Fantasy Premier League 25/26

[ROFLUTION]

It's a slavish xA/xG algorithm selection built off the spreadsheet I've been working on to be perfectly honest. There'll be things I definitely want to to ignore about it at times for sure - but I think that's about the best it's throwing up for now taking it at face value.

There's a chances it's way off, built on my own formulas and early season is always fraught with uncertainty. It's making a lot of assumptions based on last season and of course things change a lot over the summer.

For Palmer specifically we know he struggled to find the back of the net in the second half of the season and get many assists but he's still got the underlying numbers in all of this so that's why it's picking him out I suppose.

Kluivert doesn't do so well on this actually, he's behind his teammate Semenyo at the same price from open play and a few others but there's the hope he might get a pen here or there. It's also week 3 (away to Spurs) and 6 (away to Leeds) where Bournemouth are showing up well mostly, but that's based on assumptions that Spurs will continue to be poor defensively and Leeds will be similar to recent promoted sides.

The whole thing is definitely throwing up some oddities while largely looking solid. Who knows? Maybe they're strokes of genius but could also be a pile of turd. Like I said, will be ignoring it at times too, just got to know when.

Team will no doubt change a million times by the time we're underway anyway.

Didnt realize it was all stats-based. Makes the selection interesting, but also everyone I’d personally pick I’d probably also do the eye test for - like Kluivert. He’s great if he can replicate his golden period again, bit that depends so much on form. Dango is also an alternative.

Anyways, just to play devil’s advocate: if it’s a model/based on stats, there doesn’t seem to be a lot of common players with the teams I see FFScout / FantasyHub and the likes post where they do their 98% team rating (for projected points I assume).

 

[Sarni]

It's actually possible to have 4 premiums somehow with the team still looking fairly solid.

Kelleher
Dorgu AWB De Cuyper Konsa
Salah Saka Palmer Damsgaard Arias
Haaland

+ a garbage bench

 

[Pickle85]

It's actually possible to have 4 premiums somehow with the team still looking fairly solid.

Kelleher
Dorgu AWB De Cuyper Konsa
Salah Saka Palmer Damsgaard Arias
Haaland

+ a garbage bench

Thius is similar to what I'm on atm. Head says it's a terrible idea but just looking at the starting 11...

 

[Red Star One]

While nobody should really care, my spreadsheet/predictor is almost finished. It's come on nicely. Put way too many hours into it with barely any understanding of how to use spreadsheets when starting so I'm feeling quite proud in a way. A little self-taught project.

Needs a little cosmetic work and may well be a load of old rubbish but nearly there. Maybe I can get a fully office-based job now having learned my way around all this?

Spoiler: Team xG prediction page

Here I have the results and xG from recent games on the right hand-side. I'll need to copy and paste each week new scores in there as they happen. All the stuff in the middle can basically be ignored, just bunch of numbers doing some computations.

Pick the home team and away teams from drop down menus at the top and we get a bunch of predictions for xG they'd get if playing each other using various formulas that I came up with and was testing. Found the very first one, "last 12" to be the most accurate so that's what I'll be using.

Flaws/comments
No formula can ever be perfect, or even approaching it. Football is unpredicable in it's nature and it's part of why we love it. Can only try to get close.

I've no reliable way of assessing the promoted sides. Won't be picking any players from their teams initially anyway other than perhaps a cheap defender to be on my bench. For now I've had to assume they'll perform similarly to the relegated teams but that may not be the case. I've basically assumed Sunderland = Southampton for example, and used Southampton's results from last season to start with for them. As they start to play games we'll start to find out, but early season I might be overestimating how many goals teams will score against these promoted teams.

Teams change in general. Maybe United do a lot better and hopefully we do. Like the promoted sides, early season hard to predict until we actually get going we won't know. New data will start being added with each game we play until it's purely based on this upcoming season 12 weeks in.

Spoiler: Player points prediction page

Here we have a list of players in a team's squad in the left, their xG and xA per 90, and their defensive contribution points per start. This is all last season's data at the moment again which obviously needs updating as we go. Still need to fill in the position of every player and their defensive contribution stats on the very left, just wanted to get a working model up and running.

The middle part is where it all really happens. Stick the expected team xG for and against from the previous sheet along the top in the GW1F, GW1A, GW2F, GW2A etc. part and repeat for the next 7 weeks. I've decided that's about how long I want to look into the future when making transfer decisions. Everything else basically references that, and using the individual player xA, xG and defensive contribution stats it spits out some sort of estimated points tally for each player on the very right for the next 7 weeks.

