To see less ads Register or Login ----- Daily Fantasy Sports games 18+

R_NZ FPL Blog

A forum for comment and discussion on Fantasy PremierLeague.com (FPL) Teams. Post your Rate My Team (RMT) messages here!
Post Reply
FranckKessie
Grumpy Old Man
Posts: 1548
Joined: 21 Aug 2017, 21:34

Re: R_NZ RMT Blog

Post by FranckKessie »

Ruth_NZ wrote: 15 May 2018, 08:49 Principle #1 Structure for value and balance squad spend more in favour of defenders; don't chase attacking hauls to the detriment of overall performance.
Principle #2 Collect the 'easy points' as a priority; don't throw them away in the quest for 'hard' points.
Structure for value gets a plus one from here. I don't agree that beyond Alonso,Valencia and Azpilicueta that there were many predictable high-scorers. Of course you would expect the rest to deliver a decent amount of points due to the teams that they play for but most of them came at a high price or there were better options within the team at the start of the season imo.

Collecting the 'easy points' I am not sure I get. Who are the easy points? Is it more or less choose highly owned players? I.e. risk management.

User avatar
BobMem
Grumpy Old Man
Posts: 4222
Joined: 24 Dec 2010, 08:54
FS Record: The Graduates - FISO 5AS H2H Champions 13/14
FISO King of Rock and Roll - Elimination game winner 2015/16

Re: R_NZ RMT Blog

Post by BobMem »

I don't want to get in the way of this defender chat, because I think there's still plenty of mileage in it, but I'm planning a longish post on chip usage. Are you lot going to get offended if I put that here, or are all musings welcome?

User avatar
Aldershot Rejects
Dumbledore
Posts: 9586
Joined: 03 Aug 2011, 17:15
Location: Kent
FS Record: 5th Metro (2010-11); 146 - Sky (2015-16); 218 - Sky (2014-15); 386 - Sky (2020-21); 636 - FPL (2017-18); last 16 Sky Cup (2018-19)

Re: R_NZ RMT Blog

Post by Aldershot Rejects »

BobMem wrote: 15 May 2018, 11:18 I don't want to get in the way of this defender chat, because I think there's still plenty of mileage in it, but I'm planning a longish post on chip usage. Are you lot going to get offended if I put that here, or are all musings welcome?
Sounds good to me.

User avatar
Ruth_NZ
Grumpy Old Gorilla
Posts: 9156
Joined: 25 May 2015, 22:46

Re: R_NZ RMT Blog

Post by Ruth_NZ »

BobMem wrote: I don't want to get in the way of this defender chat, because I think there's still plenty of mileage in it, but I'm planning a longish post on chip usage. Are you lot going to get offended if I put that here, or are all musings welcome?
That's another I had on my list. :lol:
Either wait until we open that up or fire away now, up to you. :)
FranckKessie wrote:Collecting the 'easy points' I am not sure I get. Who are the easy points? Is it more or less choose highly owned players?
Not just that, you'd need to read through the last page. But remember, they are my lessons learned; how valuable they will be for others is up to them. Personally I clearly need that principle after the Salah saga because my calibration was out. :lol:

User avatar
Patrician
Grumpy Old Man
Posts: 1492
Joined: 13 Oct 2005, 18:32
Location: ...bears no responsibility for bad points forecasts
FS Record: 2nd TFFE 2008, FPL 3 top 10k finishes
Contact:

Re: R_NZ RMT Blog

Post by Patrician »

It would be a very long post, which might distract from the overall purpose of this blog to describe in full my stats approach (if people are interested I could write it up and start a new thread). I find stats analysis is most useful in the early part of the season (GW 3 to 12). Specifically it helps to:

1) Avoid bad early bandwagons
2) Identify high potential players quicker than the masses
3) Identify an unexpected “One True Bandwagon” early – e.g. Salah this year, Mahrez two seasons ago
4) Have a good early wildcard that sets up the season well
5) Assess if a 4 point hit has > 4 point expected return
6) Quantify the expected returns of a good fixture run for players

This year I found myself in conflict between two FISO commandments:

1) Thou shalt board the One True Bandwagon, however late ye be
2) Though shalt not worship “Carlos Kickaball” until he has proven himself in the EPL

My model identified Salah early, his expected points was lower than his actual early points, but still very high for his price. I didn’t jump on earlier than GW10 because of commandment 2) and comments I read here. Here is a summary of my model, it has many flaws and imperfections, but the outcomes have been good.

