Patrician's Crystal Ball

[BLOCKHEAD]

5 goals have been disallowed by VAR in the last 2 games at the Copa America. Brazil also had 3 goals disallowed against Venezuela in an earlier group game.

This is leading to a lot more 0-0 results. Most of the goals are disallowed due to very tight offside calls.

[dod]

Although we have a different structure yours is the most similar team to my own I have seen so far. I notice however that you have fallen a little short of the "11 of your squad should be 6m or less" that you proposed as ideal last season .

As I suggested at the time that was the ideal set up for last season and each season is different. Trying the same thing this season with 11 '6m or less players' and 4 'big hitters' doesn't look at all viable.

Good looking squad though .

[fred1266]

5 goals have been disallowed by VAR in the last 2 games at the Copa America. Brazil also had 3 goals disallowed against Venezuela in an earlier group game.

This is leading to a lot more 0-0 results. Most of the goals are disallowed due to very tight offside calls.

but some has also been given of the top of my head Argentina and Paraguay

[Mav3rick]

If you are targeting an average of 70 points per week

How do you realistically target 70 points per week though? That's an average of 5.83 over 12 players (including the captain) which is quite hard to reach. I guess you're looking for a basic range for each position in your team, with boosts for form/fixture runs and a high captaincy score?

[Patrician]

As I suggested at the time that was the ideal set up for last season and each season is different. Trying the same thing this season with 11 '6m or less players' and 4 'big hitters' doesn't look at all viable.

I don't think it can be done with any of the 12M+ players, but a very nice looking team can be assembled with Kane/Auba, Vardy, Mane and De Bruyne, plus Robertson and TAA. I think that combo plus whichever 6M or less players emerge might be a winning formula.

[Patrician]

How do you realistically target 70 points per week though? That's an average of 5.83 over 12 players (including the captain) which is quite hard to reach. I guess you're looking for a basic range for each position in your team, with boosts for form/fixture runs and a high captaincy score?

It is very tough, but an average of 70 points per week is what won it last year, and it can only be done catching the boosts for form/fixture runs as you suggest. That is what led me to do the analysis on the top players and how much they score against weak opposition. This showed that it might not be completely unreasonable to average 10pts (doubled) for captaincy over the season.

Last year I targeted 65 points per week and managed 63, this year I target 70. I am sure I will fall short but will it inspire something >65 points per week? Lets see.

[Mav3rick]

It is very tough, but an average of 70 points per week is what won it last year, and it can only be done catching the boosts for form/fixture runs as you suggest. That is what led me to do the analysis on the top players and how much they score against weak opposition. This showed that it might not be completely unreasonable to average 10pts (doubled) for captaincy over the season.

I see, I suppose setting that target is what's crucial. I have no doubt that the winning team tends to have high captaincy scores, I'm sure that's survivor bias of course, so I don't think I could personally aim for 10 points captaincy. It's just not possible to do that in my experience, but if you just do what you can to manage the captaincy as high as you can then you may get lucky and come out way ahead of what you'd expect.

[dod]

As I suggested at the time that was the ideal set up for last season and each season is different. Trying the same thing this season with 11 '6m or less players' and 4 'big hitters' doesn't look at all viable.

I don't think it can be done with any of the 12M+ players, but a very nice looking team can be assembled with Kane/Auba, Vardy, Mane and De Bruyne, plus Robertson and TAA. I think that combo plus whichever 6M or less players emerge might be a winning formula.

It could well be. I wouldn't argue against any of those players coming good but they aren't nailed on super-premiums like we had last season. My issue is the impracticality of picking all those 6m or less players you need without hindsight. There's a huge amount of chaff to sift through to get to the few grains of wheat.

I've no doubt that in hindsight you may well be able to pick a near optimum squad using that formula but only in hindsight. I also don't think it will be as good as last season's offering. The price structure this season just won't allow it. Players are going to emerge in the 7-9.5m range that are going to be the best value picks this year. You've already included 2 possibles among your 4 big hitters.

What differentiates FPL from other risk management games is that the rules and values of the pieces change, therefore it is impossible to say before the start of each season what the best strategy and structure are. That for me is the attraction of the game. What is however possible is to identify with a useful degree of certainty the strategies and structures that probably won't ever work, and avoid them.

[Patrician]

I have no particular issue with players in the 7-9.5M range, I am just very clear that I will need to armband them to hit the pts/game I need from them.

