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

Fantasy Luck Vs. Fantasy Skill

A Fantasy Football forum for news on fantasy football games run by the Premierleague (FPL).

Fantasy Skill Vs. Fantasy Luck

0 - 25 % Luck - Luck wont win you things in a Fantasy Season. Skill and Hardwork will triumph eventually.
45
21%
25% - 50% Luck - Need to be well read and well informed. Indepth Knowledge certainly Helps. Luck is not so important.
109
52%
50 % - 75 % Luck - If the football fundamentals are strong, then you can beat any one.
44
21%
75 % - 100 % Luck - Basic Knowledge of Football is More than Enough
13
6%
 
Total votes: 211

User avatar
Carlos Kickaball
Dumbledore
Posts: 7801
Joined: 04 Sep 2013, 18:02

Re: FPL Skill vs. Luck: An Epic Debate

Post by Carlos Kickaball »

:lol: - It was an attempt to try and get round peoples fondly held beliefs about the current issue.

I'm sure they'll be some more examples later. :D
Last edited by Carlos Kickaball on 27 Apr 2015, 18:16, edited 1 time in total.

User avatar
Mav3rick
FISO Jedi Knight
Posts: 20858
Joined: 20 Jul 2009, 20:35
FS Record: FPL: 1082, 1201, 1800, 10203

The stats are dark and full of errors.

Re: FPL Skill vs. Luck: An Epic Debate

Post by Mav3rick »

I don't think anyone has a fondly held belief that trees don't require sunlight...

To be honest, I don't actually know what it is you're arguing against. The Dazzler is saying that there's a mixture of skill and luck, you're agreeing??

User avatar
Carlos Kickaball
Dumbledore
Posts: 7801
Joined: 04 Sep 2013, 18:02

Re: FPL Skill vs. Luck: An Epic Debate

Post by Carlos Kickaball »

Yes, I agree that there is a mixture, and see no problem in debating the relative importance of each.

However if he does think there is a mixture, and then goes on to say that there is no reason to assume why the top performer would have had better than average luck, he's clearly incorrect.

User avatar
thekrumcake
FISOhead
Posts: 835
Joined: 25 Sep 2012, 16:15

Re: FPL Skill vs. Luck: An Epic Debate

Post by thekrumcake »

I think what the Dazzler wrote makes a lot of sense, except the percentages. We need a time frame to set a percentage. If the 10 % luck estimate is over one season, then the luck factor is ~0% over seven seasons, right?

Either way, I don't think it makes much sense trying to come up with percentages. I would rather someone came up with an estimate on how many seasons/GWs a player needs to play to achieve her expected average ranking.

User avatar
Stemania
FISO Jedi Knight
Posts: 20448
Joined: 27 Aug 2006, 11:54
Location: On the Iron Throne of xG, the seat of The Crown Prince of the Stat Perverts. Or if not, in the STC!
FS Record: Best: TFF 321st. FPL 129th. FFS Career HoF Peak 2nd (Live 1st). Ability since lost.

Re: FPL Skill vs. Luck: An Epic Debate

Post by Stemania »

Carlos Kickaball wrote:if you take the best performer then it follows that they are likely to be one of the best for skill
Isn't this all that matters regarding Ville and the reasons people are following him? I don't understand why discussing luck at all is even vaguely important if it's already been agreed he's very likely one of the best players.

User avatar
sween
FISOhead
Posts: 608
Joined: 24 Jul 2009, 07:27
FS Record: 3035 last season

Re: FPL Skill vs. Luck: An Epic Debate

Post by sween »

Hello all. My first post in ages, but this is such an interesting topic for me (I am an investment risk analyst) I thought I would share my thoughts...

Firstly, in trying to quantify "luck" we are really talking about positive standard deviation. Or rather, the times we achieve a positive result above what you could typically expect from any outcome. "bad luck" of course is simply the opposite. For example, those of you who have brought in the likes of Moreno or Schlupp this week, you can consider yourselves "unlucky" thus far due to the the fact that given the information at the time you bought them, you could reasonably expect a return from either player above zero. Likewise for those of you who have John Stones coming off the bench for 15 points, you can consider that very lucky as that players return represents an extreme positive outlier in his likely range of returns. When we talk about luck we are really talking about enjoying or suffering an unsually high degree of outliers in any sample of returns over a given period in time.

Given this, in a game that attracts almost 3.5m players, the top and the bottom of this range will be seperated by skill 99.9% of the time. To take the most extreme example, someone who has a good knowledge of the game and plays every week is almost absoultely certain to beat someone who selects their team each week by random. In the case of someone who at least tries, having very little knowledge, against someone who knows the game very well (I mean both football and FPL), a 99.9% chance of victory still allows for a few hundred people to punch way above or below their weight over the course of a season. Besides a tiny tail risk however these players are almost certainly seperated by skill rather than luck.