Flaws/comments
New signings are difficult. Players' stats may well change quite drastically from last season if they signed for a different club. Players moving from clubs outside of the PL are particularly hard to assess at the very beginning. In this Arsenal example I've used Gyokeres' Champions League xG and xA rather than his Portuguese league stats but all I can do is guess. I've used Zubimendi's xA/xG from La Liga as I believe it's a similar standard of league to the PL, and I've used Partey's defensive contribtion points estimate for Zubimendi as he'll playing in a similar position. Everything will start filtering through a few weeks from now and I'll have a much better picture.

Goalkeepers - don't exactly know what I'm doing with them yet. Was going to just use the clean sheet element of their projected points but we all know they can get points from how many saves they make as well and that the goalie with the most clean sheets isn't always the best one to have. Unsure how to proceed here, and whether I want to flesh everything out fully for them as it would require quite a bit more work.

Defensive contribution stats - I've gone with points per start which is somewhat broad. I've not tried to assess which teams players are more or less likely to achieve these points against. Suppose I could try and nail that down but it's a lot of effort.

Subs/playing time - this completely assumes that every player plays a full 90 which of course doesn't happen, and subs are becoming more and more important in football, having increased impact over the years. Still, I'll be trying to pick players who get a lot of minutes and will be looking at that on the FPL website anyway. Nobody really wants someone who goes off after 65' every week in their team, or a player who only starts occasionally.

Penalties - the taker can change! It's why I've not actually included them in the points calculation part. That's just to take a mental note of each week and add a little to a players' projected xG in my head.

Bonus points - I've not included them at all! Wouldn't even know how to begin to predict that as there's so many ways players can earn BPS, it's different depending on position and they've tweaked how it's calculated from last season anyway. Suppose I could do an average bonus points per game or something and I might actually. Anything beyond that is too complcated for me.

It's ugly! - Need to put some colour coding in there now. Did toy around with that a bit just now but I couldn't quite get anything that looked nice and pretty. Can also collapse/hide the columns for a lot of that stuff in the middle once the data is in the sheet each week which would help a lot. Would end up looking more like this:

Learning my way around a spreadsheet and how to write formulas for some of these cells was an experience! What on earth is this?

That would have made no sense to me at all a few weeks ago, it's why I'm oddly proud. Used functions like VLOOKUP too having no idea what that was either before starting.

Will probably add some cumulative next 5 weeks, next 6 weeks and next 7 weeks prediction totals at the end as well as week-by-week to put more a bit more info on it all. That's easily done and I'd be referencing that a lot.

Well that was long, but like I said kind of pleased with all of it. Hopefully it helps me!

That's very ambitious! Thanks for sharing this and I'm curious if this helps you a lot. You had an insane result last year, so I'm sure it will be difficult to even repeat that!

 

[Chipper]

Didnt realize it was all stats-based. Makes the selection interesting, but also everyone I’d personally pick I’d probably also do the eye test for - like Kluivert. He’s great if he can replicate his golden period again, bit that depends so much on form. Dango is also an alternative.

Anyways, just to play devil’s advocate: if it’s a model/based on stats, there doesn’t seem to be a lot of common players with the teams I see FFScout / FantasyHub and the likes post where they do their 98% team rating (for projected points I assume).

The thing I put together definitely has flaws.

The big flaw is assuming a continuation of last season, specifically the last 12 games when it comes to teams and the season as a whole for players.

They're huge and quite honestly false assumptions and it should only improve as we go. It kind of thinks players will fill in for whoever they're replacing at first with little to no impact on results. Brentford and Lewis-Potter would be a good example. They've lost Mbeumo (0.24 xA/0.20 npxG per 90) and Norgaard (0.04/0.14). It's kind of assuming that the players they have remaining will somehow step up to fill the void. So Lewis-Potter with his 0.14/0.09 and all the other Brentford players get a little boost on the thing I've put together. They were quite big numbers for a defender anyway.

Football doesn't actually work like that, most would assume and probably correctly that Brentford will become a worse team. Improvement can happen to an extent for individuals, think when Rooney scored more the season after Ronaldo left, and I think Benzema did the same when Ronaldo left Real Madrid but it's not a thing you can say will happen. The most common effect of losing key players is that most players who remain would go backwards or stay about the same, that's just logical.

