AS The Kangals, Jurassic League, 3DM PRINTERS TT & Trini Gunners Winner 14/15
Contact:

Re: 4-4-2 and Team Structure

Post by fred1266 »

ricardo68 wrote:Nice points. Where's Bulgarche
well we can rule out any 5 at the back cause he has never been me :D

User avatar
Archy
Grumpy Old Man
Posts: 3210
Joined: 22 Oct 2011, 10:09
FS Record: Maybe all is not lost after all

Re: 4-4-2 and Team Structure

Post by Archy »

Good post Ruth
That's the killer advantage of 4-4-2. By reducing your number of attackers to 6 you can actually shoehorn more of the predictable high-scorers into your team.
I'm not sure if I necessarily agree with this statement however.

When you talk about 'predictable' hig-scorers, are you talking about the likes of Cahill and Azpilicueta? I don't think it was ever really predictable that these guys would have kept 6 clean sheets in a row so I wouldn't consider it a given that these guys should be shoe-horned in, even now. Even with this (probably unsustainable) run, they have only beat the likes of Cook, Daniels and McCauley by crica 10 points, despite being at least £1m more expensive. The strong performing Arsenal defenders are at least £1.5m more expensive than the cheaper alternatives.

The question really then is whether the incremental cost is worth it, and whether the predictably high-scoring defenders are really that predictable in prcatice (weren't we all piling in to Man U defenders at the start of the season, plus Stones from Man City?!).

Personally I'm using 3-5-2 by the way as I think the best value is siting in midfield this season.

User avatar
baganboy
Comfortably Dumb(ledore)
Posts: 5874
Joined: 05 Aug 2008, 06:59
FS Record: 2011/12 - 212. 2019/20 - 222.
Altogether 6 top 10Ks. 8 top 20Ks. 9 top 50Ks.

Re: 4-4-2 and Team Structure

Post by baganboy »

I have been extolling the virtues of 4-3-3 (which is basically the same thing as a 4-4-2) all season. At the risk of gloating, they have served me rather well too.
I personally needed a lot of math to be convinced out of the template 3-4-3 (I must still have the numbers somewhere), but I think one sentence suffices: There are way too many 5.0M top-team CRDs to not consider a 4-man defence.

PS: I dabbled with a 5-Man defence for a while, but I don't think there are enough big-scoring big-money players for a 5-2-3/5-3-2. I still might dabble in it, if Aguero, Kane, Zlatan, KDB and Sanchez all get into a hot spell. The only way you can get all the big-money(10M+)-big-points hit form at the same time, is loading up on the defenders. In the current season, I would always rather pick an arbitrary 5.0M defender (Darmian) over an arbitrary 5.0M midfielder / forward (Stuani).

User avatar
Ruth_NZ
Grumpy Old Gorilla
Posts: 9156
Joined: 25 May 2015, 22:46

Re: 4-4-2 and Team Structure

Post by Ruth_NZ »

Aldershot Rejects wrote:Only had time for a quick skim but I've been looking at 4-3-3 for most of the reasons you suggest. Think I prefer this structure because of Defoe who is the most reliable sub 7.5 million player in my opinion. I'll have a longer look later.
Yeah, 4-3-3 is a very similar concept. Basically you are talking about Sigurdsson and Anichebe (4-4-2) against Defoe and Fletcher (4-3-3) as a rough comparison. Either look fine. Defoe gives you Deeney and Austin as get-out options in the case of injury. Siggy gives you players like Willian, Cazorla, Mata. I just think I prefer the options at 7.0 to 7.5 among the midfielders. But somewhere or other you will probably still have that one attacking slot at around that level even if the other 5 are big-hitter premiums.
Archy wrote:Good post Ruth.
Ruth_NZ wrote:That's the killer advantage of 4-4-2. By reducing your number of attackers to 6 you can actually shoehorn more of the predictable high-scorers into your team.
I'm not sure if I necessarily agree with this statement however. When you talk about 'predictable' high-scorers, are you talking about the likes of Cahill and Azpilicueta?
No, I am talking about the attackers. The point with the defenders is that they offer more predictability on value. I have watched this for 3 seasons and it is pretty much a constant.

