Strategic Thinking Cavern (18/19)

[raoul]

Regarding the enigma that is Pep, I think we are forgetting the fact that unlike us, he picked his lineups for Everton and Chelsea fixtures after the prior matches had taken place. For all we know he had indeed got Jesus pencilled in for Everton, only for Sergio to bag 3 against Arsenal. Do you drop your star striker after a hat trick against a top 6 club?

Sane is the more interesting one to me. Are we seeing a repeat of the start of the season? Are there "issues"?

I got close to doing Aguero to Auba this week, but then got lucky in that I was struggling to field a starting 10 let alone 11 so fought other fires instead (if I just called Aguero in a home match a fire that needs fighting, then I am a moron). Plus Arsenal away, Huddersfield being perhaps more stingy at home, decided best to wait. That, on top of captaining him in the DGW after what was basically a 50-50 decision with Sane (I recall someone writing on here never to captain a player who is not super-premium), has been a rather nice chunk of fortune.

Although given some of the nightmares I have had this season, I feel like I deserve it!

[Finisher1]

I just meant the fact that they are a profit-making organisation makes them open for criticism.

It was a free-to-view article you linked though, right? They have ads on those since, like FISO, they need to pay for server costs etc (and for a site their size some part time staff).

The accuracy of paid-for info is another question, but it's pretty clear all their articles are essentially opinion pieces and don't necessarily represent the view of the FFS leadership team as a whole. Anyone in the community can write and submit one for a start!

Incidentally, the BBC have started producing regular FPL articles now too - and they are....well...not better.

It was a free-to-view article written by a guy who (if I'm correct) seems to be a full-time editor in FFS.

FFS membership for a full season costs about £15 (IIRC) and a half season membership costs £10. I don't have their financial statements available but I suspect even 10,000 sold memberships per season is a very moderate estimation, could well be a way more. In addition to memberships they also sell ads, like you said. Anyway, it seems obvious they make some really good profit, so it's not just "we-have-to-pay-server-costs" charity thing. But like I said, I don't have their financial statements available so this is just my speculation!

[Stemania]

They may well make a decent profit after paying editor/owner wages (plus upfront costs like subscribing to OPTAs data for the purpose of repackaging, which will not be cheap), but that doesn't mean their opinion pieces have to be anything but that. E.g. all newspapers.

[Finisher1]

They may well make a decent profit after paying editor/owner wages (plus upfront costs like subscribing to OPTAs data for the purpose of repackaging, which will not be cheap), but that doesn't mean their opinion pieces have to be anything but that. E.g. all newspapers.

Editor/owner wages are essentially profit as such, because it means editors/owners get paid for the stuff they produce. I just said it makes their articles open for criticism, that's all. Whenever The Sun or Daily Mail post some questionable stuff, people go furious. So surely FFS is similarly open for criticism.

[Ruth_NZ]

I don't have their financial statements available but I suspect even 10,000 sold memberships per season is a very moderate estimation, could well be a way more. In addition to memberships they also sell ads, like you said. Anyway, it seems obvious they make some really good profit, so it's not just "we-have-to-pay-server-costs" charity thing. But like I said, I don't have their financial statements available so this is just my speculation!

FFS Ltd. have current assets in excess of 500k GBP. The site has made a lot of money and continues to do so.

David Munday (the guy you refer to) is the full-time editor responsible for articles as of this season and writes many of them himself. He is a journalist by profession and was employed on that basis rather than because of his (very poor) FPL record.

Many of those who have leveraged FFS to 'make a name' for themselves have created income streams outside of FFS, whether with 'patreon' subscriptions to youtube channels or via being employed to write articles for mainstream media or to participate in radio discussions on commercial radio. Or, like Crellin, by charging people for transfer and DGW planning advice. And they all make some advertising revenue on their sites/channels, of course. I could name a fair few that have done that. Mark Sutherns, the founder of FFS, became full-time employed by FPL itself at the beginning of last season and appeared on the FPL show regularly. I don't know if that is still active but I do know that FPL pays FFS for services including the ICT ratings and the fixture-difficulty ratings that are used in the game itself. Last season those services included advice on injury status but I believe that function has been taken over by Dinnery/Rotoworld.