Ran my selection through these sites after your post and they certainly don't love my picks! FF Hub scored it an 85/100, saying it could improve it to a 92 if I sign up. I'm with FFScout this season and they gave it 335 projected points in the first 6 weeks. They hate Matheus Nunes due to minutes risk in particular. Tinkering around with what they like I can make a 355 Salah team fairly comfortably.

So to kick off with here's my scoreline guesses vs bookies, and that's the building block of any of these.

A few discrepencies there already. My thing hates the promoted sides, they're all pencilled in to perform like the relegated ones. Literally just substituted them in as if they were same teams.

Mine is keying on Forest at the end of last season big time for their opening game this season. In their last 6 or 7 home games they didn't produce much xG. 0.4, 0.7, 0.5, 0.5, 0.9, 1.3 (v Leicester which is poor) and 1.2. Brentford tightened up away. They also played each other and it was 0.9-1.3, pretty much what my model predicts again. Wouldn't necessarily expect that to continue in the real world though and the bookies perhaps rightly see it very differently.

Liverpool eased off at the end of last season otherwise it would be showing a more dominant prediction v Bournemouth, it's things like this showing up into this season on the thing I put together. Other than those it's not looking completely awful but enough to skew things at times.

For what it's worth, here are players of interest on my little model for the first 6 weeks vs FFScout's top ones in each position:

So there's a lot of shared players as well as the odd weird anomaly for me. Some of the strange ones on mine are definitely minutes risks, filter those out and it's getting closer to theirs.

Filter out the issues I mentioned above on my model and it's closer still. My team won't be picked purely on what my spreadsheet is telling me, only ever meant as a guide. Some of the unusual names it throws up are certainly intriguing though and might warrant further investigation. Lewis-Potter right at the top of mine perhaps due to some of those problems with my methology, but he's still okayish and reasonably prominent on Scout, just outside the players they really highlight. They have him 21st on their list of defenders which I suppose is top 25% considering 80-85 will start each week.

 

[Sarni]

Thius is similar to what I'm on atm. Head says it's a terrible idea but just looking at the starting 11...

I usually start my season trying to differentiate and avoid template, and end up well behind very early on. I had some pretty cool builds without Salah and Haaland too. Will probably change it a dozen times in the next 3 weeks anyway.

I like the £4.5m choices in defense and there’s a few £5.5m - £6m midfielders in there as well. In the meantime I’m not really fond of most strikers, what with Isak’s future being uncertain, United not actually having a striker and Arsenal/Liverpool trying to accommodate new players.

 

[RDCR07]

Surprised Gyorkeres still isn’t priced. Also if Isak moves to Liverpool, I’m going Isak over Salah and Gyokeres over Saka.

 

[paulscholes18]

Surprised Gyorkeres still isn’t priced. Also if Isak moves to Liverpool, I’m going Isak over Salah and Gyokeres over Saka.

He has been added now 9.0

Sanchez
Udogie Van Dijk De Cuyper
Salah Wirtz Palmer Kudus
Gyokeres Solanke Beto

Darlow, AWB, Anderson, reinildo

 

[DJ_21]

Does anyone do a paid league for this?

 

[ROFLUTION]

The thing I put together definitely has flaws.

The big flaw is assuming a continuation of last season, specifically the last 12 games when it comes to teams and the season as a whole for players.

They're huge and quite honestly false assumptions and it should only improve as we go. It kind of thinks players will fill in for whoever they're replacing at first with little to no impact on results. Brentford and Lewis-Potter would be a good example. They've lost Mbeumo (0.24 xA/0.20 npxG per 90) and Norgaard (0.04/0.14). It's kind of assuming that the players they have remaining will somehow step up to fill the void. So Lewis-Potter with his 0.14/0.09 and all the other Brentford players get a little boost on the thing I've put together. They were quite big numbers for a defender anyway.

Football doesn't actually work like that, most would assume and probably correctly that Brentford will become a worse team. Improvement can happen to an extent for individuals, think when Rooney scored more the season after Ronaldo left, and I think Benzema did the same when Ronaldo left Real Madrid but it's not a thing you can say will happen. The most common effect of losing key players is that most players who remain would go backwards or stay about the same, that's just logical.

Ran my selection through these sites after your post and they certainly don't love my picks! FF Hub scored it an 85/100, saying it could improve it to a 92 if I sign up. I'm with FFScout this season and they gave it 335 projected points in the first 6 weeks. They hate Matheus Nunes due to minutes risk in particular. Tinkering around with what they like I can make a 355 Salah team fairly comfortably.

So to kick off with here's my scoreline guesses vs bookies, and that's the building block of any of these.

