Patrician's Crystal Ball

[Bao]

Hi Patrician,

Could you give me a favour? I need to replace Deeney by a FWD maximum 7.5 for only GW35. After that I will change him to Firmino for -4. Could you suggest any players have highest expected point? I am considering Rashford (eve MCI) or Wilson (FUL) but cannot decide, already have Jimenez.

Thanks.

[Patrician]

That is a close call Bao. I do think those two are the best options at that price point, and my model predicts 5.9 points for Wilson, and 5.5 for Rashford. I would lean towards Rashford for two chances to haul. Will he play both though?

[pokeface]

Your rival is captaining Sterling this week I assume then? So you’ll go Aguero for security?

Edit: What’s the big reason as to why Sterling isn’t higher in your model for these gameweeks? His expected minutes? Quite surprised as he has more points than Aguero overall.

[Patrician]

He went Son (c), I went Aguero (c) and I also played my bench boost. This is a key gameweek, and I need to outscore him by a lot.

Yes, I am treating Sterling and Aguero differently. I suspect Sterling will be benched for at least one game, there are just so many options in that midfield. Aguero I expect to start every game, and come off after the game is won (hopefully with points in the bank). I am not that fixated on the comparison though. For my team it comes down to Sterling vs Salah. I prefer the Salah run in, security of starts and form.

[Turd Ferguson]

He went Son (c), I went Aguero (c) and I also played my bench boost. This is a key gameweek, and I need to outscore him by a lot.

Yes, I am treating Sterling and Aguero differently. I suspect Sterling will be benched for at least one game, there are just so many options in that midfield. Aguero I expect to start every game, and come off after the game is won (hopefully with points in the bank). I am not that fixated on the comparison though. For my team it comes down to Sterling vs Salah. I prefer the Salah run in, security of starts and form.

You really think Salah's form is that much better than Sterling's? Sterling just nearly scored a hat trick against Tottenham in the CL. Salah hasn't had an xG over 1.0 for 12 games. Most of those games were an xG of less than .5.

[Patrician]

I find with Sterling that he is very annoying to own. Rather than a regular stream of points, you get big hauls separated by two or three games of nothing including the odd benching. Salah has provided an attacking returns in 5 of the last 6 games, and with HUD, new, WOL (all of whom have nothing much to play for) I fancy he could go on a big end of the season.

I might need to do a Sterling study and try and find if there is a common denominator for the matches where Sterling gets a big haul.

[pokeface]

I had planned to ditch Sterling this week and bring in Salah for Huddersfield, but then Deeney got sent off so I lost my free transfer, so at the moment it looks like I might be going Salahless. Will see what Sterling does (if anything) against United.

[fred1266]

I had planned to ditch Sterling this week and bring in Salah for Huddersfield, but then Deeney got sent off so I lost my free transfer, so at the moment it looks like I might be going Salahless. Will see what Sterling does (if anything) against United.

exact think happened to me so i wondering if i should save my FT this week then bring him in for GW 37

[pokeface]

I am going to see what happens with regard to the title race after this week is over if it’s going to go down to the final day I feel that Sterling might be a good potential captain pick in Gw38 as I imagine Brighton will be safe by then so have nothing to play for.

Logically right now I’m thinking, ditching Sterling for Salah for a hit is a bit risky, given City have Burnley next and they put 5 past them in the reverse fixture. I have Mané I’m covered as best I can be for a potential Salah haul. So I’ll hold for now.

[Turd Ferguson]

I find with Sterling that he is very annoying to own. Rather than a regular stream of points, you get big hauls separated by two or three games of nothing including the odd benching. Salah has provided an attacking returns in 5 of the last 6 games, and with HUD, new, WOL (all of whom have nothing much to play for) I fancy he could go on a big end of the season.

I might need to do a Sterling study and try and find if there is a common denominator for the matches where Sterling gets a big haul.

