The optimal team

[Mav3rick]

I'd think that a lot of this optimal team being pickable at the start stuff rests on captaining Salah every week, rather than any magical formula. In seasons past you'd generally need an expensive captain, but without that requirement, the extra funds flow into defence and DDG, which is where AR is correct because it's relatively easy to pick the higher scoring defenders, even if cheaper ones might be better value.

So, my instinct is that the pre-requisite of a no transfer team doing well is based on a 300 point captain, when many of us were, for some reason, preserving with other players like Kane.

That's probably the lesson more than anything else, why did so many of us not captain Salah regularly, at least after the first 10 GWs?

[MoSe]

I'd be just curious to know the source, whether for instance it was BB own research, or he got it from one of his friends, or if he found it somewhere on the internet where we could access it too.
That's not to question anything, on the contrary, it's to compliment the guy who got the idea and actually went on to search whether there was one.

The idea would be nothing original in itself of course.
In any game where you have to make choices, even before I came to FPL, it's just natural to think to check afterwards which would have been the best choices. Which would have been the best possible result, as a benchmark, a framework, a ceiling against which to assess your own performance. Of course that can be done in hindsight, and would have mostly required clairvoyance at the start

In my rookie FPL season I tried to do something like that n the first 3 gws, for my friends ML and site, before I found out Fiso existed.
In GW1 of course I saw that FPL provided the Dream Team. And IIRC it was immediately evident that it just collected highest scores without respecting a legal XV restraints, namely budget and club cap.
For GW2 I did it "manually" in excel, checking also all possible transfers (involving those who had a high score in either week, not of course literally ALL of them), which ones were profitable taking a hit for, or whether it was instead convenient to keep a player who had a high score in one week but just a medium one or a blank in the other. I remember that it was possible, approaching it with logic and not just toilwork, to determine that the solution was formally the best possible legal one for the two gws. Not just the best I could find, but 100% the best possible one.
It took me maybe a couple hours IIRC (it was 4 seasons ago...).
I went on to do it after GW3. It took me something like a couple DAYS. I think to recall it involved something like taking a hit to bring on Adebayor and then another to remove him

Of course I stopped at that, because I had tested what I wanted to check and learn, and it would have required coding a program to do it beyond 3 gws.

Here the idea is simpler, thus maybe simpler to check and find, or... to discard and overlook as you wouldn't think it's possible!
I wouldn't want to make up my memories now, but I have a vague recollection that after my rookie season I did try to check whether there was a simple "season dream team" which could improve the winner score, and soon found out there couldn't be a legal one.
So it's quite possible that you can't be doing the same in every season, and it makes it more remarkable that it could be done this one.

Remind that it is
1. a LEGAL season start XV selection
2. no transfer during the season
3. no chips used
4. no lineups changes, same XI starters and C(VC) all season
5. normal functioning tho for autosubs and VC, when a starter or the captain Did Not Play

Relaxing restraints 2 3 4 would allow to improve the final score but would make the search much more complex and impossible to find within reasonable effort and time.

We could just maybe try to deploy TC and BB to the OP optimal team in the appropriate two GWs

[Aldershot Rejects]

Pretty sure it was https://twitter.com/fplkernow that came up with it.

[MoSe]

Great trove, thanks AR (was it a known source, I mean, We were you following the guy beforehand?)

Interesting how from a previous tweet
https://mobile.twitter.com/fplkernow/st ... 5332317185
we understand that he went seeking it tentatively, i.e. he previously had a different template which he called the "dead team"

The fpl "dead team" ended up finishing 20 points behind the overall winner.
4:17 PM · 14 mag 2018

It's a hypothetical team I put together when I was doing my wildcard earlier in the season and then updated the progress of by referring to it as the fpl dead team.
9:18 PM · 15 mag 2018

It was a similar 532 team with Monreal & Kane in place of Sané & Vardy (subs all different)

I wonder if he gave out his FPL team link/id seeking further down his tweets, so that we could know when did he WC

Thus maybe later he tried to tweak a bit starting from his early season template, and found out that with a couple changes and adapting the bench to optimise subs only when they were actually needed, he could improve his "dead team" by 30p.

So maybe it's possible that with a more systematic approach, even better solutions could be found....

OTOH this puts the feat under a different light...
I didn't research his older tweets to see when he actually had declared his "dead team", but...:
actually putting together a template in early season,
discarding it,
and then realise you could have won FPL using it as a "dead team"
even more remarkable, and... he must be gnawing at his elbows... (howdoyousay being full of regrets?)