A few discrepencies there already. My thing hates the promoted sides, they're all pencilled in to perform like the relegated ones. Literally just substituted them in as if they were same teams.

Mine is keying on Forest at the end of last season big time for their opening game this season. In their last 6 or 7 home games they didn't produce much xG. 0.4, 0.7, 0.5, 0.5, 0.9, 1.3 (v Leicester which is poor) and 1.2. Brentford tightened up away. They also played each other and it was 0.9-1.3, pretty much what my model predicts again. Wouldn't necessarily expect that to continue in the real world though and the bookies perhaps rightly see it very differently.

Liverpool eased off at the end of last season otherwise it would be showing a more dominant prediction v Bournemouth, it's things like this showing up into this season on the thing I put together. Other than those it's not looking completely awful but enough to skew things at times.

For what it's worth, here are players of interest on my little model for the first 6 weeks vs FFScout's top ones in each position:

So there's a lot of shared players as well as the odd weird anomaly for me. Some of the strange ones on mine are definitely minutes risks, filter those out and it's getting closer to theirs.

Filter out the issues I mentioned above on my model and it's closer still. My team won't be picked purely on what my spreadsheet is telling me, only ever meant as a guide. Some of the unusual names it throws up are certainly intriguing though and might warrant further investigation. Lewis-Potter right at the top of mine perhaps due to some of those problems with my methology, but he's still okayish and reasonably prominent on Scout, just outside the players they really highlight. They have him 21st on their list of defenders which I suppose is top 25% considering 80-85 will start each week.

Interesting walk-through. In theory you could also take the 3 best sites and your model and do a wisdom of the crowd calculation of the most recommended players for projected points. That’s how pinnacle set their market leading odds, which the market then use as a baseline. (One that’s worked for decades for bookies being profitable). They basically take a selected few experts and ask them to put odds on a game and the average of those expert opinions are what becomes the odds.

 

[IsoJoe]

GK-Raya, Sels
Defenders- Virgil,Murillo,Guehi, Vuskovic and Alt Nouri
Midfielders - Caicedo , Rice, Salah , Kudus, Bruno G
Forwards - Isak,Gyokeres and my dark hourse ....Kroupi Jr

 

[WayneKing]

Here's my first draft. It is basically dirt goaltending and one premium defender Then 2 premium mids and Watkins/Bowen as forwards. Rogers is untouchable for me. I could be convinced to get rid of everyone else. And I have a hunch about Enzo. Enzo had hype last year and disappointed. But he has slowly rounded out his game and is a good sleeper pick imo.

Kelleher/Kinsky

Gvardiol/WBS/Guehi/Konsa/Radon

Enzo/Palmer/Ndiaye/Salah/Rogers

Watkins/Bowen/Fullkrug

I really don't like Palmer but I just want to cover him off. I could drop him and get Haaland with some Mid, D and F downgrades.

Kelleher/Kinsky

Murillo/WBS/Guehi/Konsa/Radon

Enzo/Anderson/Ndiaye/Salah/Rogers

Watkins/Haaland/Piroe

 

[WayneKing]

Cunha, Bruno, Hojlund and Mbeumo in before he changes to United.

Smarts, 4 United attackers for when we smash teams every week

(tries to add drool emoticon in, forgetting they inexplicably deleted it ages ago)

You're joking but Cunha, Mbeumo and Bruno are capable of tearing it up in this league.

Cunha and Mbeumo are zero percent owned in the expert template right now. Which I think is way overdoing the ManU skepticism. Mbeumo was a must-have asset last year. Cunha wasn't far behind. Wouldn't it be funny if they just started slaughtering teams right out of the gate.

 

[Andycoleno9]

What do we experts () think about 2 things;
1. Cunha or Mbuemo? Or none of them?
2. Is Foden worth a shot? Trying to find someone from City and can't decide. Surely Foden will be a key player this year when KDB is out.

 

[JP77]

What do we experts () think about 2 things;
1. Cunha or Mbuemo? Or none of them?
2. Is Foden worth a shot? Trying to find someone from City and can't decide. Surely Foden will be a key player this year when KDB is out.

I would honestly avoid any City midfielder to start the season. Just too hard to predict their starting lineup and it’s even tough when players have a run of brilliant games as Pep loves to throw a curveball. Just way too much of a risk imo.

 

[RaddyRed]

What do we experts () think about 2 things;
1. Cunha or Mbuemo? Or none of them?
2. Is Foden worth a shot? Trying to find someone from City and can't decide. Surely Foden will be a key player this year when KDB is out.