Luck, or deviation from the mean, really only becomes significant when players are very evenly matched. There are many instances over the course of a season where players must make an almost exact 50/50 call when it comes to picking a captain, or benching a player. Over a season the average person should come out broadly neutral when it comes to these decisions, however in a game played by so many, there will of course be many players who get significantly "luckier" or "unluckier" than average. In the same way someone could call a coin toss correctly 5 times out of 5 around 3% of the time, 3% is a signficant number of players, even if we assume the best players make up the top 10,000. Regarding everyone on the site, you will have some who belong to that 3% of lucky players and 3% that belong to the unlucky cohort.

Someone referred to poker player Jamie Gold above. He is an example of an extremely positive outlier in a game based upon making marginal calls. Lucky for him he managed to get that run of positive outliers consecutively in the biggest game he ever played in (in the same way people manage to get 6 numbers right in the national lottery).

So when it comes to seperating say the top 1000 or so players in the FPL world, over a season the definitive factor will be standard deviation (or luck). Even if betting on the Premier League was perfectly efficient and a player played FPL in a perfectly efficient manner, he is still very very unlikely to win FPL (or come close to winning) over a season. Because to do so would require a huge degree of fortune over 38 gameweeks in order to dodge the slings and arrows of misfortune while millions of players are trying to do the exact same. Simply put, a team of the most knowlegable football students, quant analysts, and logical geniuses, could put together a logically perfect FPL team based upon perfect information (if we can assume in theory such a thing could exist), and while they may start the season as favourites to win FPL against any other single team, they would still be very very unlikely to win against every other team.

So in conclusion one cannot make statements such as "the game is 70% skill and 30% luck" as that ratio changes based upon the gulf in knowlegde between yourself and any other player. Rather we can only say, the difference between the best and worst players is almost certainly 100% based on skill, and the difference between the best and the second best player in the world will be seperated at the end of the season almost certainly 100% by luck.

User avatar
Carlos Kickaball
Dumbledore
Posts: 7801
Joined: 04 Sep 2013, 18:02

Re: FPL Skill vs. Luck: An Epic Debate

Post by Carlos Kickaball »

Great post Sween, good to have some very constructive ones, and I would agree with most of what you've said, and have expressed similar in another topic for a while.

This is essentially the central point (to me anyway):
sween wrote:So when it comes to seperating say the top 1000 or so players in the FPL world, over a season the definitive factor will be standard deviation (or luck).
The real questions comes in how large this group can get until the managers we start to include are much worse, and how different in skill are say the 1000th, 10000th, and 50000th best managers, to the very best.

When evaluating the effect of luck it's clearly erroneous to selectively chose one of the most uniform records, as some seem to have done on here (for by example by choosing the best), and it would be much better done looking at a larger sample of records, perhaps even just the ones of active forum members.

User avatar
MPTree
Grumpy Old Man
Posts: 2008
Joined: 13 Oct 2013, 13:44

Re: FPL Skill vs. Luck: An Epic Debate

Post by MPTree »

Great post, sween. A really good read. Thanks.

User avatar
Red Eye
Grumpy Old Man
Posts: 4413
Joined: 07 May 2006, 07:12
Location: Republic of Mancunia

Re: FPL Skill vs. Luck: An Epic Debate

Post by Red Eye »

OIEIAO wrote:I think some people are thinking about top players wondering if they were born with some magic touch, but "skill" is defined as "learned ability" - I don't think there's much about fantasy football you can't learn.

Back on poker: I've hard people say there are probably 1000 players (maybe it was less, maybe only 100) who are good enough to win the World Series of Poker - the remainder will not beat enough of the more skilful players just through luck (they'll beat some, some of the time, but not all). The top 100, however, will fight it out and luck will be a factor in which of them make the final 10 and actually win the whole thing.

In the same way, I feel the gap between being 1 million and 10k in FPL is mainly skill/experience/dedication. The gap between there and the top 10 is probably a bit more skill and a fair dose of luck.
Was reading a Daniel Kahneman book on thinking/decision making recently and he laid out two basic conditions for the acquisition of skill:

1) The environment is sufficiently regular to be predictable.
2) There is an opportunity to learn these regularities through prolonged practice.

I would say things like poker and chess are more regular/predictable than football matches. 'Experts' who operate in very irregular environments are not significantly more accurate in their predictions than dart-throwing monkeys, but they do create/experience an illusion of expertise. In fact, those who know most are often worse as they can be deluded about their own 'skill' and are unrealistically overconfident.

I think football has a degree of predictability. For example, at the start of the season you could fairly safely predict which 5 clubs would finish at the top. Given the correlation between wage bill and league position, if you were serious about it you could probably come up with some fairly simple calculations around which players to pick based on price, who they play for, how many minutes they get etc. Beyond that I think it is largely luck whether you buy Ramsey or Ozil or captain Hazard over Costa and outcome bias is an obvious problem when trying to assess after the fact.