[Aldershot Rejects]

Great trove, thanks AR (was it a known source, I mean, We were you following the guy beforehand?)

No, I just tripped over him this morning, and thought of this discussion.

[MoSe]

For winner yusuf, from anewpla.net
Captains Used
Salah (16 times - 309 points - average 19.3 points)
Lukaku (7 times - 104 points - average 14.9 points)
Sterling (6 times - 70 points - average 11.7 points)
Kane (5 times - 58 points - average 11.6 points)
Son (1 times - 26 points - average 26 points)
Mahrez (1 times - 20 points - average 20 points)
Jesus (1 times - 2 points - average 2 points)
Aubameyang (1 times - 12 points - average 12 points)

He won FPL with1512p, 10 less than fplkernow revised dead team

[Football Hero]

Of course, maybe the best way to play the game this season was to go for a premium defence and avoid 10m+ players and to pack your midfield and attack with £8m-£10m players.

No, surely the best way to play the game this year was also the best way to play it last year and the year before right? Because you can only make decisions ahead of time, not in hindsight. How can you learn anything from thinking that you supposedly should have gone with a premium defence for this one particular season, but normally you wouldn't need to in other seasons? What would have been the clues for you last summer that would have highlighted that premium defenders were needed? If those clues were there, then why didn't you pay heed to them? Or was it that the clues weren't there, in which case you should not have gone for loads of premium defenders this season and were correct for avoiding them.

Suggest that before you comment about my side you actually have a look at what I did this season. For your information - I had three of these defenders in GW6 alongside Kolasanic and Jones.
I did the same but not to quite the same extent at the end of last season and will be starting next season in similar vein.
One of my (many) mistakes this year was having Kane for most of the year, would have been better to have freed up the value and gone with Firmino and or Vardy. Kane was good value at £11m, he's not at £13m.

The same for me, I had my worst rank in a while this year, mainly through persisting with Kane during the early season where he failed to deliver in many easy home matches. Then I dumped him around his blank when he had trips to City and Burnley, and a game against Southampton who weren't getting thrashed much at the time. Kane delivered back-to-back hatties while my replacement Lukaku was again not worth having, (other than the first few weeks of the season, he really didn't deliver much).

Despite it not working out, I believe it was still good play to persist with these players when looking at the fact these two were the best two strikers over the last couple of years, even when Lukaku was at Everton.

[Football Hero]

Pretty sure it was https://twitter.com/fplkernow that came up with it.

This is assuming that no TC or BB were ever used either right? So someone had chosen this team back in August, and then never logged back in to FPL, they would have won it, likely never even realising if they didn't check their emails much either.

[Football Hero]

Surely the fact they weren’t injured is the reason they are in this thread? By its nature, that’s what the optimum team is... whether it’s Gross or Hazard that makes the team. That’s the beauty of selecting the best team after the event.

I'm assuming that in most seasons, there isn't a ''set and forget'' 15 that would outscore all teams in FPL, (hence this thread), particularly if the best players of that season have long spells on the sidelines at certain points.

I could be wrong of course, but I haven't seen this mentioned in other seasons that there is a common 15 that you can use throughout the season that beats all other players who are using transfers.

Understood.

So it’s the fact that this years optimal team would have won (rather than the optimal team per se).

Sort of, although this year's true optimal team would have been enhanced with the use of correct transfers and would have scored 2800 points say. This set and forget team still wasn't optimal, even though it would have won.

That is because perfect optimal play would hypothetically allow you to score the maximum points available in a given season which would be in the high 2000/low 3000 points range.

[Football Hero]

Of course, maybe the best way to play the game this season was to go for a premium defence and avoid 10m+ players and to pack your midfield and attack with £8m-£10m players.

No, surely the best way to play the game this year was also the best way to play it last year and the year before right? Because you can only make decisions ahead of time, not in hindsight. How can you learn anything from thinking that you supposedly should have gone with a premium defence for this one particular season, but normally you wouldn't need to in other seasons? What would have been the clues for you last summer that would have highlighted that premium defenders were needed? If those clues were there, then why didn't you pay heed to them? Or was it that the clues weren't there, in which case you should not have gone for loads of premium defenders this season and were correct for avoiding them.