1. I've gone with both, but I did finish around 2.5m last season

2. City now have quite a big squad so the famous roulette could get even worse. I'm waiting a few weeks in to see if he has a favoured team.

 

[Andycoleno9]

I would honestly avoid any City midfielder to start the season. Just too hard to predict their starting lineup and it’s even tough when players have a run of brilliant games as Pep loves to throw a curveball. Just way too much of a risk imo.

1. I've gone with both, but I did finish around 2.5m last season

2. City now have quite a big squad so the famous roulette could get even worse. I'm waiting a few weeks in to see if he has a favoured team.

Gyokeres made me to rethink my team so i have 8,5 mil spot in midfield and i am trying to find a right midfielder.
I agree about avoiding City but United is also a big risk. I don't expect some crazy return from Cunha and Mbuemo.

Meh, feck Gyokeres for now. I am sticking with Solanke/Saka combo. For now....

 

[CassiusClaymore]

What do we experts () think about 2 things;
1. Cunha or Mbuemo? Or none of them?
2. Is Foden worth a shot? Trying to find someone from City and can't decide. Surely Foden will be a key player this year when KDB is out.

None
No

 

[ROFLUTION]

Struggling with going with Saka or Gyokeres - or just Arsenal defence in general. The fixtures from 3-6 are just too shite overall. (it's all the tough teams)

Unless it's part of a strategy somehow, to easily get them out again in GW3 or bench boost in GW1 or 2 somehow and then wildcard in GW3, but even GW3 just seems a bad week for a wildcard as that's where they face Liverpool and you'd wildcard to get the Liverpool performers in by then.

Only consideration currently is to have Gyokeres for the first 2 GWs and then switch to another forward after their Leeds game which could be tasty. Then go Watkins, Joao Pedro, Isak (+1.5), but I'm not a big fan of booking in transfers.

 

[ROFLUTION]

Not settled on a squad yet, but the main players I have mostly prioritized in my drafts are:

Evanilson: Great fixtures after GW1
Beto: Great fixtures and you'd want a good bench option too
Amad: Great value. The potential for many points is there over a full season. He has more playmates now, so maybe there's a synergy-effect.
Dorgu+Wan Bissaka combo: Their fixtures just fit each other quite nicely, but it depends on how safe Dorgu is for minutes. I think he's the preferred choice.
Milenkovic: Has the fixtures, would get a lot of extra points. 5.5 is a bit steep though, but if Forest stinks, he could become Konate in GW4.
Gvardiol/Ait Nouri: Really tempted by Ait Nouri, but rotation fears puts me a bit off. Ideally you'd want at least 1 defender in the 6.0 bracket though, so he could be subbed if he doesn't start regularly.
Marmoush: The easier route into City (if there is one), if Haaland isn't picked.
Rogers: On penalties, and great fixtures. Even their opener against Newcastle has been high-goal scoring, so you never know. Team will pass to him more too or he'll be more pivotal, I think.

I sort of want Cucurella, Van Dijk (after GW3) and a Liverpool midfielder/striker too.
Overall I think I'll aim for a wide squad with about 2 or 3 decent subs and only 1 of Salah/Haaland. Basically to not get too hung up on transfers. Seemed like this worked last year, and you can fit in a lot of good players.

Spurs are the joker here, and I've had Solanke and Kudus in my team, but I'd max have 1. Not sure I'd want too many joker-picks you might want to sub out early if they don't fire. Rather go the other way and transfer them in if they do look great.

Currently it's something like this:

Sels
Milenkovic - Wan Bissaka - Ait Nouri - Cucurella
Salah - Rogers - Kudus - Marmoush
Gyokeres - Beto

Bench: Evanilson, Amad, Dorgu (all coming in GW2) + Kinsky

 

[Andycoleno9]

Getting closer and closer.

Sanchez - Celski have good fixtures. 5 mil is a fair price.
Darlow - 4 mil
Aina - good price for NFO defender
Virgil - zero rotation risk (unlike Frimpong) and if CS, he is in bonus always
Awb - cheap
Dorgu - worth a shot
Bijol - 4 mil starter
Salah - you can't go without him
Palmer - same as Salah
Rogers - underrated
Kudus - risky but for that price worth a risk
Baleba - cheap(est) bench option
Watkins - i always have Watkins so why change
Gyokeres (☆Solanke, Pedro)
Beto - cheap starter

☆ maybe i will start will Solanke or Pedro first 3 games and then switch to Gyokeres.