You have to select the right players, though. That remains the case whatever structure you use. I am simply of the opinion that 4-4-2 gives you a somewhat better chance of doing so in a season where budget is under hard pressure.
baganboy wrote:I personally needed a lot of math to be convinced out of the template 3-4-3 (I must still have the numbers somewhere). The only way you can get all the big-money(10M+)-big-points [players if they] hit form at the same time is by loading up on the defenders. In the current season, I would always rather pick an arbitrary 5.0M defender (Darmian) over an arbitrary 5.0M midfielder / forward (Stuani).
Yep, that's the basic concept. I would rather have Alonso or Alderweireld than Bolasie or Chadli. More effective on transfers too because you would likely keep them longer, the cheaper mids are much more prone to fixture runs.

I guess the difference between us is that I am looking for the predictability of an Alonso or an Alderweireld rather than more hit-and-miss selections like Stones or Williams. It's all about predictable performance and predictable value for me. But you could certainly go a bit cheaper with the likes of VVD, Valencia, Clyne (and maybe Zouma) at around 5.5m rather than the super-premium defenders at 6.0+.

It's always a matter of permutations. But your underlying structure often defines the permutations available to you. Which is why I thought it worth discussing.

User avatar
baganboy
Comfortably Dumb(ledore)
Posts: 5874
Joined: 05 Aug 2008, 06:59
FS Record: 2011/12 - 212. 2019/20 - 222.
Altogether 6 top 10Ks. 8 top 20Ks. 9 top 50Ks.

Re: 4-4-2 and Team Structure

Post by baganboy »

Ruth_NZ wrote: I guess the difference between us is that I am looking for the predictability of an Alonso or an Alderweireld rather than more hit-and-miss selections like Stones or Williams. It's all about predictable performance and predictable value for me. But you could certainly go a bit cheaper with the likes of VVD, Valencia, Clyne (and maybe Zouma) at around 5.5m rather than the super-premium defenders at 6.0+.
True that. My premiums are costlier premiums (Aguero et al), the defenders will have to be a bit cheaper...
Ruth_NZ wrote: It's always a matter of permutations. But your underlying structure often defines the permutations available to you. Which is why I thought it worth discussing.
Yes. Team structure is always worth discussing.

User avatar
redchemist
Grumpy Old Man
Posts: 1743
Joined: 29 Nov 2013, 23:19
Location: A massive Brinkhoff

Re: 4-4-2 and Team Structure

Post by redchemist »

This season I am very fond of 3-5-2. Mostly because of the Liverpool mf resurgence.
Hell, with Sanchez, Hazard, Gylfi, and other oop options scoring 5 points a goal (rather than a fowards 4); I would play 3-6-1 if you'd let me.

The Dazzler
Grumpy Old Man
Posts: 1248
Joined: 03 Dec 2008, 20:26
FS Record: 9th overall in FPL 2005/06, 50th 2010/11, 288th 2014/15

Re: 4-4-2 and Team Structure

Post by The Dazzler »

I saw you say last year that other systems haven't been equivalently tested. This is not correct and I can't understand how you can possibly think it is.
I've been playing this game for 12(?) years and of course they have all been tested. You think over a decade of millions of people playing, every system hasn't been tried? They have. There is always someone who comes on here every year saying, "isn't there more value in defenders? Shouldn't we be looking at spending more in defence?" And yet, 3-4-3 has always prevailed.
Although your initial premise is clearly false, you may have stumbled onto something. Not because the other formations haven't been tested but because the pricing structure has changed.
I sent a pm to TheSilkworm preseason on this subject;
The Dazzler wrote: I do think that FPL's constant tinkering may have caused a paradigm shift. Up to 2 years ago, you could buy playing £4m rotating keepers. And the top 4 keepers were £6.5m. Buying them was wrong and hurt the casuals who did so. Now the cheaper keepers are more expensive and the expensive ones are cheaper.
You could also buy playing £4m defenders. Now they are gone. Cheaper ones are more expensive and expensive ones are now cheaper (plenty @ £5.5m). Everything is moving to the middle. This creates a situation where casual's mistakes in defence are punished less.
But it also may have another effect. Up until now, every single season I have played the game, 3-4-3 was undisputed king. The price tinkering may have shifted the balance. Perhaps for the first time ever, 4-4-2 might be best or 3-5-2 or 4-3-3 or even 5-2-3! Or maybe two or more different formations have essentially the same potential. We'll see.
This price change shift towards the middle might(?) have been the reason Ville decided so early last season to go to a Strong 8 but that's veering onto a slighter different subject.
Last edited by The Dazzler on 25 Nov 2016, 18:09, edited 1 time in total.