Core Assumptions

· The number of shots taken and chances created is a better predictor of FPL points potential than number of goals or assists (much more variance in the latter)
· Conversion of shots to goals, and chances created to assists is consistent amongst players in similar positions (not really true, but simplifies things to build an algorithm)

Baseline expected points is a function of:

· Position
· Number of shots per 90 (and from where, e.g. shots in box have much higher conversion rates than shots outside)
· Number of chances created per 90
· Chance of a clean sheet, based on team chances conceded (using Poisson distribution)
· Number of minutes expected to play

(I ignore bonus points, red/yellow potential, set piece duties, which are all factors to consider)

Forecast expected points is a function of

· Baseline expected points
· Upcoming fixtures - (Home matches I give a 15% boost, and then a boost or penalty based on opposition chances created/conceded)

Worked example

· A forward has a baseline expected points of 5.0, and should play full 90min
· Is playing at home = plus 15% = plus 0.75
· …and is playing against a defence that concedes 20% more chances than average = plus 1.0
· Forecast expected points = 5.0 + 0.75 + 1.0 = 6.75




Sent from my iPad using Tapatalk

User avatar
Mr Clarinet
Grumpy Old Man
Posts: 2253
Joined: 30 Sep 2008, 17:35
Location: Athens

Re: R_NZ RMT Blog

Post by Mr Clarinet »

Aldershot Rejects wrote: 15 May 2018, 10:39
Indeed, I guess my point is that it is possible to load up on your defence and still have 3+ decent captain options each week.

Absolutely, point taken.

I meant to say that at the time, but in the end I severely edited down what had been a much longer post on the pros and cons of big-money forward lines (beyond the captaincy issue) and acknowledgement of your illustration that captaincy choice need not suffer in a defensive formation was inadvertently lost.

User avatar
Aldershot Rejects
Dumbledore
Posts: 9586
Joined: 03 Aug 2011, 17:15
Location: Kent
FS Record: 5th Metro (2010-11); 146 - Sky (2015-16); 218 - Sky (2014-15); 386 - Sky (2020-21); 636 - FPL (2017-18); last 16 Sky Cup (2018-19)

Re: R_NZ RMT Blog

Post by Aldershot Rejects »

Mr Clarinet wrote: 15 May 2018, 12:06
Aldershot Rejects wrote: 15 May 2018, 10:39
Indeed, I guess my point is that it is possible to load up on your defence and still have 3+ decent captain options each week.

Absolutely, point taken.

I meant to say that at the time, but in the end I severely edited down what had been a much longer post on the pros and cons of big-money forward lines (beyond the captaincy issue) and acknowledgement of your illustration that captaincy choice need not suffer in a defensive formation was inadvertently lost.
Indeed. And one of the things that I love about these threads is that we can explore different approaches and learn from each other.

User avatar
BobMem
Grumpy Old Man
Posts: 4222
Joined: 24 Dec 2010, 08:54
FS Record: The Graduates - FISO 5AS H2H Champions 13/14
FISO King of Rock and Roll - Elimination game winner 2015/16

Re: R_NZ RMT Blog

Post by BobMem »

This turned out to be quite long, so apologies in advance!

The value of the chips.

The ever-moving feast that is trying to understand the value of each chip, and when best to use it. I’m considering the 3 ‘performance enhancers’ here, but not the Wildcards, as that’s a whole discussion in itself.

Looking at my own performance data, and anecdotal comments from FISO’s finest, I can draw the following conclusions:

Triple Captain

Probably the least valuable chip. Easy to use whenever you want, but the most susceptible to random variance. I played it on Kane in DGW22, for a massive 3 point gain over standard captainers. I still think it is best used in a DGW, but my other 2 DGW captains scored 8 and 3 (Lukaku 34 and Sterling 37 respectively). There might be a big dollop of bad luck combined with poor captaincy choices on my part. But essentially the TC can be chucked in on any gameweek, and is worth anything from 1 to 29 points! I think anyone who scored double figures with their TC this season should be thankful.