There are also occasionally some 'predictable irregularities' which can be taken advantage of if you're on the ball. If Gary Cahill broke his leg then people would buy Zouma in droves so obviously you need to have that level of knowledge and to pay attention consistently. I'm fairly sure I make better decisions when there is an 'absence of neglect' - if I do my transfers in the space of 2 minutes on a Saturday morning and don't pay attention to price changes during the week then obviously I'll do worse. So you need a base level of application but that certainly isn't skill and the vast majority of people who play the game don't have that - they just tinker/play for fun and sometimes they'll pay close attention other times they won't.

User avatar
Carlos Kickaball
Dumbledore
Posts: 7801
Joined: 04 Sep 2013, 18:02

Re: FPL Skill vs. Luck: An Epic Debate

Post by Carlos Kickaball »

If anyone is interested Redeye is referring to Thinking, Fast and Slow, Chapter 22: Expert intuition when can we trust it, worth a read. Indeed Kahneman himself says:
Daniel Kahneman wrote:If subjective confidence is not to be trusted, how can we evaluate the probable validity of intuitive judgement? When do judgements reflect true expertise? When do they display an illusion of validity? The answer comes from the two basic conditions for acquiring a skill:

- an environent that is sufficiently regular to be predictable
- an opportunity to learn these regularities through prolonged practice

When both these conditions are satisifed, intuitions are likely to be skilled. Chess is an extreme example of a regular environment, but bridge and poker also provide robust statistical regularities that can support skill.
I have often found the comparison between FPL and poker useful, but also quite an unusual one due to some of the differences. The volume of decisions and regularity in poker are strikingly different.

User avatar
Stemania
FISO Jedi Knight
Posts: 20448
Joined: 27 Aug 2006, 11:54
Location: On the Iron Throne of xG, the seat of The Crown Prince of the Stat Perverts. Or if not, in the STC!
FS Record: Best: TFF 321st. FPL 129th. FFS Career HoF Peak 2nd (Live 1st). Ability since lost.

Re: FPL Skill vs. Luck: An Epic Debate

Post by Stemania »

Carlos Kickaball wrote: When evaluating the effect of luck it's clearly erroneous to selectively chose one of the most uniform records, as some seem to have done on here (for by example by choosing the best), and it would be much better done looking at a larger sample of records, perhaps even just the ones of active forum members.
The posters who picked out Ville's record were not trying to evaluate the effect of luck.

User avatar
Carlos Kickaball
Dumbledore
Posts: 7801
Joined: 04 Sep 2013, 18:02

Re: FPL Skill vs. Luck: An Epic Debate

Post by Carlos Kickaball »

It was selected due to how good it was, and then used in this way.

User avatar
Stemania
FISO Jedi Knight
Posts: 20448
Joined: 27 Aug 2006, 11:54
Location: On the Iron Throne of xG, the seat of The Crown Prince of the Stat Perverts. Or if not, in the STC!
FS Record: Best: TFF 321st. FPL 129th. FFS Career HoF Peak 2nd (Live 1st). Ability since lost.

Re: FPL Skill vs. Luck: An Epic Debate

Post by Stemania »

As you well know, this thread was a splitting off from the Discussion of genius players thread. People brought up Ville's record there due to it's unusually high standard. A discussion of luck framed around Ville's individual record was essentially initiated there by you, so to now critisise others for including Ville's record in the discussion seemed a bit cheap to me, 'tis all. ;)

User avatar
Carlos Kickaball
Dumbledore
Posts: 7801
Joined: 04 Sep 2013, 18:02

Re: FPL Skill vs. Luck: An Epic Debate

Post by Carlos Kickaball »

Nope I didn't solely use Ville's record to make my conclusions about the amount of luck involved in the game in general. I looked at the variability across many records from good managers. A very cheap shot, both are misrepresentations of what I have said. Just as some facts and interesting debate was starting to happen too. I'd be happy for the last 4 posts to be deleted as they offer nothing to the debate.
Last edited by Carlos Kickaball on 28 Apr 2015, 17:00, edited 1 time in total.

User avatar
Red Eye
Grumpy Old Man
Posts: 4413
Joined: 07 May 2006, 07:12
Location: Republic of Mancunia

Re: FPL Skill vs. Luck: An Epic Debate

Post by Red Eye »

Haven't followed that thread in detail but think to do something like that you need insight on how decisions are being made - is he using intuitive judgement/'expertise' or is it stats driven/calculated 'for him' in some way? And I think you need to be able to show desirable outcomes were somehow predictable before the event.

User avatar
Stemania
FISO Jedi Knight
Posts: 20448
Joined: 27 Aug 2006, 11:54
Location: On the Iron Throne of xG, the seat of The Crown Prince of the Stat Perverts. Or if not, in the STC!
FS Record: Best: TFF 321st. FPL 129th. FFS Career HoF Peak 2nd (Live 1st). Ability since lost.