User avatar
Ruth_NZ
Grumpy Old Gorilla
Posts: 9156
Joined: 25 May 2015, 22:46

Re: 4-4-2 and Team Structure

Post by Ruth_NZ »

Dazzler, I find that pretty condescending. I don't really get why you take that attitude but never mind.

I think "most experienced managers structure their squad for 3-4-3. That could mean that 3-4-3 is indeed the best solution; alternatively it could mean that it is generally assumed to be best because that's what nearly all the experienced managers do" from my OP is accurate. 4-4-2 hasn't been equivalently tested because it hasn't been used anywhere near as much. 3-4-3 is the default and 3-5-2 is the sometimes-used alternative. A properly structured 4-4-2 is relatively uncommon but I never said it hadn't been tested or tried at all. I have tried it myself (the season before last) with good effect.

All I was trying to say is that the concept of 3-4-3 being the best system isn't set in stone.

As for "inadvertently stumbling" on the idea of a different structure suiting a different price/performance landscape, I don't really see what's inadvertent about it seeing that's what I was assessing in the first place.

It's no big deal. But it is a bit bewildering to always get responses like this from you, I have to say.

User avatar
baganboy
Comfortably Dumb(ledore)
Posts: 5874
Joined: 05 Aug 2008, 06:59
FS Record: 2011/12 - 212. 2019/20 - 222.
Altogether 6 top 10Ks. 8 top 20Ks. 9 top 50Ks.

Re: 4-4-2 and Team Structure

Post by baganboy »

Triggerlips' assertion that it's 'best by test', might very well be partially right, as I do know that other systems have been heavily tested. The 'best' part, is doubtful, IMHO. %**

Here at FISO at least, as a relative long-timer myself (like The Dazzler, though not quite as good an FPL player as he) I agree with him that many of the other formations have been tried and tested to death, by many players along the way. indeed, during my first season at FISO, 4 5 1 (with Henry - then just departed to Barcelona - as the permanent captain) was a formation very much in vogue. Perhaps searching old postings from guys like Wyld et al, right here at FISO would reveal as much. Those guys did a lot of the earliest FPL related thinking, including the price change algorithm --- many of these ideas form the bedrock of current FPL thinking. Tongue-in-cheek, I'd say that just like there was football before the Premier League, there was FPL before Fantasyfootyscout.

In my opinion, generally, when a third of the season has passed (i.e. about now), there is a cheap monster striker / midfielder that appears, to upset the applecart and make only the 3-4-3 viable... if you can afford 6 heavy hitters + cheap-monster (who is effectively as good as the big guns, in other words, Riyad Mahrez. Amr Zaki :D ), you will do it of course - in terms of sheer points irrespective of value, mids and strikers rock. There has been no true cheap monster mid/str this season. Cheap monster defenders, now, are really rare, simply because defenders are part of a pack and cleanies are afforded to all, and you would not find one defender from a weak team stand out from the others *&*. Inference: Therefore 3-4-3. IMHO, 3-4-3 is an effect, not a cause.
Also, (and this is my personal observation, backed up by little data) there has been a point when FPL made the prices go in such a manner that for midfielders and strikers, only the costly (~9M+) and the cheap (<6M) were feasible options *** + CRDs of the top teams were jacked up from 5M to 5.5 and now 6M. My emerging hypothesis is that: This is also the time that FFS became the super-popular one-stop shop for FPL, thus 3-4-3, something which is an effect (a result of structural changes to the game/pricing), because accepted as the ONLY way to play the game.
I remember seasons when defenders (especially CRDs) were cheaper, 4-defender teams were common at the top, and even 5-defender teams were not entirely uncommon. I also remember the post-2010 days, when playing anything apart from a 3-4-3 would mean that one is a noob. This season's pricing, to me seemed a little bit similar to when I started playing this game. 3-4-3 is still a good formation to employ, but not the only path.