Free Hit

Has been elevated in my opinion from “not sure how to use it” at the start of the season, to “the best chip by far” by seasons end. I played it for DGW34, allowing a smoothish transition to an acceptable team for the short GW35, but 11 doublers for DGW34. Looking at the points my starting XI would have scored, even factoring in the most likely transfer (I tried to be honest with myself) I estimate the Free Hit chip was worth 31 points to me. I think it really had to be used either in DGW34 or SGW35, because of how the fixtures fell. You might argue my approach compromised the team for GWs leading up to 34. My GW33 rank wasn’t great (2.6m) but the 2 preceding GWs were some of my better ones (482k and 683k GR).

All said, if a Free Hit is worth anything like 30 points, then it’s the best chip!

Caveats:
1.       GW31 (482k GR) was captain Salah @29 points
2.       You could use the WC to negotiate DGW34 :arrow: SGW35 transition.

 

Bench Boost

Probably the hardest to really put a value on. I coupled it with a wildcard, for a late season boost: Wildcard 36, BB37, Flourish 38.

My bench scored 13 points in GW37. My average bench score is 5.84 points, with 5 gameweeks exceeding the 13 points benched in DGW37. Interestingly 3 of these came in the first 6 GWs, so the argument you should just burn it early might have some merit. GW1 bench was only 5 points though.

There are a lot of single figure scores in my ‘points benched’ column, so again I suspect a double figure Bench Boost is decent enough. At 10-15 points, it’s really not worth Wildcarding for though.


That’s all I have on those 3 chips. I would love your comments/observations/own anecdotes. I look forward to a long discussion on WC use too. The late WC/BB combo may not be optimal overall, but it isn’t half handy for the knock-out stages of the FISO Spring 16s tournament! :)

User avatar
RomynPG
Dumbledore
Posts: 9639
Joined: 13 Oct 2005, 18:32
FS Record: FPL 134 overall 06/07 ... never close since

Re: R_NZ RMT Blog

Post by RomynPG »

BobMem wrote: 15 May 2018, 11:18 I don't want to get in the way of this defender chat, because I think there's still plenty of mileage in it, but I'm planning a longish post on chip usage. Are you lot going to get offended if I put that here, or are all musings welcome?

I know I'm going to come off as a grump here - but shouldn't these converstaions, defender usage, chip usage etc, be topics on the main forum?

duck-and-cover.jpg
You do not have the required permissions to view the files attached to this post.

User avatar
Ruth_NZ
Grumpy Old Gorilla
Posts: 9156
Joined: 25 May 2015, 22:46

Re: R_NZ RMT Blog

Post by Ruth_NZ »

RomynPG wrote: I know I'm going to come off as a grump here - but shouldn't these converstaions, defender usage, chip usage etc, be topics on the main forum?

I'm just doing my lessons learned appraisal, Romyn, but it has sparked a conversation. To be honest, I think we should be glad that it is going on at all - FPL sites (including FFS as well as FISO) are notoriously quiet in the summer, or will be till the World Cup game gets released anyway. If this was in-season I'd agree with you but at the moment I don't think a gift horse should be looked in the mouth as it were.

User avatar
BobMem
Grumpy Old Man
Posts: 4222
Joined: 24 Dec 2010, 08:54
FS Record: The Graduates - FISO 5AS H2H Champions 13/14
FISO King of Rock and Roll - Elimination game winner 2015/16

Re: R_NZ RMT Blog

Post by BobMem »

You're probably right, I might post something up. I got a bit excited by my first decent gameweek of the season, and all the interesting responses in here to other musings.

Edit: Done.

FranckKessie
Grumpy Old Man
Posts: 1548
Joined: 21 Aug 2017, 21:34

Re: R_NZ RMT Blog

Post by FranckKessie »

Ruth_NZ wrote: Not just that, you'd need to read through the last page. But remember, they are my lessons learned; how valuable they will be for others is up to them. Personally I clearly need that principle after the Salah saga because my calibration was out. :lol:
I did read through the last page and maybe lost a bit due to the advanced use of the English language :) But from what I read "easy points" are introduced by Mr Clarinet here. He uses Salah as a reference...
Mr Clarinet wrote: 14 May 2018, 22:52 ...