Suggest that before you comment about my side you actually have a look at what I did this season. For your information - I had three of these defenders in GW6 alongside Kolasanic and Jones.
I did the same but not to quite the same extent at the end of last season and will be starting next season in similar vein.
One of my (many) mistakes this year was having Kane for most of the year, would have been better to have freed up the value and gone with Firmino and or Vardy. Kane was good value at £11m, he's not at £13m.

Appeared to be talking in generalities to me...so if you re-read what he wrote with the word "one" replacing every "you" it makes reasonable sense no?

Apologies if I did misread. Having said that I don't think the performance of those four defenders is any surprise at all. Azpil & Alonso pretty much repeated their performance from the previous season. Davies is playing for a solid defence and has proven attacking ability. Man City central defenders were scoring regularly at the end of last season and in pre-season and also seemed to have tightened up considerably defensively - for me it was a toss up between Kompany and Ota before the season began. I would argue that the performance of the highest scoring defenders is one of the most predictably things in the game.

I disagree that the high scoring defenders are easily predictable each year, (save for OOP 'defenders' like Alonso). Plus Azpil and Alonso doing well wasn't that great when they are priced appropriately too, using up more budget than you would ideally like in the process.

Look at Alderwiereld's attacking returns a few years ago compared to last season and this. Or Otamendi's. Or McAuley's. Or Monreal's. Randomly each season there will be defenders that fluke into getting 6 goals and 3 assists when in the previous three season's combined, they mustered 2 goals and 0 assists. They come out of nowhere with hauls, where corner kicks keep landing on their heads with them being in the right place at the right time to score.

[Smurphy Paw]

Could someone check my maths?
I think that playing Pope, Lossl, Robertson is 5 points better than DDG, Hennessey, Mariappa

[Billy Bongo]

Pretty sure it was https://twitter.com/fplkernow that came up with it.

Yep just saw it on Twitter and posted here for something to talk about over summer

Sent from my SM-G965F using Tapatalk

[Billy Bongo]

Pretty sure it was https://twitter.com/fplkernow that came up with it.

This is assuming that no TC or BB were ever used either right? So someone had chosen this team back in August, and then never logged back in to FPL, they would have won it, likely never even realising if they didn't check their emails much either.

Yes that's right, it's a team you could have picked, done nothing, chips or transfers, fiddly, and won fpl

Sent from my SM-G965F using Tapatalk

[MoSe]

Could someone check my maths?
I think that playing Pope, Lossl, Robertson is 5 points better than DDG, Hennessey, Mariappa

From mobile, but let's try it.

DDG Utd 5.5m 172p dnp 31 35 38
Hennessey cpl 4.5m got 22p in the 3 gw when ddg dnp
Total 194p for the GK spot

Pope bur 4.5m 152p dnp 1 2 3 31
Lössl hud 4.5m in those 4 GW got 9 8 6 2 = 25p
177p so -17p on fplkernow combo, but 1m to improve on Mariappa 4m

Mariappa wat 4m was needed in 14 26 31 35 38 when he got -1 2 0 5 2 = 8p
Robertson liv 5.0 (at season start too) got - 1 11 - 18 = 30p

So yes
Fplkernow trio 202p
Smurphy's trio 207p

As I said above, fplkernow tweets suggested he didn't have a systematic approach, and probably even better solutions for a "dead team" could have been found.

[Smurphy Paw]

Yay! I win FPL!

[MoSe]

So far
Anyone with something better?

[Smurphy Paw]

It should be possible. My starting point was Mariappa vs Robertson in GW#35&38, and then raising the cash. DDG seemed the obvious place to look. I didn’t try anything else so somebody must be able to do better

[Aldershot Rejects]

Not done the sums and it might sound crazy but simply starting with Robertson and making Azpilicueta (or Otamendi) 1st sub might increase the total.

[Football Hero]

What about increasing the points total with the odd choice transfer here and there, like having Otamendi to start with, and then transferring him out for Robertson at the halfway mark.

[MoSe]

What about increasing the points total with the odd choice transfer here and there, like having Otamendi to start with, and then transferring him out for Robertson at the halfway mark.

hmm... maybe you have me on ignore and can't read my posts...

Remind that it is
1. a LEGAL season start XV selection
2. no transfer during the season
3. no chips used
4. no lineups changes, same XI starters and C(VC) all season
5. normal functioning tho for autosubs and VC, when a starter or the captain Did Not Play

Relaxing restraints 2 3 4 would allow to improve the final score but would make the search much more complex and impossible to find within reasonable effort and time.