However, formation talk at the start of the season is always reasonable - especially because of seasons like this current one, when CRDs from top teams are at 5M, no cheap-monster midfielder in sight, and many (indeed, most) of the big guns firing. And this season is especially fun for old timers like us because it is distinctly un-template.
Personally: for better or for worse (and perhaps due to my early influence), I have always preferred the 4-3-3 over the 3-4-3. I played my first FISO season with 4-4-2. And I tend to gravitate towards the 4-defender format often during every season I played.
PS: Talking of 5.0 CRDs, I played a half of a season with Wes Brown, 5.0M at right back for a defensively dominant Man United - Rio, Vidic, Evra, Van Der Sar, Howard Webb, and a 5.0M CRD.. yes that's what I am talking about


%** though I have never interacted with him - I have never really dared to venture out of my nice little cocoon here at FISO.
*&* It has happened though - really old timers will remember the 'How are you doing this Laursen' days.
*** Gylfi ***** Sigurdsson has always been the exception, and thus is his reason for surprising FPL popularity.

IceCreamFirm
Red & Blue Braces
Posts: 397
Joined: 06 Aug 2015, 12:59

Re: 4-4-2 and Team Structure

Post by IceCreamFirm »

Looking at having 2 up front too.
Not happy with Lukaku and Benteke and can't see anyone I want below 10m.
Last edited by IceCreamFirm on 27 Nov 2016, 10:56, edited 1 time in total.

User avatar
Stemania
FISO Jedi Knight
Posts: 20448
Joined: 27 Aug 2006, 11:54
Location: On the Iron Throne of xG, the seat of The Crown Prince of the Stat Perverts. Or if not, in the STC!
FS Record: Best: TFF 321st. FPL 129th. FFS Career HoF Peak 2nd (Live 1st). Ability since lost.

Re: 4-4-2 and Team Structure

Post by Stemania »

Historically, the FPL pricing structure has long favoured 343 (which is why it's pretty much been a standard for everyone for years). FPL Towers have been really improving their pricing the last few years though, especially with both base price and premium defenders/keepers. The cost of 'good' third strikers on the cheap has also been 'written out' by FPL (there were no 5.5m penalty taking Callum Wilsons this year) impinging further on 343's 'obviousness'.

This is the first season I can remember in a long time when other formations have sometimes looked viable (for reasons other than the anomalous "there's Harry Kane at 5m starting as lone striker for Spurs every week"-type ones). I personally still think the pricing favours 343 in the long term, but for spells of the year I'm open to other formations being very effective depending on personal circumstance. If it were to happen soon to my lot I think I'd be more likely to head in the 352 direction than the 442 one mind you. :D

The Dazzler
Grumpy Old Man
Posts: 1248
Joined: 03 Dec 2008, 20:26
FS Record: 9th overall in FPL 2005/06, 50th 2010/11, 288th 2014/15

Re: 4-4-2 and Team Structure

Post by The Dazzler »

Ruth_NZ wrote:Dazzler, I find that pretty condescending. I don't really get why you take that attitude but never mind.
Ok, I have taken out the "somehow inadvertently" part if that's what is causing the problem. I don't think my tone was any stronger than the way you normally address me or anyone else but happy to modify it.
Ruth_NZ wrote:I think "most experienced managers structure their squad for 3-4-3. That could mean that 3-4-3 is indeed the best solution; alternatively it could mean that it is generally assumed to be best because that's what nearly all the experienced managers do" from my OP is accurate. 4-4-2 hasn't been equivalently tested because it hasn't been used anywhere near as much. 3-4-3 is the default and 3-5-2 is the sometimes-used alternative. A properly structured 4-4-2 is relatively uncommon but I never said it hadn't been tested or tried at all. I have tried it myself (the season before last) with good effect.