This perspective on the graphic might suggest that there isn't a simple choice to be made between going for "easy points" and going for "hard points" - to be competitive, you first need the "easy points", just to be in the game; then you need to outplay competitors for the "hard points". Put it another way - if you eschew the "easy points", you won't see the desired impetus from your hard-won "hard points".
To me this is not really clear and therefore asked for a more detailed understanding of the principle. I may have missed something, but "easy points" for me is best addressed as do like every one else or pick according to ownership/template. This limits the risk you take and you make up a strategy where you rely on your skills to identify the "hard points". This is a sound and widely used strategy. In essesence: Simplify things as much as you can!

Do I miss something or?

In reality I do not think it is easy, because by nature you only know "easy points" once they have emerged. Hopefully you have them onboard, but if not then you are sacrificing points (indirect by using a free FT or by taking a hit) to eliminate risk and often a different options seems viable.

One example could be Salah vs Coutinho around gameweek 13. Up until gameweek 13 Salah had performed magnificent in FPL terms, but from what I had seen he had shown nothing that indicated he would be a contender for Ballon d'or. My impression was that he needed lots of chances to score his goals (he came to many chances though) and he had luck on his side examplified by the goal against United (if I recall correctly). Sanchez in my book had similar luck last year. (It may be off but those are the impressions I remember but never have studied anything) Since GW13 Salah has just kept on improving and added things like the chip with his head which I wouldn't have imagined him capable off.
Can you classify him as "easy points" and when would you have done so?

I like the principle of Squad Structure and "predictable" value, because it is generic and applies to the game mechanics. I am still not sure "easy points" principle does, but this is down to not fully understanding it :)

The approach to picking defenders is something I find interesting. Should you pick the safe "bet" or the more risky bet?
I do not remember further back, but this year it's been Trippier/Davies vs Alderweireld/Verthonghen and Robertson vs Van Dijk which has been the most clear cut cases but also Chilwelll vs Maguire from GW28 onwards. Winner in all occasions have been the risky bet. I have mainly leaned towards the safe expensive bet.

A quick look at understat and their xG and xA show that CB's generally have been unlucky (Van Dijk,Verthonghen, Maguire and Alderweireld) and FB's (Robertson and Davies) has been lucky. So is there more to it than variance, hmm? :?

Generally I believe that this game has much more variance to it than people think. There really aren't that many decisions to be made over the course of a FPL-career.

User avatar
Mr Clarinet
Grumpy Old Man
Posts: 2253
Joined: 30 Sep 2008, 17:35
Location: Athens

Re: R_NZ RMT Blog

Post by Mr Clarinet »

FranckKessie wrote: 16 May 2018, 10:36
I did read through the last page and maybe lost a bit due to the advanced use of the English language :) But from what I read "easy points" are introduced by Mr Clarinet here.

I adopted this terminology specifically with reference to a prior post by Ruth_NZ himself. It was just a way of progressing a discussion, it's not a law of physics. I think the discussion was maybe helpful in framing a general strategic principle, that's all. You can use the same terminology in a different way if you wish.

FranckKessie
Grumpy Old Man
Posts: 1548
Joined: 21 Aug 2017, 21:34

Re: R_NZ RMT Blog

Post by FranckKessie »

Mr Clarinet wrote: 16 May 2018, 11:21
FranckKessie wrote: 16 May 2018, 10:36
I did read through the last page and maybe lost a bit due to the advanced use of the English language :) But from what I read "easy points" are introduced by Mr Clarinet here.

I adopted this terminology specifically with reference to a prior post by Ruth_NZ himself. It was just a way of progressing a discussion, it's not a law of physics. I think the discussion was maybe helpful in framing a general strategic principle, that's all. You can use the same terminology in a different way if you wish.
Arh sorry. Found the original post.