I took a cursory look at this when I was deciding whether to have him in my wildcard. There's a definite home bias. Games with multiple attacking returns:

Bri (H)
Sou (H)
WH (A)
Ars (H)
Che (H)
Wat (H)
Cry (A)

City's xG is meteoric at home though, so it's not a surprise that Sterling has more hauls at home. On the road, they're not all that much better than the other top 6 teams. It would be interesting if someone did a deeper dive and found that he tends to haul against teams with a poor left back or poor defensive midfielders, etc. My guess is that there's not a huge pattern, he just tends to haul when the City attack plays well and he's on the receiving end of some tap ins as a result.

Ultimately for me, Salah isn't hitting the heights of last season. Once City went out of the CL, it reduced the rotation pressure City will face and I backed Sterling to outscore Salah over the remainder of the season (although blanking at home against Tottenham may have already tilted those odds a bit against me).

[Pirlo's Beard]

he just tends to haul when the City attack plays well and he's on the receiving end of some tap ins as a result.

This is the frustrating crux of it I feel. He'll end up in the right place at the right time quite often but good luck predicting exactly where or when.

[pokeface]

Patrician, how does Laporte compare to VVD for the final 3 games in your model? VVD seems more of a goal threat to me but I also expect 3 City cleansheets.

[Patrician]

Yes, I forecast VVD ahead of Laporte. Greater goal threat and easier fixtures.


Sent from my iPad using Tapatalk

[Patrician]

GW35 Review: 91pts, GWR 441,728, 2,175 pts, OR 11,984 BENCH BOOST

A small red arrow in the grand scheme of things, but a decent result considering I was up against many teams who had wild carded into this double gameweek, and my squad held its own. It was as good a spot as any to chuck in my bench boost. Most importantly, my rival had a horror show scoring just 57 points and decided not to play his bench boost. Tough shit for him, as his bench scored 37 points (Ryan, Digne, Barnes). This swing of fortunes put me 15 points ahead overall.

GW36 Review: 104pts, GWR 244,532, 2,279 pts, OR 8,783

First 100 pointer of the season and a good time to do it, small green arrow only as lots of key players did well. Salah (c) plus TAA and Robertson got the week off the a flying start, ably supported by Wilson and Jimenez. I hope that the last few gameweeks reward my approach to the chips. My rival scored ok with 88 points (and left another 18 points on the bench) keeping my lead within reach but I am now 30 points ahead with two games to play. He brought in Aguero who I have, eliminating a differential (good for me, differentials = variance). This means it is now quite easy for me to play spoiler tactics and minimse the differentials between our teams.

Key differentials are

My Salah, Haz, Pog vs his Mane, Sterling, Fraser
My Jimnez vs his 9M FWD (Firmino, injured will probably bring in Vardy)
My TAA vs his Digne
He has his Bench Boost to come, so the lead is really more like 15 points.

The main worry is Mane and Sterling, both of whom are capable of a monster score that could eliminate the gap. I have two free transfers. What to do about Pogba? He has Huddersfield, but seems to have checked out, could even be dropped for his bad attitude now 4th is looking unlikely. I think on balance I expect Mane + Sterling to outscore Salah + Pogba in the final two games so I think those two transfers are the best bet for both overall points potential and to reduce the risk of being caught. The problem is if his remaining differentials / bench boost do well in GW37, then I would have very little room to manouver in response for GW38.

Decisions, decisions.

GW36.JPG

[Patrician]

Copying from another thread for later reference...

Doing a quick scan, here is a benefit of hindsight list of key decisions, that I think were somewhat predictable. When analysing my own season, I took advantage of a few of these, and lost out on a lot of points with the ones I missed. I am especially disappointed about missing most of the early Wilson/Fraser points (20ish points), Salah 16-18 (50ish), the Son run (40ish) , and the Vardy Bounce (20ish). Thats about 130 points which would have me in or around the top 100. I am disappointed because in all those cases I went through the thought process but prioritised some other decision.