All I was trying to say is that the concept of 3-4-3 being the best system isn't set in stone.
We have over a decade of play, with maybe 20+ million teams and maybe 100 million GWs. All formations have been tested. Extensively. And 3-4-3 has prevailed repeatedly. Best by test.
You have repeatedly said they have not been tested. This is incorrect.
Ruth_NZ wrote:As for "inadvertently stumbling" on the idea of a different structure suiting a different price/performance landscape, I don't really see what's inadvertent about it seeing that's what I was assessing in the first place.

It's no big deal. But it is a bit bewildering to always get responses like this from you, I have to say.
It's inadvertent because your original hypothesis (that other formations haven't been sufficiently tested) is false. It is the changed pricing points that is causing the situation whereby the 4-4-2 formation (and others) may now (and only now) be viable.
All I'm doing is pointing out the flaw in your methodology. You may be right, it's just for the wrong reasons. It's a little like the 'What the Bookies, Fixtures and Form Say - Captains pick' article on FFS, or as I like to call it, "The Lucky Dice - Captains pick" article. The methodology is so flawed it's unbelievable. Nobody wants to hear that as it's been widely 'successful'. The Emperor has no clothes.
Anyway, now you know why the conditions might indeed be in place for an alternative formation to the 3-4-3, so it can only aid you in your research.

So with that out of the way, we can move onto the actual content of your post? I might have a few issues there too.

User avatar
Beerfuelledman
FISO Knight
Posts: 13220
Joined: 13 Oct 2005, 18:26
Location: In Norn Iron
FS Record: FISO 17/18 FPL Cash Draft League Winner

Re: 4-4-2 and Team Structure

Post by Beerfuelledman »

Ruth_NZ wrote: As a last example, here's a sample 4-4-2 wildcard squad that I could afford right now (and my TV isn't huge):

Pickford Mannone
Clyne Alderweireld Evans/Mee Alonso
Firmino Coutinho Özil Hazard Fletcher
Costa Kane Anichebe


It's only a sample and I'm not really talking about which are the best players to have, some may prefer KDB and/or Aguero in there. But in terms of shape I think that squad ticks all the boxes. To transition towards something like that from 3-4-3 takes some time unless you want to take a pile of hits. But I think that will probably be my general direction of travel.
I like it, but I think I prefer a 352 variant as long as Allen is in the No 10 role. If Allen plays deep, the 442 is more optimal IMHO.

Pickford Mannone
Alderweireld (Evans/Mee/McAuley )Alonso
Firmino Coutinho Özil Hazard Allen
Costa Kane Anichebe

User avatar
baganboy
Comfortably Dumb(ledore)
Posts: 5874
Joined: 05 Aug 2008, 06:59
FS Record: 2011/12 - 212. 2019/20 - 222.
Altogether 6 top 10Ks. 8 top 20Ks. 9 top 50Ks.

Re: 4-4-2 and Team Structure

Post by baganboy »

The Dazzler wrote: We have over a decade of play, with maybe 20+ million teams and maybe 100 million GWs. All formations have been tested. Extensively. And 3-4-3 has prevailed repeatedly. Best by test.
Hey TD, I don't agree with this BTW. As in, I do, but 3-4-3 was best by test within a caveat of a window of time. around 2010 to 2015. Well, maybe 2015-16 too, but how would I know?
I remember the 09-10 season well (my first really good one), and I remember having played 4-3-3 all season - and remember quite a few in the top playing 4 at the back too. The previous was the Laursen season, so 4 in defence had to be de jure. Correct me if I am wrong, but Laursen came in at 4M? Of course you would remember it too, you were already an elite player then.
Those where the seasons when Joe Hart at Birmingham and Mark Schwarzer at Fulham (?), even that Isecssen or something guy at Stoke, were first team GKs at 4M apiece, and rock solid defences. You could build a team with 5 X 4M defenders, and many did.
In the Lampard seasons, everyone would have Lampard, PADregas and Ronaldo in their teams, even 3-5-2 or 4-5-1 were options too (with the Mikael Forssels of the world solid first team 4.5M enabler strikers)
My argument is that FPL killed all that, with increasing the base defender + goalkeeper prices, with the substitute captain rule, penalties for conceding a penalty, and all those silly s**t reasons. I think they are balancing it somewhat now.