In other words, as the trough gets bigger, more pigs appear to feed from it.

[Finisher1]

Thank you Ruth for a very detailed post.

Based on this, I'd certainly expect better quality from David Munday. Yes, they are just opinions but they appear to be based on some vague and more or less irrelevant factors.

I just hate it when at one moment he says Aguero is "very unlikely" to start both matches and then at the second moment he says Aguero is basically safe from rotation. Those are strong claims and they need strong evidence to support them. As a professional FPL-journalist he should do better.

Well I guess I just have to take everything he says with more pinch of salt from now on, and give a good laugh if "suddenly-nailed-for-good" Aguero is actually rested either in GW28 or GW29

[Ruth_NZ]

Nobody checks back, F1, that's the problem.

Funnily enough, I'd have thought that you of all people would have gone for Aguero. I would have done myself - I thought he'd likely start twice - but it was the effective -8 involved in Aubameyang Kun Aubameyang that persuaded me to go another way. Harsh retribution really because that single decision has cost me 66 points, no word of a lie.

[Stemania]

Whenever The Sun or Daily Mail post some questionable stuff, people go furious.

In those two purveyors the challenge is usually to find something that is not questionable.

[Smurphy Paw]

Whenever The Sun or Daily Mail post some questionable stuff, people go furious.

In those two purveyors the challenge is usually to find something that is not questionable.

The Daily Mail is currently and formally considered a ‘questionable source’ by a new media bias checker.
https://mediabiasfactcheck.com/daily-mail/
The Mail are understood to not be happy about this. Their readers...may or may not be aware, it hasn’t been reported
Not sure to what extent this translates to their back pages

[Finisher1]

Whenever The Sun or Daily Mail post some questionable stuff, people go furious.

In those two purveyors the challenge is usually to find something that is not questionable.

It will soon be the same with FFS if this recent trend continues

[Bobby Fetta]

Finisher - I'm surprised you put any faith in that FFS article. You are one of the most knowledgeable people around about lineups etc. In my opinion its not designed for people like you, just disposable articles for people who are not really paying attention. But it is just an opinion piece, not a claim on some inside knowledge - which is really the thing where people do get furious (and actually that was the Guardian earlier in the season that did that - supposedly a respectable newspaper, depending on political allegiance!).

Not to denigrate David Munday, who is actually having a very good season (maybe he didn't really pay much attention to his fpl team previously?), but personally I would put more weight in something you wrote than something on FFS. Hence why I come on FISO but don't bother reading FFS articles any more. I had my worst 2 seasons ever when I used to read everything on FFS.

Back to your original comment. Irrespective of whether it was the right decision to get him, clearly people with Aguero have got lucky with the number of goals he scored. Personally, I thought in a dgw at least he'd get one start and maybe more so was worth the risk. Arsenal was quite a juicy fixture - lots of us captained Salah against Arsenal and were amply rewarded with that at christmas; Everton look terrible at the moment; Chelsea I didn't have down as a good fixture.

[Bobby Fetta]

Exactly CL and EL was what I ment. What would interest me is to compare players form the same team against each others e.g. B/DSilva and Mahrez or Firmino, Mane and Salah rather than compare Aguero to Aubameyang for example. This would give more insight to the question weather one is the main man of certain tea. Also how does a team function when they are not fielding the traditional first 11. Weather lacking a player like Hazard would benefit the fullbacks or the midfielders etc.

This is a very old post but I came across an answer eventually. FiveThirtyEight (Nate Silver's website which covers politics and lots of other things) has an xG model that includes the Champions League and Europa League. The data is free to download (https://data.fivethirtyeight.com/).

There is also lots of other interesting football prediction stuff in there, e.g. predictions for matches and league positions (https://projects.fivethirtyeight.com/so ... er-league/). Even covers the championship - Ipswich 97% to be relegated

[Bobby Fetta]

I haven't posted the form charts for a while. So here they are for the FA Cup break.