 

[Jaap]

First draft was pretty quick:

Kelleher
Frimpong Muñoz Romero
Kudos Palmer Mbeumo Cherki
João Pedro Beto Mateta​

Subs all around 4.0s and 4.5s.

 

[Andycoleno9]

Not settled on a squad yet, but the main players I have mostly prioritized in my drafts are:

Evanilson: Great fixtures after GW1
Beto: Great fixtures and you'd want a good bench option too
Amad: Great value. The potential for many points is there over a full season. He has more playmates now, so maybe there's a synergy-effect.
Dorgu+Wan Bissaka combo: Their fixtures just fit each other quite nicely, but it depends on how safe Dorgu is for minutes. I think he's the preferred choice.
Milenkovic: Has the fixtures, would get a lot of extra points. 5.5 is a bit steep though, but if Forest stinks, he could become Konate in GW4.
Gvardiol/Ait Nouri: Really tempted by Ait Nouri, but rotation fears puts me a bit off. Ideally you'd want at least 1 defender in the 6.0 bracket though, so he could be subbed if he doesn't start regularly.
Marmoush: The easier route into City (if there is one), if Haaland isn't picked.
Rogers: On penalties, and great fixtures. Even their opener against Newcastle has been high-goal scoring, so you never know. Team will pass to him more too or he'll be more pivotal, I think.

I sort of want Cucurella, Van Dijk (after GW3) and a Liverpool midfielder/striker too.
Overall I think I'll aim for a wide squad with about 2 or 3 decent subs and only 1 of Salah/Haaland. Basically to not get too hung up on transfers. Seemed like this worked last year, and you can fit in a lot of good players.

Spurs are the joker here, and I've had Solanke and Kudus in my team, but I'd max have 1. Not sure I'd want too many joker-picks you might want to sub out early if they don't fire. Rather go the other way and transfer them in if they do look great.

Currently it's something like this:

Sels
Milenkovic - Wan Bissaka - Ait Nouri - Cucurella
Salah - Rogers - Kudus - Marmoush
Gyokeres - Beto

Bench: Evanilson, Amad, Dorgu (all coming in GW2) + Kinsky

Beware of Marmoush. They are stacked in midfield. He will more play in attack so rotate with Haaland.

 

[ROFLUTION]

Beware of Marmoush. They arevstacked in midfield. He will more play in attack so rotate with Haaland.

I don’t think Haaland will be rotated, not in the beginning of the season.

Pep usually gives his players a year to settle in, but Marmoush has been played constantly since his arrival, also when Haaland was injury free. I think he’s a core of his team and the safest bet for an attacking City player, albeit it always comes with a minutes risk with Pep. He likes to keep his players on their toes too, if they don’t perform.

If rotated, he’s an easy way to Wirtz/Cunha/Mbeumo, etc.

 

[Andycoleno9]

I don’t think Haaland will be rotated, not in the beginning of the season.

Pep usually gives his players a year to settle in, but Marmoush has been played constantly since his arrival, also when Haaland was injury free. I think he’s a core of his team and the safest bet for an attacking City player, albeit it always comes with a minutes risk with Pep. He likes to keep his players on their toes too, if they don’t perform.

If rotated, he’s an easy way to Wirtz/Cunha/Mbeumo, etc.

On FIFA WC Marmoush started 2 games and both as a striker. I want Marmoush also but that rotation risk is too much for me.

 

[ROFLUTION]

On FIFA WC Marmoush started 2 games and both as a striker. I want Marmoush also but that rotation risk is too much for me.

I don't think the FIFA WC is anything to go by at all. It's for money, and Pep knows that he needs to save the player's energy and therefore rotated - specifically your biggest player Haaland you don't want to burn out off season. Every player got about the same amount of game time overall, except for maybe a few new players he'd want an extra look at (Cherki, Ait Nouri). But still that was maybe only 10-25% more game time than everyone else.

 

[Chipper]

Interesting walk-through. In theory you could also take the 3 best sites and your model and do a wisdom of the crowd calculation of the most recommended players for projected points. That’s how pinnacle set their market leading odds, which the market then use as a baseline. (One that’s worked for decades for bookies being profitable). They basically take a selected few experts and ask them to put odds on a game and the average of those expert opinions are what becomes the odds.

Certainly an idea to use a few different sites and consult them all plus my own numbers. It does go against the orginal intention of what I put together in a way but that's not necessarily fixed.