This is like two old men fighting over how the grand old times were differently grand :D :D :D

User avatar
Ruth_NZ
Grumpy Old Gorilla
Posts: 9156
Joined: 25 May 2015, 22:46

Re: 4-4-2 and Team Structure

Post by Ruth_NZ »

The Dazzler wrote:It's inadvertent because your original hypothesis (that other formations haven't been sufficiently tested) is false. It is the changed pricing points that is causing the situation whereby the 4-4-2 formation (and others) may now (and only now) be viable. All I'm doing is pointing out the flaw in your methodology. You may be right, it's just for the wrong reasons. It's a little like the 'What the Bookies, Fixtures and Form Say - Captains Pick' article on FFS, or as I like to call it, "The Lucky Dice - Captains Pick" article. The methodology is so flawed it's unbelievable.
There is no flaw in my methodology because I'm not trying to 'prove' anything. And the 'hypothesis' you refer to is nothing of the sort. It was a small commentary designed to say that it isn't a given that 3-4-3 is automatically the best structure. From what you say, that isn't news to you. Great. So we agree.

I don't think 4-4-2 becomes viable or not because of overall price points. What makes a structure efficaceous is its ability to deal with the specific price/performance landscape at play at any time. If there were one or two players priced like Mahrez, Vardy, Alli, Payet were last season and performing at similar levels you caould forget all ideas of 4-4-2. But this season they aren't there. Or if they are they are yet to become visible.

At the same time, 4-4-2 really needs a decent, playing striker at close to 4.5m to be at its most effective. I had my eyes on Barrow but 5.0 is a bit pricey for a 1st sub. The emergence of Anichebe is therefore another jigsaw piece appearing for a good 4-4-2.

As for the FFS article, I totally agree with you and have said so. In fact I was banned from FFS last season partly because I refused to back down about it. Now I just ignore it.
The Dazzler wrote:Anyway, now you know why the conditions might indeed be in place for an alternative formation to the 3-4-3, so it can only aid you in your research. So with that out of the way, we can move onto the actual content of your post? I might have a few issues there too.
You have straightened me up, corrected my methodology and helped me to understand the world better. That's what your words imply. And that's what I mean by condescending.

I would welcome your views but I'd also appreciate it if we could get beyond the "who knew it first" thing. I don't think that adds anything.

User avatar
Cliffard
FISOhead
Posts: 810
Joined: 08 Oct 2008, 00:00
Location: Trondheim, Norway

Re: 4-4-2 and Team Structure

Post by Cliffard »

I love FISO.

User avatar
Stemania
FISO Jedi Knight
Posts: 20448
Joined: 27 Aug 2006, 11:54
Location: On the Iron Throne of xG, the seat of The Crown Prince of the Stat Perverts. Or if not, in the STC!
FS Record: Best: TFF 321st. FPL 129th. FFS Career HoF Peak 2nd (Live 1st). Ability since lost.

Re: RE: 4-4-2 and Team Structure

Post by Stemania »

Ruth_NZ wrote: ....

And that's why I would say that as things stand this season, 4-4-2 is the ideal structure. Not in every season. Last season 3-4-3 was certainly better. But last season's price/performance landscape was different to what we have now.

As a last example, here's a sample 4-4-2 wildcard squad that I could afford right now (and my TV isn't huge):

Pickford Mannone
Clyne Alderweireld Evans/Mee Alonso
Firmino Coutinho Özil Hazard Fletcher
Costa Kane Anichebe
Having just reread the opening post in full, is it not the case that the whole thing (if not the whole thread) could be summerised by the following short simple question:

"Is the best 6.5m-range striker (say Austin) with a backup 4-4.5m defender better than the best 6.5m defender (say Alderwiereld) with a 4.5m striker now that there is a 4.5m striker (Anichebe) that has started a few games?"

Other that that your example team seems to be the very same team structure as the 343, the only other difference from one of the current most common 343 setups being that in your case it's obviously Agueroless.