[Billy Bongo]

What about increasing the points total with the odd choice transfer here and there, like having Otamendi to start with, and then transferring him out for Robertson at the halfway mark.

It's a hypothetical team proving its possible to win by doing nothing

Sent from my SM-G965F using Tapatalk

[Aldershot Rejects]

Pretty sure making Azpil S1 and starting Robertson would work.
Also wonder about Sane to Silva and then Mili to Gross. Although Sane to Silva on its own might do the trick.

Busy day tomorrow so I might not get chance to have a look.

[Smurphy Paw]

I’m enjoying the non-FPL time so I can’t see me doing the extra detail. My aim was just to see if improving it was possible - we’d all accepted the score at face value.
If somebody is so-inclined, it is worthwhile spending a bit of time though before we think it gives us the backing to pick, for example, 5/6 premium defenders next time (with Robertson that makes it all five defenders +/- DDG). Ditto, sub-£10m strikers.

[Billy Bongo]

The one thing I've learned from this is that I need to spend more on defenders next season

Sent from my SM-G965F using Tapatalk

[Sutter Kane]

Which defenders though? They aren't all going to be big points scorers - when they score their first goal, if they do, no idea whether they'll get 2 goals or 6 goals and that makes all the difference.

[Billy Bongo]

Well, I went budget last year and it didn't work out, I need more money there, Mendy Pva Roberston Rudiger Monreal amongst could be great value

Sent from my SM-G965F using Tapatalk

[Mav3rick]

It really doesn't say big defence to me, it says, pick the right captain!

If Salah is 12m next season, the equation will be different, very different if someone like Aubameyang or Kane also reaches the expected heights.

I think FPL have really balanced out the prices of premium defenders, 4 man backlines and cheaper strikers over recent years, so well in fact that most (non-extreme) approaches are pretty even as a start point. An extra midfielder, an extra premium defender, a third striker, it's much of a muchness and likely not the key to the dead team.

The overwhelming factor, IMO, is Salah (C) for 9m and no second captain pick. You then have loads of free budget with no Kane, Hazard, Sanchez etc so it's obvious to spend the cash in defence.

Next season, as ever, to start the year you should ensure you have the right captain, pick the value options from transfers/underpriced players, and then build the squad out from there.

The one lesson I would consider is the merit of a second captain option, in seasons gone by we've often seen pre season FPL sides that try to squeeze 3 super premium options (even this year, we had the Kane, Lukaku, Morata type of front line). My feeling has always been that two super premiums is the right balance (ideally one striker and one mid), but maybe it should just be one that you back consistently.

I'm pretty sure my captaincy switching is very poor, perhaps everyone's is, and that if I just pick one super premium player and leave it on them, I'd do better (fairly sure that for two if not three seasons now I have done worse by switching captain) even before considering the benefits of the extra fund spread.

[Billy Bongo]

Nevertheless, defence in this team is a lot more than I spent

Sent from my SM-G965F using Tapatalk

[Aldershot Rejects]

I'm all for big defence, but as Mav says the fact that the best captaincy option was 9m obviously affects the shape of this team. I also agree re captaincy switching, I'm seriously considering auto-captain once next season settles down a bit.

In terms of super premium option, I suspect the maths mean that it is hard to get value from having more than one in your team, unless of course there are one or more £6m attacking players scoring very big. In fact, I suspect that any super premium player has to be scoring at least 6.5 ppg to fully justify their place in their side. That's my hunch anyone. One of my job's for the next month or so is to think a little bit about that.

[Smurphy Paw]

Mav might well be on to something about the captaincy - not instead of a discussion about defenders, in my view, but on to something nonetheless.
I have just compared my team to the Top 50on Fix’s tool. It’s a bit of fun, and shows that in a number of positions I outdid the Top 50, and in a few I didn’t. Taking fewer hits is something else to consider.
However, the single biggest difference between my team and the Top 50 average was Captaincy choice. I lost 81 points on their average (301 vs 220) and I was only 60-odd points behind.
Or, to present that another way, had I been as successful as the Top50 in picking my Captain, I would have been in the Top25.

At this point the assumptions about Salah fall apart. One of the main reasons that I was behind is that I captained Salah a lot more than the Top50 average. They gained far more points from Kane who, typically, I also had.
One anecdote doesn’t make it true, or the instincts on the thread untrue, but it’s food for thought.