User avatar
Zimmerman
FISO Jedi Knight
Posts: 30211
Joined: 13 Oct 2005, 18:42
Location: having a picnic at the Bear Mountain

Re: R_NZ RMT Blog

Post by Zimmerman »

And yes, I think early on in the season all the talk was of the amount of chances Salah needed. He certainly became more clinical as the season progressed.

User avatar
Patrician
Grumpy Old Man
Posts: 1492
Joined: 13 Oct 2005, 18:32
Location: ...bears no responsibility for bad points forecasts
FS Record: 2nd TFFE 2008, FPL 3 top 10k finishes
Contact:

Re: R_NZ RMT Blog

Post by Patrician »

To illustrate my point
· Salah scored a whopping 8.4 points per 90 minutes played
· Kane scored 5.4 per 90 min as a comparison

Typically ~6 points per 90 is what you would hope from a super premium player. Very rare that a player scores much more consistently over an entire season.
My calculation of expected points using the method above gives me:

· Salah expected points per game based on stats = 6.2
· Kane expected points per game based on stats = 5.8

So Kane’s stats are consistent with his actual points, whereas Salah massively out performed his stats, which Ruth correctly identified. But 6.2 is still an outstanding scoring rate that you only normally see in a handful of super premiums – he was a must have throughout the season at his price.

Why did Salah outperform his stats? He converted 22% of his shots into goals. Average for a striker is 16% - in fact Kane’s conversion % was exactly 16%. Far from being wasteful as reported here, Salah was extraordinarily lethal! Also, 16% of the chances he created became assists, whereas 11% is the average.

During the first 13 games of the season he scored 10 from 49 = 20%. A little lower, but still very high, especially for a midfielder earning 5pts per goal. This idea that he was wasteful in the early part of the season is a myth.


Sent from my iPad using Tapatalk

User avatar
Zimmerman
FISO Jedi Knight
Posts: 30211
Joined: 13 Oct 2005, 18:42
Location: having a picnic at the Bear Mountain

Re: R_NZ RMT Blog

Post by Zimmerman »

Genuinely surprised at that.

Distinctly remember early on being despondent at his wasted chances.

User avatar
BobMem
Grumpy Old Man
Posts: 4222
Joined: 24 Dec 2010, 08:54
FS Record: The Graduates - FISO 5AS H2H Champions 13/14
FISO King of Rock and Roll - Elimination game winner 2015/16

Re: R_NZ RMT Blog

Post by BobMem »

Zimmerman wrote:Genuinely surprised at that.

Distinctly remember early on being despondent at his wasted chances.
He was wasteful.

Just not a wasteful as everyone else has been since stats records began! :D

I'd really like to get into the underlying stats a bit more next season, as the above analysis really interests me. Can you guys recommend good sources for up-to-date stats please?

User avatar
Patrician
Grumpy Old Man
Posts: 1492
Joined: 13 Oct 2005, 18:32
Location: ...bears no responsibility for bad points forecasts
FS Record: 2nd TFFE 2008, FPL 3 top 10k finishes
Contact:

Re: R_NZ RMT Blog

Post by Patrician »

Thats why I have learned to not trust my eye. After all he missed 39 shots in that period, maybe some of them were eye-catching howlers! What mattered was he was getting lots of shots (most in the box), and he was worth having even if his conversion rate dropped from 20% to the 16% average.

At that stage of the season I was more excited about Richarlison. He had had 45 shots and scored 5 (conversion 11%) - I had him pegged to score at a similar rate to Salah. That didn't work out though, so the stats do sometime sell a dud.

User avatar
Patrician
Grumpy Old Man
Posts: 1492
Joined: 13 Oct 2005, 18:32
Location: ...bears no responsibility for bad points forecasts
FS Record: 2nd TFFE 2008, FPL 3 top 10k finishes
Contact:

Re: R_NZ RMT Blog

Post by Patrician »

I pay for members access to FF Scout - it allows you to create your own stats tables with any opta stats.

User avatar
Sutter Kane
Dumbledore
Posts: 7522
Joined: 05 Aug 2010, 12:13
FS Record: Unknown.