In your example the cash from Kun's downgrade to the worst of Costa/Kane is basically put into the extra 9.5m range mid (Ozil) over the currently more common Sterling/Walcott (with anything left over beefing up Clyne over a more standard 4.5-5m defender) where others may prefer to fund something like just KDB in that Walcott/Sterling slot instead. :?

Stower79
Grumpy Old Man
Posts: 1366
Joined: 13 Oct 2005, 18:32
FS Record: Previously placed Top 20 Sun DreamTeam, Top 20 TFF, TFF Divisional winner, Top 500 FPL

Re: 4-4-2 and Team Structure

Post by Stower79 »

Ruth_NZ wrote:
As a last example, here's a sample 4-4-2 wildcard squad that I could afford right now (and my TV isn't huge):

Pickford Mannone
Clyne Alderweireld Evans/Mee Alonso
Firmino Coutinho Özil Hazard Fletcher
Costa Kane Anichebe


It's only a sample and I'm not really talking about which are the best players to have, some may prefer KDB and/or Aguero in there. But in terms of shape I think that squad ticks all the boxes. To transition towards something like that from 3-4-3 takes some time unless you want to take a pile of hits. But I think that will probably be my general direction of travel.
I haven't posted for while, but thought I'd pop my head in on this thread as I'm very much having a similar debate in terms of structure. Rather than 4-4-2 though I'm considering whether 3-5-2 or 3-4-3 is the way to go and I am struggling to justify having a third striker at this point in time when there is so much value in midfield and very little up front.

To simplify things, essentially the conclusion you appear to arriving at Ruth is that you'd rather have a fourth defender than a fifth midfielder or third striker.

The first point I’d make regarding your post is that you state:

“We have a better chance of predicting value in terms of points per £1m with defenders than with attackers.”

But where are the stats or evidence to back this up? Team structure has to be directly associated to the various player options at the time so how are you predicting the points?

You also refer to an example that we could not have possibly predicted that a number of midfielders would have scored the points they had:

“Among the midfielders, well you could have predicted that Hazard, Coutinho, Firmino had a good chance of doing well. But most of the others? Not a chance in hell.”

I disagree with this statement and to be honest when you are talking about structuring a team 12 weeks into the season, when we have a lot more info and stats now to base our decisions, discussing what we knew or didn’t know before GW1 is largely irrelevant in my view. We can probably make better predictions now given what we have seen after 12 gameweeks.

I’ll give some basic examples to support my view regarding value and what we could have reasonably expected at the start of the season:

Chadli – 48 points from 7 matches. In two previous seasons he had a similar points value including a 160 point season from just 2407 minutes.

Snodgrass – 54 points from 11 matches. In a previous season he had 152 point season and 105 from 2569 minutes.

Could we not have reasonably predicted that both these players could have returned approximately the points they had?

Especially given their increasing relevance to their teams compared to their former teams.

Moving on to predicted points from this gameweek onwards and your example of a 4-4-2 structured team:

Pickford Mannone
Clyne Alderweireld Evans/Mee Alonso
Firmino Coutinho Özil Hazard Fletcher
Costa Kane Anichebe

For me, I can't see any reasonable argument for holding a Clyne/Alderweireld or even Alonso, over a fifth midfielder in a similar or even cheaper price range at this point in time. I'm referring to the likes of Snodgrass/Chadli/Allen/Capoue/Stanislas who are all outperforming Clyne and Alderweireld by some margin in terms of points per million. Alonso is roughly on a par with these midfielders in terms of value but we surely don't expect Chelsea to keep a clean sheet every week.

In particular, if you look at Chadli and Snodgrass, they are returning the points that could have reasonably been anticipated based on past performances in seasons gone by and their respective roles in their teams. Some of these midfielders are much cheaper as well enabling an upgrade in other areas of the team.

For example:

Snodgrass – 5.5m. It is fairly reasonable to expect at least a 160 point season based on past performance. 5 points per game currently.

Alderweireld 6.3m. Scored 166 last season and his price point is higher than Snodgrass. Currently scored 31 points from 8 matches. 4 points per game currently.

Clyne 5.6m. 38 points from 12 matches. Best points season – 142. Currently just over 3 points per game.