Re: FPL Skill vs. Luck: An Epic Debate

Post by Stemania »

sween wrote:Simply put, a team of the most knowlegable football students, quant analysts, and logical geniuses, could put together a logically perfect FPL team based upon perfect information (if we can assume in theory such a thing could exist), and while they may start the season as favourites to win FPL against any other single team, they would still be very very unlikely to win against every other team.
I'm not actually so sure about this statement, though I guess it depends on what you mean by 'very very unlikely'. I think it really underestimates just how incomplete every fpl manager's information is, and how many logically imperfect decisions every one of us makes in a season. I certainly don't think the statement is obviously true.

How many of the players on this forum really had an objective good detailed assessment of all the Leicester players going into the double? Most of our knowledge came from the odd other poster on the forum, the odd background stat, or a vague memory of a few match highlights. That's an awful long way from what a real football expert might know. How are we each honestly going to pick between Ramsey and Ozil when the decision comes around?

So, for example, what would be a logically perfect way to choose a captain? The only really quantifiable method would presumably be to gather every piece of information, and watch every match, available to us and then produce a set of very accurate points projections (some attempted but imperfect examples might be the numbers provided by FFFix or FFS, or betting odds) and pick the highest projected player every time. But in my experience almost every fpl player at the end of the day chooses their captain on a 'feeling', especially when it seems 'close call', and then sometimes backs it up with projections and odds if they agree. When the projections or odds don't agree, it's usually the projections or odds that are critisised as being slightly inaccurate due to some bit of information being overlooked or them being skewed by recent events. :?

In the crack the code thread, nigel mentioned that he's essentially used an algorithm/computer to run his team this year and despite it being fairly untested and needing ''a lot of work'' his team is around 9000th place - make of that what you will. :|


One of the problems with the whole luck skill debate (every time it's had) is that it almost always assumes that every player tries his hardest throughout the season (or rather, is consistent in the type of decisions they make and thought they put into each decision) every week, year on year, and that their fpl ability essentially never changes. It's also usually assumes that either every 'skilled' player has roughly the same football knowledge and methodology and that these don't vary either. Sometimes the level of knowledge and level of consistency might be included in the definition of 'skill', but that's usually forgotten and 'skill' quickly just becomes decision making ability again. :(

User avatar
sween
FISOhead
Posts: 608
Joined: 24 Jul 2009, 07:27
FS Record: 3035 last season

Re: FPL Skill vs. Luck: An Epic Debate

Post by sween »

Finisher1 wrote:
MPTree wrote:
Finisher1 wrote:So CK, are you suggesting that it is not certain if thesilkworm is actually the best FISO manager this season, despite the pure fact that he has the best overall rank? It may just be that he has been more lucky than some other, more skilled FISO manager?
Depends how you define "best". It's not always correct to define the quality of a decision is by its outcome. Would be interesting to hear silkworm's own take on the role luck has played in his season so far.
Well, he captained Kane in that famous gameweek when Kane trashed Aguero. So did I.

I think that was a very skillful decision, but some people here said it was luck.
In what way was it based upon skill though? Did you have access to information that other players, or betting exchanges did not? Did you quantify the data you did have better than others who picked Aguero? If so, can you verify your methodology by repeating the process over many seemingly 50/50 calls over a season?

If you can, there may be enough evidence to suggest there is a skill factor involved in picking between two very close options (and you would also be a very rich man!). If you cant you most likley have flipped the coin and won.

Carlos Kickaball wrote:Great post Sween, good to have some very constructive ones, and I would agree with most of what you've said, and have expressed similar in another topic for a while.

This is essentially the central point (to me anyway):
sween wrote:So when it comes to seperating say the top 1000 or so players in the FPL world, over a season the definitive factor will be standard deviation (or luck).
The real questions comes in how large this group can get until the managers we start to include are much worse, and how different in skill are say the 1000th, 10000th, and 50000th best managers, to the very best.

When evaluating the effect of luck it's clearly erroneous to selectively chose one of the most uniform records, as some seem to have done on here (for by example by choosing the best), and it would be much better done looking at a larger sample of records, perhaps even just the ones of active forum members.
Cheers Carlos,
I think over a season there is really no way of seperating say the top 1000 players in a sort of skill to luck ratio. There are simply too many variables to consider. I think comparing very long term records is an obvious way of proving skill but even then it is difficult to make any real comparisons given so few players are likley to dedicate themselves fully to FPL over the very long term. It would demonstrate long-term dedicated rather than intelligence, knowlegde, or skill.
Stemania wrote:
sween wrote:Simply put, a team of the most knowlegable football students, quant analysts, and logical geniuses, could put together a logically perfect FPL team based upon perfect information (if we can assume in theory such a thing could exist), and while they may start the season as favourites to win FPL against any other single team, they would still be very very unlikely to win against every other team.
I'm not actually so sure about this statement, though I guess it depends on what you mean by 'very very unlikely'. I think it really underestimates just how incomplete every fpl manager's information is, and how many logically imperfect decisions every one of us makes in a season. I certainly don't think the statement is obviously true.