Re: R_NZ RMT Blog

Post by Sutter Kane »

Patrician wrote: 17 May 2018, 07:06 I pay for members access to FF Scout - it allows you to create your own stats tables with any opta stats.
And has it helped since joining?
Ruth_NZ wrote: I'm just doing my lessons learned appraisal, Romyn, but it has sparked a conversation.
Indeed and after another non-top 10k season, I again hark back to a number of things that haunt my dreams (well maybe not quite). In terms of lessons to be learned:

1) I heard Mark Sutherns (or Triggerlips), one of them talking 3 years ago about price points and trying to stick to a spend in each area of the pitch. I've never believed in that as every season is different but their results are top drawer over the years. Are they right? (flexibilty of transfers more or less of a problem with this approach?)

2) I have 2 friends in various MLs that are roughly where I am in terms of finishes. Their skill set is different to mine, they see the game differently, the one guy taking more hits, the other swearing by watching live games and MOTD every week. On speaking with them, neither is particularly in tune with the major 'goings on' right up until the last minute so their planning is well...not good. Therefore there is something that they are doing that I am not or something I shouldn't be doing that I am. I believe that thing could be removing players I shouldn't because I think the grass is greener. Thus using a transfer I needn't, for often negligible effect.

3) It's like a broken record but I have a habit of sticking stubbornly to my guns and not getting in Suarez, Michu, Fellaini (anyone remember that one, same year as Michu), Yaya, Salah, etc until too late. I think this has an enormous impetus/momentum/whatever factor. You have to be extremely lucky to finish top 10k whilst avoiding Salah this season for eg.

User avatar
Sutter Kane
Dumbledore
Posts: 7522
Joined: 05 Aug 2010, 12:13
FS Record: Unknown.

Re: R_NZ RMT Blog

Post by Sutter Kane »

Of course all that could be hot air. I am not as good at planning as I think I am. (many on fiso think they are good planners because of discussion like this and vast time spent analysing) It's very difficult to boil down reasons why some players manage top 5k mostly every season and others are always mostly short of that, with the same info available.

User avatar
Tacalabala
FISO Knight
Posts: 19010
Joined: 07 Sep 2008, 01:03

Re: R_NZ RMT Blog

Post by Tacalabala »

I hope nobody minds me posting this here, this may be of interest to a lot of you - Important information about Spring League entry for 2018/19

User avatar
Ruth_NZ
Grumpy Old Gorilla
Posts: 9156
Joined: 25 May 2015, 22:46

Re: R_NZ RMT Blog

Post by Ruth_NZ »

Some interesting comments on here over the last few days, thanks. I plan to continue with the 'lessons learned' process in due course and expect to look at captaincy, chip usage and fixtures/form as well as taking up the discussion with Patrician and others about how best to use stats (I want to create some benchmarks). But I'm going to park all that for the time being and do my review of the top 12 teams and top 20k teams from season 2017/18 while all the information is available. It's quite a lengthy operation but I have some time today so will see how far it can be progressed. :)

User avatar
Ruth_NZ
Grumpy Old Gorilla
Posts: 9156
Joined: 25 May 2015, 22:46

Re: R_NZ RMT Blog

Post by Ruth_NZ »

2017/18 TOP 12 & TOP 20k ANALYSIS

This study compiles some key data from the top-12 FPL finishers and from a cross-section of the top-20k and will also make some short commentaries. To ease the comparison, Green has been used for the Top 12 and Blue for the Top 20k throughout. Brown is used for season 2016/17 numbers.

To give a cross-section of the top 20k I have selected 12 teams from the FISO league ranked at (or close to) the following points in the OR: 1k, 2k, 3k, 4k, 5k, 6k, 8k, 10k, 12k, 14k, 17k, 20k (as last season). So slightly biased towards the top 10k. Fusen's tool has been used, as well as FPL Discovery (and a fair bit of manual labour). :)

For detailed comparison, the section with last season's top-12 data begins here and for the top-20k here. Reference will be made to last season's data in some of the commentary.

1. Bench Boost

The BB scores of the top 12 were: 14; 12; 22; 18; 10; 25; 19*; 19; 19; 21; 7*; 10*; Average 16.3 (last season 20.0).
The BB scores in the top 20k were: 8; 27*; 20; 17; 6; 5; 16; 8; 13; 9; 20; 21; Average 14.2 (16.4).