New Manager Bounces

Pogba - GW18-26 (especially 18-20) HIT
Vardy - GW28-36 MISSED

Easy Fixture Runs

Son - GW13 to 25 (especially 16-21) MISSED
Aguero - GW2-11 HIT
Salah - GW 9-23 MISSED
Auba - GW6-24 MISSED
Anderson GW11-19 MISSED
Wilson/Fraser - GW1-15 MISSED and GW30-36 HIT

Double Gameweek Exploit

Aguero GW23-32 HIT

Identify early and keep (key enablers)

Hazard, especially GW3-8, GW16-19, GW30-33 (best value premium I think, and fixture proof ) HIT
Jimenez MISSED
W-B HIT
TAA when fit HIT
Doherty HIT

[Patrician]

When I look at my season like that, it is clear to me now how important fixtures and new manager bounces can be. Key lesson learned.

[Fuzzy]

I know it’s spurious cause-and-effect, however it’s interesting that the 2 examples you cite of new manager bounce have some common (not particularly endearing) personality traits!

[Aldershot Rejects]

New Manager Bounces

Pogba - GW18-26 (especially 18-20) HIT
Vardy - GW28-36 MISSED

Of course, with all these things that are players where there hasn't been a new manager bounce, but having also caught Pogba and missed Vardy, I am frustrated by the Vary miss - especially the early weeks before the chip madness set in. I guess the key, as always, is nailing who will get the bounce and given their fractious relationship with their previous manager and the character of the new guy coming in, these two were probably easier to spot than most.

I guess the big consolation for me vis-a-vis Vardy is that my forwards have generally scored pretty well over the period he has been doing well.

[pokeface]

I’m a bit mixed about this. And it’s something that is maybe obvious but has burnt me a bit this season.

The fact is, you are never ever going to be able to jump on every single bandwagon/in form player.

I think it’s important to look at why you missed them.

For me, for some reason or other, I felt it very important to get Vardy/Maddison in for GW 10-18 as they had a great fixture run. And the players I decided to do ditch for them was Pogba/Wilson.

Pogba was a good ditch but Wilson then went onto get double digits over the next few weeks, whilst Vardy/Maddison did nothing until about GW18.

Anyway, main point being, I altered the structure of my team to try and get on a bandwagon early and it cost me hugely!

I think it’s difficult to definitively say if something was a hit or miss or not until you look at who you would have had to take out in order to make it happen.

[Joccki_10]

I think we are forgetting how important squad structure is at times a new valuable asset arises. That is one of the most difficult things to do when you look back at how ones season has panned out: how did your structure look like, what did it enable/disable?

[pokeface]

I think we are forgetting how important squad structure is at times a new valuable asset arises. That is one of the most difficult things to do when you look back at how ones season has panned out: how did your structure look like, what did it enable/disable?

Agree, I think that is what I was trying to say (poorly) above.

[Patrician]

I have a funny feeling my Salah and Pogba to Sterling and Mane transfers are going to backfire. One of his few remaining differentials is bloody Digne


Sent from my iPhone using Tapatalk

[Patrician]

GW37 Review: 47pts, GWR 3,352,61, 2,326 pts, OR 8,671

A very mediocre gameweek, that deilivered a tiny green arrow. My transfers were Salah and Pogba to Sterling and Mane, and I captained Aguero. These moves were designed to nullify the threat of my rival's differentials and protect my 30 point lead. The result was fine from that perspective, I now have a 34 point lead going into the final gameweek. He does have his bench boost, so I don't think I can breathe easy yet.

This is now quite a good example of something Dod blogged about, playing differently depending on your specific goal. When trying to beat a rival, especially late in the season, if you are a healthy number of points ahead, then the sensible play is to mirror your opponent and eliminate dangerous differentials. If you are a number of points behind you try to create dangerous differentials. You do this even if it reduces your expected points which harms overall rank potential. This week, had I been following my normal approach, I would have kept Salah, and captained Hazard (Hazard was forecast to score the most points by my model), which would have given me an extra 15 points.