How many of the players on this forum really had an objective good detailed assessment of all the Leicester players going into the double? Most of our knowledge came from the odd other poster on the forum, the odd background stat, or a vague memory of a few match highlights. That's an awful long way from what a real football expert might know. How are we each honestly going to pick between Ramsey and Ozil when the decision comes around?

So, for example, what would be a logically perfect way to choose a captain? The only really quantifiable method would presumably be to gather every piece of information, and watch every match, available to us and then produce a set of very accurate points projections (some attempted but imperfect examples might be the numbers provided by FFFix or FFS, or betting odds) and pick the highest projected player every time. But in my experience almost every fpl player at the end of the day chooses their captain on a 'feeling', especially when it seems 'close call', and then sometimes backs it up with projections and odds if they agree. When the projections or odds don't agree, it's usually the projections or odds that are critisised as being slightly inaccurate due to some bit of information being overlooked or them being skewed by recent events. :?

In the crack the code thread, nigel mentioned that he's essentially used an algorithm/computer to run his team this year and despite it being fairly untested and needing ''a lot of work'' his team is around 9000th place - make of that what you will. :|


One of the problems with the whole luck skill debate (every time it's had) is that it almost always assumes that every player tries his hardest throughout the season (or rather, is consistent in the type of decisions they make and thought they put into each decision) every week, year on year, and that their fpl ability essentially never changes. It's also usually assumes that either every 'skilled' player has roughly the same football knowledge and methodology and that these don't vary either. Sometimes the level of knowledge and level of consistency might be included in the definition of 'skill', but that's usually forgotten and 'skill' quickly just becomes decision making ability again. :(
I think "very very unlikely" would mean <1%, and I would happily give anyone odds of 101/1 that they will win FPL over any given season. There is simply too much noise in the stats over the course of a season for anyone to be a significant favourite (101/1 would actually be a significant favourite in the context of 3.5 million players). Using your own example of a detailed knowledge of Leicester players going into this gameweek, I am sure the vast majority of players do not have a complete knowledge. But then, there is no way that crack team of experts could know Schlupp was injured (excluding insider knowledge). Picking Schlupp rather than Morgan resulted in a 9 point gain/loss at the weekend. Schmeichel ended up with his highest point score of the season because a penalty clipped the wrong side of a post. While Vardy managed 7 points rather than 2 because the whole ball wasnt over the line when he poked it in the net. These are tiny margins, and even if we can assume that a single player may have the most complete knowledge of any player, the game has far too much standard deviation to allow him to be clear favourite when he is playing against tens of thousands of other active players.

As regards, building a model to finish around 9000th, that is impressive and absoultely believable. I think taking a quant approach with perfect knowledge of the game should get you confortable within that range if you are smart and dedicated enough to apply it. But then, as impressive as it is, that team is still 9000 places from winning! And it is in this space where the noise in the data is most significant.

User avatar
Red Eye
Grumpy Old Man
Posts: 4413
Joined: 07 May 2006, 07:12
Location: Republic of Mancunia

Re: FPL Skill vs. Luck: An Epic Debate

Post by Red Eye »

Stemania wrote:In the crack the code thread, nigel mentioned that he's essentially used an algorithm/computer to run his team this year and despite it being fairly untested and needing ''a lot of work'' his team is around 9000th place - make of that what you will. :| (
I think that sums it up quite well - despite it being not a perfect model, if you limit/constrain subjective judgements you can surely achieve a very good level of performance simply by applying some rules. Pretty sure there would be diminishing returns on this though and even if Nigel did indeed to "a lot of work" he may not perform significantly better simply because there is too much random variation in football (i.e. its a funny old game :)).

You could also try a 'wisdom of crowds' approach and just follow the most popular transfers each week but you would think doing either of these must take an element of the fun out of it.

User avatar
Carlos Kickaball
Dumbledore
Posts: 7801
Joined: 04 Sep 2013, 18:02

Re: FPL Skill vs. Luck: An Epic Debate

Post by Carlos Kickaball »

It comes back to:
Red Eye wrote:1) The environment is sufficiently regular to be predictable.
Some of these posts seem to imply that vast quantities of statistical data, being able to watch every minute of every match, being a football mastermind, being a superb statistician, will give you a big prediction edge over in these close calls, over someone who is an avid Premier League football fan, that watches some games and highlights, and has access to all the information in the online online stats, odds, forums and media. I just don't think that is true, once you get to a certain point the extra data becomes quite superfluous, especially with how many random factors there are. We just don't have models and football experts who can really tell you who will get a clean sheet, we just have rough ideas.

In something like poker or chess it is very different, players have to make decisions themselves, and the calculations and skills needed to make the best decisions are much higher. There will be things that seem more obvious to a grandmaster, than a committed club player, and the grandmaster can explain and demonstrate why. But I don't think this is true for FPL, the guess of a well informed committed reasonable FPL player who has access to the same things we do, is not going to be much worse than a expert football obsessed statistician who watches every game and has UEFA coaching badges. If you go onto a chess or poker forum, you'll probably observe a consequence of this, which is that the ideas being shared in comparison to an FPL forum will be much more complicated.