Lets take Siggy as an example, 8M, great player, will score well. But for 8M I need to average 6.3 pts per week. Last season he averaged 4.8. He is miles off value for money, unless I captain him some of the time. How often am I realistically going to captain him? Hardly ever. So no Siggy for me, even if his starting fixtures are amazing.

[Sutter Kane]

He has great fixtures though, so the defences may leak enough for him to get 6ppg.

As for 10ppg per week captain, you need massive luck I think - wasn't Mane electric at home last season though? If he repeats that pattern then it's a predictable advantage that could result in a huge ppg for the captain if you can get the captain scores decent from the other games where Mane isn't captained. In fact, having Salah would be pointless in that squad. I think certain players react more to playing at home that perhaps playing different levels of opposition. It's tough to know these things yet, possibly not until the season is almost over.

[Sutter Kane]

Good idea to search for patterns though. If they exist that is!

[Billy Bongo]

I have no particular issue with players in the 7-9.5M range, I am just very clear that I will need to armband them to hit the pts/game I need from them.

Lets take Siggy as an example, 8M, great player, will score well. But for 8M I need to average 6.3 pts per week. Last season he averaged 4.8. He is miles off value for money, unless I captain him some of the time. How often am I realistically going to captain him? Hardly ever. So no Siggy for me, even if his starting fixtures are amazing.

Hmmm, but second half both him and the team improved significantly, that's more relevant imo

Sent from my SM-G975F using Tapatalk

[Patrician]

Second half he averaged 4.8 pts per week, 5.2 if you count gameweeks 18 and 19. Still miles off.

[dod]

I have no particular issue with players in the 7-9.5M range, I am just very clear that I will need to armband them to hit the pts/game I need from them.

Lets take Siggy as an example, 8M, great player, will score well. But for 8M I need to average 6.3 pts per week. Last season he averaged 4.8. He is miles off value for money, unless I captain him some of the time. How often am I realistically going to captain him? Hardly ever. So no Siggy for me, even if his starting fixtures are amazing.

Well at least you've thought it through. I'm not convinced but I haven't put in the work you have so will be watching your team with great interest.

I just don't believe in a linear relationship between points per million spend as it relies on your available discretionary budget. Just because a player has a higher ppm does not mean that owning them is a better use of your DB than one with a lower ppm. In the extreme case of an unlimited budget ppm becomes irrelevant as you would just have all the highest point scorers irrespective of value and captaincy.

Forgive me if I have got the wrong end of the stick and am misrepresenting your position.

[Patrician]

Maybe I should add another principle 12) Be flexible, rules are there to be broken, there is no such thing as a fundamental truth in FPL

[Patrician]

STARTING SQUAD PRINCIPLES UPDATED FOR 19/20

1) Assess contextual situation, e.g. VAR, Wolves in Europe, Burnley out of Europe, Chelsea in flux.
2) Assume a GW6-8 wildcard - prioritise teams with easy first 6 fixtures
3) Select three or four 9M plus players that rotate well for captaincy
4) Focus on the first XI. Base price players for the bench, but ensure a decent 12th man.
5) Pick a premium priced player in each position except goalkeeper, for flexibility
5) Spread price points for flexibility
7) Identify and cover the key price points (e.g. 2+ good options), and avoid price points with just one good option (e.g. maintain 1 move transfers)

SEASON PRINCIPLES UPDATED FOR 19/20

Primary Goal: 2,700 points, which means quickly establishing a team which will on average score ~70 points per week

1) Relentlessly seek and drive towards the highest average xPts combination of players
2) Value for money without the armband comes at 6M and below, eleven of your squad should cost 6M or less (can break this rule for certain premium defenders like TAA and Robertson). If they cost more than 6M and you never plan to captain them, they have no place in your team.
3) Identify the core VfM players as soon as possible in the season (more on this later), get them in and stick with them.
4) Ignore the crowd. Never chase points unless they are supported by underlying stats
5) Don't hesitate to make the obvious plays (e.g. Aguero transfer in 17/18)
6) If the stats are supportive, be patient (e.g. Kane 17/18, Wilson 18/19)
7) Wildcard early (but try to have > four weeks of data) to take advantage of the models strength
8) Don't let the chips drive strategy, and don't be adverse to playing them outside of a DGW
9) Avoid hits unless there is a clear expectation of neutral xPts or better
10) Avoid early GW transfers, unless price changes really force the issue - wait until Fri/Sat
11) Maintain price point flexibility, a decent enough bench, and a strong 12th man. Injuries and rotation are inevitable
12) Monitor form closely (last 4 gameweeks), to spot emerging players and changes in role/tactics
13) Fixtures matter, especially for the premium players, so rotate for good fixture runs
14) Be flexible, rules are there to be broken, there is no such thing as a fundamental truth in FPL