Original goal was a free points predictor that looked ahead a few weeks and was solid so thought I'd have a go at making one. A big part of the inspiration is because I was using this one which was everything I wanted but they stopped updating around Feb or March and it doesn't look like it's coming back - https://fplanalyser.com/prices

FPL Review's free one only goes 3 weeks in advance and I didn't want to pay.

There's a couple of other free ones I know of that I don't think are fantastic but have their uses:
https://managerbotten.dk/player-picks - FPL Team were using these guys data at one point but I think they've switched
https://www.fantasyfootballpundit.com/fpl-points-predictor/ - not updated for the new season yet?

I kind of was digging into all of these at time last season for that wisdom of the crowds effect.

In the end I paid for FFScout this summer when the idea was always to keep it free! It was for their data rather than their model so I guess I could even pay for a couple more at this point.

These guys possibly look interesting and I think it's all free but am not sure - https://www.elevenify.com/p/premier-league-data The team they run does very well every year, played 6 seasons and their worst result is top 30k. Team goal predictions up for the first 6 weeks which certainly could be of use to me plus a few other things.

 

[LukeTheRed]

Getting closer and closer.

Sanchez - Celski have good fixtures. 5 mil is a fair price.
Darlow - 4 mil
Aina - good price for NFO defender
Virgil - zero rotation risk (unlike Frimpong) and if CS, he is in bonus always
Awb - cheap
Dorgu - worth a shot
Bijol - 4 mil starter
Salah - you can't go without him
Palmer - same as Salah
Rogers - underrated
Kudus - risky but for that price worth a risk
Baleba - cheap(est) bench option
Watkins - i always have Watkins so why change
Gyokeres (☆Solanke, Pedro)
Beto - cheap starter

☆ maybe i will start will Solanke or Pedro first 3 games and then switch to Gyokeres.

Will Beto actually start? They spent big on Barry from Villarreal

 

[Maluco]

Certainly an idea to use a few different sites and consult them all plus my own numbers. It does go against the orginal intention of what I put together in a way but that's not necessarily fixed.

Original goal was a free points predictor that looked ahead a few weeks and was solid so thought I'd have a go at making one. A big part of the inspiration is because I was using this one which was everything I wanted but they stopped updating around Feb or March and it doesn't look like it's coming back - https://fplanalyser.com/prices

FPL Review's free one only goes 3 weeks in advance and I didn't want to pay.

There's a couple of other free ones I know of that I don't think are fantastic but have their uses:
https://managerbotten.dk/player-picks - FPL Team were using these guys data at one point but I think they've switched
https://www.fantasyfootballpundit.com/fpl-points-predictor/ - not updated for the new season yet?

I kind of was digging into all of these at time last season for that wisdom of the crowds effect.

In the end I paid for FFScout this summer when the idea was always to keep it free! It was for their data rather than their model so I guess I could even pay for a couple more at this point.

These guys possibly look interesting and I think it's all free but am not sure - https://www.elevenify.com/p/premier-league-data The team they run does very well every year, played 6 seasons and their worst result is top 30k. Team goal predictions up for the first 6 weeks which certainly could be of use to me plus a few other things.

When looking at that overview of key data, the right combo is Haaland/Palmer rather than Salah/Palmer. And you could even go attack heavy with Haaland/Watkins/Isak.

I wonder if any of the influencers will start posting Haaland teams

 

[Andycoleno9]

Will Beto actually start? They spent big on Barry from Villarreal

He finished season as no1 striker and in a solid form. I doubt he will lose that spot straight from the start.

 

[WayneKing]

What do we experts () think about 2 things;
1. Cunha or Mbuemo? Or none of them?
2. Is Foden worth a shot? Trying to find someone from City and can't decide. Surely Foden will be a key player this year when KDB is out.

Neither is worth picking yet apparently

But you know who us experts should be considering ? You know who's the top defensive contribution mid fielder ?

Casimiro And he's only 5.5 million. These defcon points add up.

 

[Swiss_Red89]

Granit Xhaka for 5.0 is tempting. Is this only me because he is my National Teams Captain?

I'm currently on a threemium Draft:

Kelleher
AWB VanDeVen Dorgu
Salah Palmer Xhaka Rogers Kudus
Pedro Haaland

Bench: 4.0 keeper, Konsa, Esteve, 4.5 striker

Think Dorgu could become a must have for 4.5. Not sold on our offensive options yet. We dont even know who will play where. Bit of a wait and watch job imo.