User avatar
Red Eye
Grumpy Old Man
Posts: 4413
Joined: 07 May 2006, 07:12
Location: Republic of Mancunia

Re: FPL Skill vs. Luck: An Epic Debate

Post by Red Eye »

Not a statistical expert but it occurred to me you could perhaps use a control chart to understand whether a player's scores are stable and predictable. So I did one with Hazard's scores and this is the result - it shows the mean score of 6.48 and an 'upper control limit' of 21.20, which as I understand it means anything above that is statistically unlikely. Anything within that would be classed as 'random cause variation' so if you have him in your side you can reliably expect him to score between 0 and 21 in any given week :? (which has happened every single week this season). In FPL terms that is obviously quite a lot of variation and Hazard is surely one of the most stable/consistent players in the game.

One possible limitation though is it does assume all matches represent the same 'process' and doesn't differentiate by standard of opposition. So I might look at that but generally this says to me we spend the season trying to 'catch waves' and it is basically luck whether an individual player scores 2 or 21.

From a broader perspective I guess the issue is how best to compose your team to maximize the amount of money you have and we all have simple rules of thumb about that (i.e. the player is picked consistently etc). Other than that I guess you are looking for unusual events that skew things, like someone playing out of position or Kane getting promoted to the first team etc.
You do not have the required permissions to view the files attached to this post.

Razzler
FISOhead
Posts: 633
Joined: 09 Feb 2014, 12:41

Re: FPL Skill vs. Luck: An Epic Debate

Post by Razzler »

I don't consider myself a 'lucky' player but I do ok

Yep, 10K area and I finished 165th or something years back probably before it became a huge obsession for the entire world!! :D

I brought Schlupp in for his clean sheet against Swansea and getting nuthin this week. I also brought in Moreno would I could have picked Skrtel or Lovren. Ho-Hum

I looked back at my GW scores recently and found that my captain probably misfired near half the time. I can't recall if other players did better but I think in a lot of instances they did even when the captain I selected seemed the logical choice (Aguero at home, Kane at home, etc)

I now await Hazard getting a 1 or 2 tonight when I REALLY need him to fire as captain as my lead has been closed down to a few point and opponent doesn't have Hazard at all ... a decent score would make a huge difference in light of Schlupp and Moreno but ... well, let's all see :D

Who do I have coming off the bench ... no fancy-schmancy Stones for me, I have Alders and O'Shea ... both conceeded and also booked this weekend :p

User avatar
Red Eye
Grumpy Old Man
Posts: 4413
Joined: 07 May 2006, 07:12
Location: Republic of Mancunia

Re: FPL Skill vs. Luck: An Epic Debate

Post by Red Eye »

So I broke Hazard's scores down by home/away and 'opposition category' (1-5) based on last season's league positions. Chelsea have only played two away games against category 5 (which I put Leicester in) and he averaged 4 in those games. Include both home and away against cat 5 and the mean is 7.7. Either way you can reliably expect him to score between 2 and 14-15 in tonight's game. Anything above that is statistically unlikely but anything within it is random variation (i.e. 'luck').

User avatar
sween
FISOhead
Posts: 608
Joined: 24 Jul 2009, 07:27
FS Record: 3035 last season

Re: FPL Skill vs. Luck: An Epic Debate

Post by sween »

Red Eye wrote:Not a statistical expert but it occurred to me you could perhaps use a control chart to understand whether a player's scores are stable and predictable. So I did one with Hazard's scores and this is the result - it shows the mean score of 6.48 and an 'upper control limit' of 21.20, which as I understand it means anything above that is statistically unlikely. Anything within that would be classed as 'random cause variation' so if you have him in your side you can reliably expect him to score between 0 and 21 in any given week :? (which has happened every single week this season). In FPL terms that is obviously quite a lot of variation and Hazard is surely one of the most stable/consistent players in the game.

One possible limitation though is it does assume all matches represent the same 'process' and doesn't differentiate by standard of opposition. So I might look at that but generally this says to me we spend the season trying to 'catch waves' and it is basically luck whether an individual player scores 2 or 21.