CAPTAINCY PRINCIPLES FOR 19/20

1) Perhaps unsuprisingly, generally captain the best premium playing at home against the worst defence...
2) ...unless it is Sterling, who seems to do better against more average defences (know the player)
3) Generally favour midfielders. Big hauls come most frequently from the top midfielders
4) Don't forget TAA - he got 9 big hauls, 6 of which came at home for an average of 7.3

[Aldershot Rejects]

I have no particular issue with players in the 7-9.5M range, I am just very clear that I will need to armband them to hit the pts/game I need from them.

Lets take Siggy as an example, 8M, great player, will score well. But for 8M I need to average 6.3 pts per week. Last season he averaged 4.8. He is miles off value for money, unless I captain him some of the time. How often am I realistically going to captain him? Hardly ever. So no Siggy for me, even if his starting fixtures are amazing.

Interested in your benchmarks/targets for each price range. I'm targetting 5.1 ppg for an 8m midfielder which is quite a bit lower than you, but I'm working on an overall target of 57 ppw before captaincy/chips etc (and 2,600 for the season).

[Patrician]

Here you go...this is what is roughly required to hit 70 pts per week. Once you hit 7.5, the pts/week passes a line where only the top players hit that target without the armband. Ergo, higher than 7M needs the armband to keep up.

Pts Targets.JPG

[itslikebrandnew]

I assume that means that your first 11 must be max price allowed

[itslikebrandnew]

I have no particular issue with players in the 7-9.5M range, I am just very clear that I will need to armband them to hit the pts/game I need from them.

Lets take Siggy as an example, 8M, great player, will score well. But for 8M I need to average 6.3 pts per week. Last season he averaged 4.8. He is miles off value for money, unless I captain him some of the time. How often am I realistically going to captain him? Hardly ever. So no Siggy for me, even if his starting fixtures are amazing.

Interested in your benchmarks/targets for each price range. I'm targetting 5.1 ppg for an 8m midfielder which is quite a bit lower than you, but I'm working on an overall target of 57 ppw before captaincy/chips etc (and 2,600 for the season).

57ppw looks fairly easy to hit. A little bit of tweaking with my RMT and i'm about 57.5

[Patrician]

Before Captaincy 57 is great, if you can average 8pts per week for captaincy, then you are at 65, which could win the whole thing!


Sent from my iPad using Tapatalk

[Patrician]

I assume that means that your first 11 must be max price allowed

Yes the table assumes base price bench.


Sent from my iPad using Tapatalk

[itslikebrandnew]

I assume that means that your first 11 must be max price allowed

Yes the table assumes base price bench.


Sent from my iPad using Tapatalk

Thanks, i'm at 59.1 with some tweaking.

[itslikebrandnew]

59.5 but it looks like a team i can't rely on.

[Aldershot Rejects]

Here you go...this is what is roughly required to hit 70 pts per week. Once you hit 7.5, the pts/week passes a line where only the top players hit that target without the armband. Ergo, higher than 7M needs the armband to keep up.



Pts Targets.JPG

Thanks; will have a look at these later.

[Patrician]

Draft 2, four premiums

This draft sticks much more to to principles learned from last season. Quite reliant on Kane and KdB returning to former glories. Trippier becomes Maguire if he moves to Man City. This team is super strong I think.

Draft2.JPG

[Patrician]

3 Premium Version

Lower upside, more certainty.

Draft 3.JPG

[Stemania]

I love this thread!

But, but, but. I'm afraid to say I still don't get the whole idea of aiming to hit the 'winning' score by trying to pretend to be the team that gets the winning score, if that makes sense. Already proof of it's somewhat arbitrary nature is that 2500 was the total that I recall having previously seen suggested in the 'tiger-suit' methodology, yet now we have 2700 because last year was oddly high scoring. My point is basically that surely the best team is the best team, regardless of how last season went or who will win FPL this year!