It should be noted that the BB scores shown have been adjusted to represent the weakest bench that week in cases where the manager hadn't done so. Hits have not been taken into account because it's hard to tie a hit specifically to the BB but the scores shown therefore don't necessarily represent true net values for the BB, they represent the maximum actual value. Those marked * were played outside of a DGW. The average with a SGW BB in this group was 15.8, so barely different to the DGW BB. It may also be worth noting that the highest BB score was from the one manager that played it in GW1. :wink: :lol:

Overall BB averages are down from last season, perhaps representing the tricky nature (in terms of rotation) of the most popular weeks for the BB this season (20 of the 24 teams utilised the BB in DGW34 or 37). My assessment remains that a GW1 BB that can achieve 10-12 points, easing the complications and compromises later on (and reducing any need for hits and/or a 'forced' wildcard) is a perfectly sound way to use the BB chip. If anything, that idea is further confirmed by what can be seen here.

As we will probably see in most categories, the top 20k cross-section average undershoots the top 12 but not hugely.

2. Triple Captain

The TC scores of the top 12 were: 11*; 4; 3; 3; 3; 3; 4; 3; 2; 8; 3; 10; Average 4.75 (13.25).
The TC scores in the 20k group were: 29*; 8; 8; 3; 3; 3; 8; 8; 3; 8; 4; 8; Average 7.75 (9.4).

The averages here are obviously freakish; the top-12 number undershoots even a par score for a normal captaincy. This is despite the TC being theoretically the best (or one of the best) captaincy opportunities a manager can find in 38 weeks. :shock:

What is interesting is that only 10 of 24 managers here followed the 'standard' FFS consensus of a Kane TC in GW22. Last season the popularity of the FFS recommendation (Aguero in GW27 for 9 points) was bigger. But those that swerved the standard route this season mostly didn't gain much because the alternative opportunities (Kane in GW37 for 8 points, for example) didn't do much better. The big winner was the brave manager that TC'd Salah in GW31 for 29 points. :lol:

3. Wildcards

The 1st Wildcard was played in the following gameweeks: 8; 8; 15; 8; 6; 10; 8; 8; 10; 8; 19; 8.
The 1st Wildcard was played in the following gameweeks: 2; 8; 12; 6; 4; 10; 10; 14; 8; 8; 4; 13.
The 2nd Wildcard was played in the following gameweeks: 25; 32; 32; 32; 33; 27; 28; 32; 28; 32; 28; 28.
The 2nd Wildcard was played in the following gameweeks: 23; 32; 33; 32; 32; 36; 33; 32; 33; 28; 32; 32.

There is no data available for these groups for 1st wildcard usage last season but it seems clear that this season, at least, the early 1st wildcard wasn't ideal. GW8 was the most popular amongst the most successful teams (7 of 12) and only one of them went any earlier. Amongst the 20k group it was more mixed but still with 2/3 of them wildcarding in GW8 or later.

The 2nd wildcard timing is interesting, though. Last season only 1 team in the top 12 used the 2nd wildcard before GW32; this season 50% of them did. Perhaps the assumption that the 2nd wildcard should be held for the DGW period (perhaps to set up the BB, for example) is weakening its hold? Similarly to the BB (where 25% of the top 12 used it outside of a DGW), it seems that the most successful teams were perhaps more willing to diverge from the conventional wisdom.

4. Free Hit

Obviously there is no data from last season as this was the first appearance of the FH chip. Looking at scores achieved would be misleading because the FH chip is more about how much you can beat the average by in the week when you play it.

The FH chips of the top 12 were played in GW: 31; 35; 35; 35; 35; 35; 31; 35; 31; 35; 31; 34.
The FH chips of the top 20k group were played in GW: 22; 35; 35; 35; 35; 34; 35; 35; 35; 35; 35; 35.