From a broader perspective I guess the issue is how best to compose your team to maximize the amount of money you have and we all have simple rules of thumb about that (i.e. the player is picked consistently etc). Other than that I guess you are looking for unusual events that skew things, like someone playing out of position or Kane getting promoted to the first team etc.
Its an interesting way to look at things but a range in itself doesnt really tell you much. You need to have a breakdown of likelihood assuming a normal looking distribution. So in a stat sense, if you look at the range of Hazard or any player so will see that most weekly point total sit within a pretty tight range (Im guessing 2-9 in the case of Hazard. Interesting you might find that within that range you see a spead at the high and low end of the range, and little in between (more 2, and 3, and more 8 and 9, while less 4 and 5). You look at the distribution and you put a % confidence interval at various points in that range. So for example you would say that you can have 65% confidence Hazard will earn between 2-7 points, 95% confidence that he would earn between 0-15, and 99% he would earn between -2- 25. As you highlight though you then need to look at the ranges within the whole range depending on the opponent. And then, you need to use this data relative to the value of the player. Finally you need to consider if he is flagged doubtful, if he is likely to be rotated, if he one yellow off a suspension, etc, etc.

These are made up numbers, but hopefully highlights just how tough it would be to apply a risk-based statistical approach to picking an FPL team and the degree of variance within the game. It wouldnt be impossible by any means, but it would be a full time job to apply this kind of work to a professional level. As I said earlier, this is why the position of the top players will always be determined by variance rather than any knowledge or skill.

If you are interested in applying stats to football, find the "World Cup and Economics Report" by Goldmans (just google it). It is economic based rather than statistical but still gives an idea of the kind of tools you can apply if you have the time

User avatar
Carlos Kickaball
Dumbledore
Posts: 7801
Joined: 04 Sep 2013, 18:02

Re: FPL Skill vs. Luck: An Epic Debate

Post by Carlos Kickaball »

It won't be normal looking, for a one game week event, it will be multimodal. Part of what makes it so unpredictable.

User avatar
Red Eye
Grumpy Old Man
Posts: 4413
Joined: 07 May 2006, 07:12
Location: Republic of Mancunia

Re: FPL Skill vs. Luck: An Epic Debate

Post by Red Eye »

Ok (kind of) follow you. The point I think I was hinting at was that using a control chart gives you rules you can apply to differentiate between random cause variation (i.e. nothing happened - that score was within expected range) and 'assignable cause variation' (something may have happened - investigate). The rules are...

1) Outside Control Limit (3 standard deviations from mean (I think)) i.e. for Hazard to score 22+ tonight then something exceptional will probably need to happen, like he plays striker and/or Leicester have 2 players sent off.
2) Run of eight consecutive scores above or below mean - that's significant so you may want to re-assess expected returns. (obviously in football you wouldn't actually wait 8 GWs if someone broke their leg before deciding something significant happened :) )

Now over the course of the first 7 GWs this season more than 200k people transferred Hazard out. Statistically there was nothing significant about his scores over that time so really there was no good reason to do so. I can only assume people must have seen bigger scores from others, falsely attributed some significance to them and decided other players were 'better' than Hazard (i.e. Di Maria, Sanchez, whoever). It seems what many people do is chase their tails hopping on and off players when doing nothing/sitting tight would be just as good, if not better.

Thing is we can't help it - we want to take that punt and we delude ourselves when the boat comes in that 'yes I knew it'. So the 'skill' might not lie in making more good/'lucky' decisions but in limiting the number of decisions and therefore the number of bad ones.

The Dazzler
Grumpy Old Man
Posts: 1248
Joined: 03 Dec 2008, 20:26
FS Record: 9th overall in FPL 2005/06, 50th 2010/11, 288th 2014/15

FPL Skill vs. Luck: An Epic Debate

Post by The Dazzler »

Carlos Kickaball wrote:
The Dazzler wrote:
Carlos Kickaball wrote:Some really good points raised recently, and worth continuing. I've seen a lot of straw man arguments in this thread, some people seem to think that those saying that a mixture of skill and luck are needed to succeed in FPL, are really saying skill is not involved or that all managers near the top are of exactly the same ability. That is not what is being said, well certainly not by me.
The Dazzler wrote:
Valeron wrote:I'm saying the following.

Ville is a skilled player
Ville has got way more than his share of the breaks over a number of seasons.
No. Prove it. The most likely thing is he is doing as well as he should be.
Actually Dazzler, your logic comes unstuck here, that is not the most likely thing and you are failing to interpret the evidence properly.

I think we all agree that both skill and luck play a part in FPL performance, and the debate centres around how much it plays a part. If you accept that both skill and luck can influence performance, then by picking the highest performers, you not only get the some of the most skilful, but also the some of the most lucky. Incidentally you could also fail to find either the most lucky, because as you pointed out the most lucky manager could be a poor one, and you could fail to find the most skilful, as the most skilful manager may not have been particularly lucky and have a good rather than outstanding record.
Unsurprisingly, I disagree.
Your statement above implies you are giving equal weight to skill and luck.
Actually my statement doesn't imply anything about some ratio of skill to luck, and would be true so long as skill and luck both contribute towards performance. In fact I think giving some percentage of skill and luck without a timeframe is actually pretty meaningless. You seem to be taking things from my words that aren't really there, which I admit has been a problem for a lot of your straw man arguments.
I think any reasonably minded person could see how stating;
"If you accept that both skill and luck can influence performance, then by picking the highest performers, you not only get the some of the most skilful, but also the some of the most lucky."
could be interpreted as you implying that you are valuing skill and luck equally.
"Some of the most skilful", "some of the most lucky". It's not explictly stated, it's implied. It's certainly not pointed out, as it should be, that there is a huge gap between the impact of those 2 variables. But then that wouldn't suit your agenda, hence that's why it's not mentioned.
And of course it's all vaguely worded, as is your wont, to allow easy deniability.
In any case it's not a point I want to labour but it's certainly not a strawman by me. By any reasonable minded persons standards.