Mainly, I just wonder, if your first principle is maximising the highest average xPts combination of players (which is mine too!), and you already have your own xPts calculations for each player of interest, they why not just packing-problem the solution to find the best team based on your numbers? There's a spreadsheet or two in the STC of Mav's creation from last year that could easily be modified for that (or TSO's site teamfill.com perhaps). Once you have produced projections you believe in, the game is pretty much finished (via packing) as far basing it purely on points projection goes - OK, you might want to afterwards add in your flexibility principles etc to modify - so I don't see how the 2700/2500 points benchmark things actually come in to it or are helpful in finding the best team in practice.

Lets take Siggy as an example, 8M, great player, will score well. But for 8M I need to average 6.3 pts per week. Last season he averaged 4.8. He is miles off value for money, unless I captain him some of the time. How often am I realistically going to captain him? Hardly ever. So no Siggy for me, even if his starting fixtures are amazing.

Further, to be a bit more specific on the benchmarks thing, another potential dodgyness is using these 'season long' points projections/ppg. I struggle to see how the Siggy "needs to average 6.3 pts per week" is so useful an idea to have, for example. If you were just using season long ppgs, surely Siggy's overall 4.8ppg is misleadingly low as:

1) Of course 70*38=2660, sure, but many of the points we gain during the season are from transfers and some bench rotation (plus we have DGWs, which genuinely add extra games if we assume we avoid blanks through good management, plus the chips). So in practice, if you were using this method, you'll surely 'need' less than the average of 6.3 in the way you calculated it.

2) We also own players for good fixture runs and not for others. You'd hope to use transfers to own Siggy for a good run(s) - like perhaps the opening few weeks of this season - and not his average runs; meaning you'd expect to get more than his 'average ppg' out of him when in your team (if indeed ppg can be assumed to be a good predictor of future performance, which is another conversation entirely ).

3) If you think it is genuinely realistic for a collection of players to hit the benchmarks producing, say, a 70ppw team - personally I would be very sceptical - then presumably you should take into account that some players would overperform the 'average benchmark' required to be in this 70ppw squad - the 'benchmark delta' won't be spread evenly across all 11 outfielders. The implication being that the team producing 70ppg would contain a number of players that underperformed their own benchmark but were still in this 'optimal' team. I.e., having a ppg below a benchmark does not preclude a player from being in the 70ppw team.

4) The season isn't linear in terms of squad price, and therefore points potential is not uniform each week - so hoping for 70 points in GW1 is hugely different to expecting 70 points in GW38 (theoretically speaking). Ergo, if working from a ppg standpoint, bigger squad scores will naturally come towards the end of the season and so at the start of the season you would require a lower ppg from Siggy compared to a similarly priced player come GW30+. Equivalently, if you were looking to simplify and attain a uniform 70 points every week, then your players would require a lower ppg in later weeks (lower than any 'required projected average ppg' you might calculate at the start of the season from the initial 100m budget, that is).

That wasn't meant to be an argument for picking Siggy incidentally (though I do like him), but just for attempted illustration. Hope that it all makes some semblance of sense.

[itslikebrandnew]

I'll post my 59.5ppw team in my rmt shortly for people to pick through.

[Jameselaprendi]

2) We also own players for good fixture runs and not for others. You'd hope to use transfers to own Siggy for a good run(s) - like perhaps the opening few weeks of this season - and not his average runs; meaning you'd expect to get more than his 'average ppg' out of him when in your team (if indeed ppg can be assumed to be a good predictor of future performance, which is another conversation entirely ).

This articulates nicely a couple of things I was going to attempt to ask

ppg is obviously a useful metric in any decision making framework - but I don’t always see it used in a way that’s aligned to how the game is played and how points are accumulated...

Some players are consistent, some streaky and some hit sporadic hauls

It would be useful start to analyse a seasons’ worth of data by player by match to identify patterns and start to classify different players or types of players. As an example, I like the idea of a heatmap 4-6 weeks rolling average alongside single match scores. Perhaps aggregatable to team level and filterable to different points types (goal, assist, bonus, clean sheet, etc). Need to build in fixture classification too, v top 6 v bottom 6, etc.

Sounds like a monstrosity but there’s 6 weeks to kill

And to contrast the above with ‘points’ based on xG

If someone can tell me where I can find the data...