I gave them a makeover:
- now divided into separate charts for defence and attack
- includes actual goals as well as expected goals
- shading to indicate over or underperformance relative to xG, inspired by the charts on the experimental 3-6-1 website
- including data for 2016/17 (PL) and 2017/18 (Championship for promoted teams) from fivethirtyeight

Let me know what you think (I still have the old format if people prefer that, but I think the new ones are much better!)

gw26 defence charts.PNG

gw26 attack charts.PNG

[Ruth_NZ]

Let me know what you think (I still have the old format if people prefer that, but I think the new ones are much better!)

Lovely to see as always.

I think the old ones give a better overall picture for the very reason that it's all on the one graph. I liked them for that reason, also they were simpler because they showed less. But these are probably more useful for FPL purposes.

One thing I would suggest is to rank the teams; rather than to have the same order in each set of graphs, to rank the teams in order 1-20 so that the top left is #1 and the bottom right is #20. I'd suggest that ranking be done in terms of actual goals scored/conceded over a longer period (last 19 games?) because the over/under-performance element you now have would then show how those actual numbers relate to xG/XGC. But as an alternative you could rank them in terms of xG over a shorter period (10 games maybe) to create more of a form guide.

The one thing I'd be careful not to lose is the impact of being able to read team trends. The more information you add, the less impact that one element has. The original charts made that stand out, whereas now I have to remind myself to look at that because my brain is busy absorbing more data.

Anyway, that is all meant as constructive input and I really like these charts, always have. Thanks.

[Ruth_NZ]

Let me know what you think (I still have the old format if people prefer that, but I think the new ones are much better!).

Oh yeah, and another thing I forgot. The new over/under-performance element begs the question of why? It's not strictly something about the charts, because they show data rather than trying to explaining it, but it is about how best to understand what you are seeing. So it relates to using (interpreting) the charts I guess.

This is current in my mind because I recently had a look at Leicester and came across the fact that Schmeichel is one of the worst GKs in the PL in terms of xG prevented. He conceded (from recall) 7 goals more than he 'should' have done last season and has conceded 4 goals more than 'should' have been the case so far this season. That doesn't necessarily mean that he is a bad GK, perhaps the chances being conceded by Leicester are actually better/bigger chances than the xG model is recognising? But it's also possible that this consistent defensive under-performance at Leicester is down to a GK that is more prone to errors than some others.

Strangely enough, my impression of Begovic when he was at Chelsea was that he always had a mistake in him and seemed to very often manage to find a way to concede once. That goal didn't matter so much when Chelsea were winning 3-1 but matters very much more for FPL. So the recent replacement of Begovic with 38 year-old Boruc didn't entirely surprise me.

There are GKs on the other side of the equation, who I think can be relied on to beat xGC over a period; Fabianski and Heaton are two examples. But I didn't check the data to see if I am right, that's just off the top of my head.

I guess what I am saying is that some analysis of GK performance over a longer period would be a very useful adjunct to the defensive charts because it might go a fair way down the road of explaining the over/under-performance element. This is the kind of stuff I used to spend hours doing manually and haven't had time for this season.

[Stemania]

I'd agree with all that, ordering might be good idea. Much appreciated BF - but I'm slightly greedier. I vote for both formats as they examine two slightly different, but both interesting, aspects.

[Fuzzy]

Firstly, BF I think these charts are great. Apologies for not acknowledging that before now.

No chart can capture everything (fixture difficulty across a certain period, change in manager , etc). That's possibly an anathema in the STC , but it's fact.

What I really like about the charts, for one who doesn't see as much football as I'd like, is it points me towards possible trends/anomalies to investigate further (to then enable better informed decision making).

Ordering would be great (by G, by xG, by overperformance - you will get a range of opinions), as would relative trend (i.e. rank improving or deteriorating), as a bit of a “nudge” for further investigation.

Thanks again for sharing.

[Bobby Fetta]

Anyway, that is all meant as constructive input and I really like these charts, always have. Thanks.

Thanks for comments - really useful. I'll respond properly when I have some time. I agree about the ordering - it had occurred to me before - just need to automate it. I think I'll reduce the timeframe back to 2 seasons (76 games) as there is a lot of info squeezed in now.

In the meantime, here are the old format charts (long format for now) and table.