A vast consensus towards use in a BGW and every one of those that used the FH in such a way achieved a high GW rank in that week. Those that used it differently (3 of 24) appeared to do less well with it in terms of GW rank, although it's hard to assess the net effect; playing the FH in a DGW, for example, would have been unlikely to result in a huge GW score against others using BB and TC but might nevertheless have protected the manager against what would otherwise have been a significant fallback.

To be continued.
Last edited by Ruth_NZ on 22 May 2018, 06:48, edited 1 time in total.

User avatar
Billy Bongo
FISO Knight
Posts: 12000
Joined: 21 Jul 2013, 22:18

Re: R_NZ RMT Blog

Post by Billy Bongo »

Excellent post

Sent from my SM-G965F using Tapatalk


User avatar
Patrician
Grumpy Old Man
Posts: 1492
Joined: 13 Oct 2005, 18:32
Location: ...bears no responsibility for bad points forecasts
FS Record: 2nd TFFE 2008, FPL 3 top 10k finishes
Contact:

R_NZ RMT Blog

Post by Patrician »

We get seduced (myself included) by the DGW because we intuitively assume double the expected points, but the facts are quite different if expected points correlate to minutes played.

In a normal game week, the average number of minutes played per player is 71 min assuming all three subs are used (90 x 11 / 14). Furthermore 8 out of the 14 players (57%) play the full 90, and they are easier to predict.

Here are actuals from GW34. Average number of minutes per match per playing player was 60 min. Just 30% of DGW players played the full 180 (quite heavily dominated by GK and DEF).

The bigger the team, the less likely to have players doing the full 180. Number of players playing the full 180 in GW34:

8 BUR
6 BOU, SOT
5 BHA
4 LEI (Maguire, Chilwell, Mahrez, Vardy), CHE (Courtois, Azpi, Cahill, Kante), TOT (Lloris, Vertonghen, Davies, Eriksen)
2 MUN (De Gea, Smalling)

Seems like quite a roll of the dice to base a major strategy around.



Sent from my iPad using Tapatalk
Last edited by Patrician on 21 May 2018, 22:18, edited 1 time in total.

User avatar
Zimmerman
FISO Jedi Knight
Posts: 30211
Joined: 13 Oct 2005, 18:42
Location: having a picnic at the Bear Mountain

Re: R_NZ RMT Blog

Post by Zimmerman »

What about GW37?
What about previous seasons?

User avatar
Patrician
Grumpy Old Man
Posts: 1492
Joined: 13 Oct 2005, 18:32
Location: ...bears no responsibility for bad points forecasts
FS Record: 2nd TFFE 2008, FPL 3 top 10k finishes
Contact:

R_NZ RMT Blog

Post by Patrician »

Similar story in GW37, average mins played was 59min per match, and 32% got a full 180min. Might look at previous seasons later.

Number of players playing 180min

8 BHA
6 SWA, SOT, HUD, LEI, NEW
5 ARS
4 TOT, CHE
3 MUN (De Gea, Pogba, Smalling)
2 MCI (Fernandinho, Sane)


Sent from my iPhone using Tapatalk

User avatar
Archy
Grumpy Old Man
Posts: 3210
Joined: 22 Oct 2011, 10:09
FS Record: Maybe all is not lost after all

Re: R_NZ RMT Blog

Post by Archy »

I think before people throw too much cold water on the DGW BB idea, I'd like to point out the following:

I scored 28 points from a GW37 BB and, if anything, think this was slightly unlucky. Only a fortunate deflected goal in the last minute of injury time from Everton cost Southampton a 2nd clean sheet, which would have added a further 10 points or so to the tally.

Also, a carefully selected team can avoid rotation issues to a large extent. I think 12 of my 15 played both games (and all 4 of my bench played both games), and most of those played the full 90 both times. Players like Kane, Eriksen, Mahrez and Vardy simply don't get rested so careful planning can help avoid some of the pitfalls (at least it did on this occasion)

On the negative side, the late use of the wildcard and BB had an opportunity cost as I held on to weakened team for too long (espcially through GWs 29-35), so i still think the GW1 BB idea carries merit.
Last edited by Archy on 22 May 2018, 18:24, edited 1 time in total.

View Latest: 1 Day View Your posts
Post Reply

Return to “FPL Team Diaries & RMTs”