 

[RobinLFC]

Feels harder than last years to get a good team together if you want Salah, Palmer, Watkins and Isak in your team. That's even omitting Haaland and Saka at the moment.

Surprised (but then again not really) by the lack of Wirtz in the drafts in this thread. Think he'll put up very decent numbers and is "only" a £8.5m at the moment.

Might also be sensible to lock Isak in at £10.5m before he transfers to Liverpool (big if but still) - I assume you can just keep your team then if you don't already have 3 Liverpool players?

 

[Pickle85]

Feels harder than last years to get a good team together if you want Salah, Palmer, Watkins and Isak in your team. That's even omitting Haaland and Saka at the moment.

Surprised (but then again not really) by the lack of Wirtz in the drafts in this thread. Think he'll put up very decent numbers and is "only" a £8.5m at the moment.

Might also be sensible to lock Isak in at £10.5m before he transfers to Liverpool (big if but still) - I assume you can just keep your team then if you don't already have 3 Liverpool players?

Who on EARTH would want three Liverpool players, let alone four?!

 

[Chipper]

Knocked this together for what it's worth, think there's a few things we can read into it.

This thing still bugging me.

Defensive contribution points are going to have at least some significance, and right now I'm just looking at it all on a very basic level. Caicedo got a lot of these points, Tarkwoski too etc. so take a look at them.

Fixtures, and who you're playing specifically has to come into it. Your opponent's style of play or how good (bad) they are at certain things, and I'm wondering if there's some sort of way of trying to get to the bottom of everyting.

Why was it so hard to earn these DEFCON points against Chelsea? Why was it easy against Spurs?

There's so many variables to it all. We've obviously got lists of the stats from FBref but trying to make sense of it all and why these things might have happened is so complex.

Feel there's got to be a way of answering all these questions and turning that into some sort of future predictions but it would take so much time and effort.

 

[Chipper]

Thinking on, I guess you don't necessarily need the why.

Perhaps some sort of matrix table could work things out.

Something like "player X makes this number of clearances on average, team Y forces their oppoents into 10% more clearances than the average team or 7% less clearances than the average team" etc.

Then repeat that for each of the stats. Maybe that works on some level to come up with a prediction of sorts.

 

[Chipper]

Starting the deep dive.

Defenders made a total of 10534 defensive actions at home last season.
Defenders made a total of 11603 defensive actions away last season.
Home/Away split = 47.6%/52.4%

Midfielders made a total of 15723 defensive actions at home last season.
Midfielders made a total of 15983 defensive actions away last season.
Home/Away split = 49.6%/50.4%

So that's interesting it's own way. Where the game is taking place doesn't seem to matter too much overall, although that could change team by team, player by player. Still, for shorthand purposes I wouldn't be too bothered about differnetiating based on that.

The slight difference between home/away repeating when looking at some of the top DC players, but it's not major at all and quite a few of the top guys actually did more defensive work at home.

Happy to say Home or Away doesn't really look like it makes much of a difference to concern ourselves with.

 

[Lennon7]

I feel like I’ve got about as good as I’m gonna get

Henderson/Sanchez
Dorgu - Virgil - Gabriel
Mbeumo - Palmer - Caicedo - Wirtz - Salah
Wissa - Mateta

Bench:
Marc Guiu (just loaned to Sunderland)
Tuanzebe
Huggins

I had Haaland and Bowen up top before with Mitoma instead of Salah but honestly thing that’s better

As others have said though deffo getting Isak when we know what’s going on

 

[Chipper]

More of a home/away split when looking at actual DC points that would have been awarded instead of just actions. Only for defenders though. It's still not massive.

Defenders home = 550 DC points
Defenders away = 754 DC points.
Home/Away split = 42.2%/57.8%

Midfielders home = 572 DC points
Midfielders away = 604 DC points
Home/Away split = 48.6%/51.4%

 

[Chipper]

Got a little working model of trying to predict DC points, not sure it's worth proceeding or if it's accurate.

All very basic stuff.

So Saliba would have earned 12 DC points last season, or 0.37 per 90.

Feeding in his fixtures and last season's data I got this for the first 7 gameweeks as predictions (DC1-DC7 columns):

It's 2.53 over the first 7 weeks at 0.36 per 90. There's variations in there, and some games are better prospects than others but as that's so close to his average anyway I'm thinking I might have just stuck with that.

Maybe that's just the way Arsenal's fixtures are though. Could look more worth it for other teams/players. His best prospect not surprisingly in GW3 (DC3) vs Liverpool.