[Bobby Fetta]

The new over/under-performance element begs the question of why? It's not strictly something about the charts, because they show data rather than trying to explaining it, but it is about how best to understand what you are seeing. So it relates to using (interpreting) the charts I guess.

Yes, this is part of the reason I posted these. I'm interested in how other people interpret them. I agree with your comments about goalkeepers and I think the chart supports that too to some extent.

Here are some of my thoughts:

- The big 6 teams consistently overperform xG. There is a lot of blue shading in the attacking graphs for those teams. Maybe that is not surprising (they're successful teams, they have the best attackers, etc.) but still important to remember when assessing xG stuff. Was part of the reason I wanted to show actual goals too on here.
- Under or overperformance of xGA is not so clear and seems (apparently) more random. There is much more orange shading at the top of the defence chart. Having said that, some teams are consistently bad vs their stats (e.g. Leicester, Bournemouth, teams with Joe Hart in goal).
- The range of xGA values is actually quite narrow (much narrower than xG). Maybe we pay too much attention to fixtures when considering attackers - fixtures seem more important for defenders.
- The Bournemouth xG vs actual goal trendlines this season are a thing of beauty...
- These team stat charts are of course actually made up of lots of component individual over/underperformances. E.g. the Chelsea attack have pretty much matched their xG values so far this season. But that comes from Hazard and Pedro generally considerably overperforming xG but Willian and Morata (particularly) underperforming.

In terms of recent trends:
- Spurs have improved a lot defensively
- Chelsea attack has struggled a lot
- Man United improvements are apparent (don't need these charts to notice that though)
- Burnley have improved massively
- Everton, Bournemouth, Watford and West Ham have regressed
- Wolves defence has gradually worsened over the season
- Southampton improvement under new manager has been entirely due to getting real goals for/against to match xG/xGA (no apparent improvement in underlying stats, which historically they have always previously underperformed)

[Stemania]

So, I was having a play and was thinking along the lines of


GK
Chelsea, TAA, Maguire
Salah, Mane, Hazard, +1
Higuain, Arnautovic, Wilson

for BGW31 (and 33), possibly trying to leave enough cash in the bank for an Arsenal attacker for 33 (or Spurs if they have no blank). With FH32 obvs taking care of the double. (Or WC32, FH33 if, say, City and Spurs both no blank).

Not sure on the +1, maybe Maddison or Fraser. There are a few millions free so there could also be a wiggle towards Vardy up top.

That is 7 transfers or so from where I am now unfortunately, which is basically all of them (and some). So will have to weigh each individual move carefully against the player they are replacing in the runup. Some GW31 players may be sacrificed to keep a better GW28-30 squad.

I'm certainly thinking Kun to Lacaz this week is a lock-in over Aub, as he's such a great Higuain placeholder and frees up cash for the possibility of Arnie in this week too - who has a great run if fit.

[Sutter Kane]

Is it worth a hit on the keeper for you? Watford probably miss GW31 and 33, although you won't know until too late. I'm unsure about Arnie. He could well be a better purchase than Higuaín or he could get injured again. Also, he has Chelsea away GW33 which is not awful but not good either.

I'll be saving 2 transfers for GW30-31 so I don't alter my squad too much up until that point. But I don't have any City players so pretty easy to play it any which way and not hurt the lead-up squad. Auba/Laca to Higuaín in GW29 is an easy swap regardless.

Lots of people inc me, are left in a bit of limbo with regards Son. Maybe worth removing him bang on GW31 then bringing him back in GW33 if Millwall win.

[Stemania]

Yeah, I'm also thinking of trying to leave most of the moves as late as poss, probably 4 transfers in GW30-31 combined. Though Chelsea in particular could come earlier with their fixtures (Lacaz to Higuain in GW29 is likely indeed).

Tempted to try to keep Son on bench just in case, just because money isn't so stretched. But will inevitably fold for shinier things.

He's been great, but Foster's fixtures get ratty around GW28 anyway, so will probs chuck then - maybe even for Kepa instead of a Chelsea def!

Arnie I agree could go either way - but he has a nice run now so is almost independent of GW31 for my team (having Ings as I do). Shame Wilson is out as I'd probs have just got him in now otherwise, despite a few tricky games.