You could make similar cases for Allen/Capoue/Stanislas etc and provide other examples with defenders at the same price range.

The examples above are fairly crude but support the 3-5-2 over 4-4-2 approach in my view - for now at least. And I don’t see any defenders like we have seen in the past (Terry or Baines for example) that would force a change in approach.

As for 3-4-3 it will be interesting to see if a striker in the range of 5.5-7.5 starts to become real value. At the moment the best of the bunch appear to be Defoe and Austin but I wouldn’t describe either as must-haves.

In summary until I see evidence to back-up the 4-4-2 approach I’m not convinced.

User avatar
Stemania
FISO Jedi Knight
Posts: 20448
Joined: 27 Aug 2006, 11:54
Location: On the Iron Throne of xG, the seat of The Crown Prince of the Stat Perverts. Or if not, in the STC!
FS Record: Best: TFF 321st. FPL 129th. FFS Career HoF Peak 2nd (Live 1st). Ability since lost.

Re: 4-4-2 and Team Structure

Post by Stemania »

Welcome back, S79. :D

User avatar
Cliffard
FISOhead
Posts: 810
Joined: 08 Oct 2008, 00:00
Location: Trondheim, Norway

Re: 4-4-2 and Team Structure

Post by Cliffard »

Good post, S79, my thoughts as well. Don't have much to add except my personal experience. I'm on 3-4-3 now, but will look at turning Lukaku and my 5th mid into Anichebe and a max 9.1 mid for 3-5-2.
As for 3-4-3 it will be interesting to see if a striker in the range of 5.5-7.5 starts to become real value. At the moment the best of the bunch appear to be Defoe and Austin but I wouldn’t describe either as must-haves.
This, for me, is key. Austin made me shift from 3-5-2 to 3-4-3. He's on 1 goal in 4 now, but Soton have Eve (H), Palace (A), Boro (H), Stoke (A), B'mouth (A), so I'm thinking of keeping faith and making him one of my two forwards in a 3-5-2.

With a fairly good Team Value, I just think it's easy to pick 5 proven mids and 2 forwards, when you include one or two bargains like Austin/Siggy/Walcott. Used to be Antonio. Plus the even cheaper ones you mention, in Chadli, Snodgrass, and the slightly less predictable Stanislas and Phillips.

User avatar
Stemania
FISO Jedi Knight
Posts: 20448
Joined: 27 Aug 2006, 11:54
Location: On the Iron Throne of xG, the seat of The Crown Prince of the Stat Perverts. Or if not, in the STC!
FS Record: Best: TFF 321st. FPL 129th. FFS Career HoF Peak 2nd (Live 1st). Ability since lost.

Re: 4-4-2 and Team Structure

Post by Stemania »

Yep, the issue all year has imo really been the lack of depth in those 5.5-6.5m mid and striker brackets, from where we've traditionally we've seen either the season bandwaggons or grossely underpriced players. We seem to have been discussing that issue pretty much constantly since preseason though. :D

Fair play to FPL for adjusting the pricing in those areas as best they could, though obviously the lack of super-duper bandwaggons also has a lot to do with happenstance and the fact that we never have been guaranteed that cheap bandwaggons would emerge every year - we've been spoiled in recent seasons.

TheoRiginal
FISOhead
Posts: 643
Joined: 24 Jul 2016, 19:39
FS Record: Virginal

Re: 4-4-2 and Team Structure

Post by TheoRiginal »

While I appreciate all the points being made here, I think that there is another factor being pretty much ignored, which has become increasingly significant as the depth of squads increase with the extra TV money and more PL teams are involved longer in Europe - and idiosyncratic managers (not just Pep, Arsene and Jose, but also Bradley and even Koeman and Puel) rotate for the fun of it, or because they do not know who their best team is, or play games with internationalitis.

Take today, when Pep is choosing to give Stones and Gundogan (preferred European starters!) the weekend off completely....

Which is the formation that mitigates against the chances of not fielding eleven? It seems to me that 343 minimises the risk, given the way that automatic substitutions are managed.

View Latest: 1 Day View Your posts
Post Reply

Return to “Fantasy PremierLeague.com (FPL)”