As a Son owner, my transfer is pretty much decided for me. Son is leaving for sure, and most likely Bilva coming in. Just need to decide whether to play my Wolves players (liv) or my Southampton players (HUD)...

GW37.JPG

[Patrician]

Rivals Team

I can't see him closing 34 points even with a bench boost. His key differentials play MCI, liv, tot, CHE, Firmino fitness in question, Boruc may not play, Son to deal with.

Rival GW37.JPG

[dod]

I think you are absolutely safe. There's no way he is going to close the gap unless he hacks your FPL account.

[Patrician]

...and so it came to pass. I extended my lead by 30 odd points (which I guess proves that my 30 odd point lead was possible to breach). Finished the season just inside the top 7k, third top 10k finish in the last four seasons.

Feels bittersweet. Obviously top 10k is always a decent season, winning such a substantial bet is always great, 2,399 points is my highest ever, and I beat the majority of the “famous” managers interviewed by FFS.....but after being close to the top 1k around GW6, I do feel like I squandered a genuine chance of challenging to win the whole thing, and maybe I will never have such a good start again. Lots to dissect over the summer.


Sent from my iPad using Tapatalk

[bigcliff2]

Exactly the same total score as me...

[fred1266]

Congrats Pat and thanks for all the help and player suggestion throughout the season

[Patrician]

I find with Sterling that he is very annoying to own. Rather than a regular stream of points, you get big hauls separated by two or three games of nothing including the odd benching. Salah has provided an attacking returns in 5 of the last 6 games, and with HUD, new, WOL (all of whom have nothing much to play for) I fancy he could go on a big end of the season.

I might need to do a Sterling study and try and find if there is a common denominator for the matches where Sterling gets a big haul.

Thanks to FK for posting the big FPL data dump. I am going to do quite a bit of data crunching over the summer to try and identify principles for fixtures and captaincy. These are the two areas that I think could take my game to the next level.

I don't think it is as simple of bottom 6 at home. My hunch is that some players are flat track bullies and do well against bottom 6, others do best against mid table sides who open up defensively at home, for example.

Bringing it back to Sterling, I mentioned in the quote above that he seems to randomly come up with these big hauls and nothing in between. If we can predict those big hauls better, it would be a major edge.

Sterling has had 9 hauls of 10 points or more (amounting to 134 pts)

GW4 NEW 10pts
GW7 BHA 14pts
GW11 SOU 21pts
GW13 whu 16pts
GW23 hud 10pts
GW25 ARS 10pts
GW26 CHE 18pts
GW30 WAT 21pts
GW34 cry 14pts

By my own analysis (combo of chances and goals conceded), here are the best, average, and worst defences...

Best.... LIV, MCI, TOT, CHE, EVE, LEI, WLV
Average..... ARS, MUN, WHU, WAT, CPL, SOU, NEW, BHA
Worst.. BOU, BUR, HUD, FUL, CAR

What is most striking is that Sterlings hauls do not tend to come when he plays the worst defences on paper (1 haul out of 9, hud), but rather against the more average defences (7 out of 9, NEW, BHA, SOU, whu, ARS, WAT, cry). Then they are more likely at home (6 out of 9).

This is an unusual pattern compared with the other high scoring midfielders.

Callum Wilson is another interesting one. 5 of his 6 big hauls came away from home.

[Oxford NZ]

With the exception of Newcastle, Stirling only only scores when City score three or more goals. He did score all three in the 3-1 at home to Watford but I am going to look and see if he only scores when City are all ready winning and then gets his chances. I do not watch much City because their players pick them selves for FPL purposes, the eye test ( I believed) was therefore not required. I think i need to watch a few more matches.