[FranckKessie]

The only certainty I have at the moment is that I will spend many transfers late.

Formation is as always of importance as it usually take up an additional transfer to go from one to another. There are several match ups where I am not certain which is best,
Fraser/Wilson/Brooks
Arnautovic/ Anderson/ Antonio
?
Timing of when the transfer is to be made is of course important. With Wilson and Brooks out Fraser pick himself if needed early.
I be inclined to prefer Anderson to Arnautovic if an early transfer is needed. Anderson imo provide greater certainty and the mental state of Arnautovic seems questionable atm and I dont like an Arnuatovic who is somewhere else mentally.

Barnes in attack looks interesting imo.

[Aldershot Rejects]

So, I was having a play and was thinking along the lines of


GK
Chelsea, TAA, Maguire
Salah, Mane, Hazard, +1
Higuain, Arnautovic, Wilson

for BGW31 (and 33), possibly trying to leave enough cash in the bank for an Arsenal attacker for 33 (or Spurs if they have no blank). With FH32 obvs taking care of the double. (Or WC32, FH33 if, say, City and Spurs both no blank).

Not sure on the +1, maybe Maddison or Fraser. There are a few millions free so there could also be a wiggle towards Vardy up top.

That is 7 transfers or so from where I am now unfortunately, which is basically all of them (and some). So will have to weigh each individual move carefully against the player they are replacing in the runup. Some GW31 players may be sacrificed to keep a better GW28-30 squad.

I'm certainly thinking Kun to Lacaz this week is a lock-in over Aub, as he's such a great Higuain placeholder and frees up cash for the possibility of Arnie in this week too - who has a great run if fit.

There's a lot of team value in that team, presumably you've got only cheapies on the bench. I've got lots of value tied up in Doherty, Pogba & Rashford which I would rather not lose - presumably with that type of team you would be biting the bullet. I do like it up as a line-up though.

[Pulpy]

I had planned to bring in Arnautavic this week but now want to see him starting and scoring again first, even if it means potentially missing a haul versus Fulham.

This article is quite an interesting read:

https://talksport.com/football/491416/m ... urst-park/

[Stemania]

In view of Jesus' hamstring I'm a little tempted to swerve the Lacazette hokey-cokey and just go Vardy, in order to keep Aguero a little longer.

So Ings, Kamara, KDB to Vardy, Arnie, 4.5m mid (-8).

It would save a -4 later, so is in some sense a false -8. Most likely means Vardy over Wilson for 31, but given the fitness issues that might not be a bad thing.

Question is, do I want two potentially stroppy strikers?

[Sutter Kane]

And do you want to spend -8? If your transfers are serving GW33 as well, I can understand it but looking that far into the future, with as you say, stroppiness, is a risk. Arnie at Chelsea GW33 too. 2 points likely there.

I can see why you'd want to keep Kun though. I was toying with bringing him in GW28 as part of a hokey-cokey but the expected returns just doesn't quite outweigh the nuisance of it all. I said that about gw25 as well though...

[Sutter Kane]

If I had Kun I'd definitely keep him now Jesus is injured. No question for me. That Jesus injury is quite possibly the third most annoying occurrence of the season so far, after Aguero's GW25 abomination and then subsequent GW26 hatrick.

[Stemania]

And do you want to spend -8? If your transfers are serving GW33 as well, I can understand it but looking that far into the future, with as you say, stroppiness, is a risk. Arnie at Chelsea GW33 too. 2 points likely there.

Nope, would prefer not to -8 but am fielding 9 + Bednarek (ars) as it stands, and could afford neither transfer on it's own if I keep Aguero - so even just as a consideration for this week alone it would in effect be -6 to upgrade Bednarek (ars) to Vardy (CRY) + Arnie (FUL), which looks a great deal.

Chelsea's defence doesn't look the scariest, but 33 isn't ideal I agree - but the transfer wouldn't really be about then. Vardy and Arnie have amazing runs all the way till 33 (excepting Arnie in 28 & 33) so would be well up in my list of potential transfers even if it were not for the blanks. Leicester's run is particularly disgusting.

[Finisher1]

I think it's about 95% likely that Arnautovic will be at least yellow flagged at least once before GW33.