"I've seen a lot of straw man arguments in this thread, some people seem to think that those saying that a mixture of skill and luck are needed to succeed in FPL, are really saying skill is not involved or that all managers near the top are of exactly the same ability."

Accuse others of creating strawmen and in the same breath, create your own strawman argument which consists of, "some people seem to think", followed by some distortion of a (or various) posters words ("are really saying"), which fits your own purpose.
'Some people seem to think...' or 'some posters seem to be saying....', and people "are really saying..."
All vague wording, not attributed to any particular poster or post so can't be easily disputed.
This is a common tactic of yours which we see over and over again. Quote what people are actually saying rather than this constant misrepresentation please.
Carlos Kickaball wrote:I'll try and explain a little clearer to you.

For example (and this is I hasten to add this paragraph is an example so we are assuming these things are true) imagine that two of the largest factors for a tree's height after 50 years, are it's genetics, and how much sunlight it gets, and with better genetics and more sunlight both causing a tree to grow taller however one may be a more important factor than the other, and these are both variable factors (that is trees differ in their genetics and the amount of sunlight). If you get a sample of 50000 trees all 50 years old and find the tallest one, you'd be right to assume that it was likely to be ones of the best genetically but also one of the ones that had the most sunlight. If someone said "oh, it must be all genetic and there is no reason to assume that one tree had more light than the other they probably just got the average amount" they'd be wrong.

In the FPL setting, if you know two factors of performance are luck and skill, if you take the best performer then it follows that they are likely to be one of the best for skill but also one of the ones that have had the most luck. This should be obvious so if you still haven't understood I'll leave it there.

What is the point of an analogy about trees whose growth depends on 2 valueless variables? I don't know who is arguing against the scenario you are describing. Is it, "some people..."?
I have assigned values for our 2 variables. I have said 90/10 for the skill/luck ratio, the exact ratio can be debated. And yes, I meant over the course of a season, as I stated, "The most likely thing for every player is that they are doing as well as they should be, plus or minus a 10% margin in any one season."

So in your strawman, yes it is obvious that the best performer should score high for skill and for luck. Strawman obliterated. Victory is achieved over the imaginary enemy. Well done, good job!

However back in the real world, where luck has a much lower value than skill and where 7 seasons have passed thus diluting the luck element further and where you have not disproved Stemanias theory that lesser skilled players are more likely to experience the variance attached to luck, it isn't so obvious.

But even if we accept your 'the best performer is the best guess for best player AND best guess for most lucky', it STILL doesn't really have any significance.
Over 7 years, the best performer is the best guess as best player. He may be the best player, he may not be. But he's almost certainly in the top 100 of best players.
Over 7 years, the best performer is the best guess as most lucky. But he's the best guess from 3.5 million players, with very little difference between all their claims. Certainly not the same as best guess among 100 players. Therefore the luck is statistically insignificant and irrelevant.
And further to that, even if it was statistically significant, we still can't say that he definitively has been lucky. We can suggest it's possible but it's possible he's been unlucky too (although less likely).
Carlos Kickaball wrote:All I am really saying is to have a record so sensational, it likely requires skill and luck, even if one is a larger factor than another. Reasoning has been explained, and I think regardless of the ratios most agree.
The Dazzler wrote:Ville, in his 7th season, is very likely doing as well as he should. Deal with it.
:lol: :lol: :lol:
And I'm saying luck is likely negligible after 7 years.
And Ville, in his 7th season, is very likely doing as well as he should.
You think most people agree with you? I'd be surprised if that's true. But I'm not surprised at all that you think it. :D
Carlos Kickaball wrote:
On luck, while most of us will regularly beat more casual managers, I actually think it's a reasonable factor between serious players, illustrated by the fact that quite a few of us are ahead of Spiderm4tt and Ville this season. You should almost always finish in the top 20% of players, but top 5k and above finishes don't happen if you're unlucky.
The above is a quote from October, hence why Carlos was ahead of superior players that he now trails behind.
I see you're not ahead of Ville or Spiderm4tt now. Perhaps the luck levelled out? Regression presumably?
But your comments about, "most of us will regularly beat more casual managers", "always finish in the top 20%" and you 'don't get top 5k if you're unlucky'. Where is this coming from?
You don't regularly beat anyone. You've never even finished in the top 20%.
3 completed years and not even close! When you say, "most of us", you do realise you're not in "us", right?
What do you know about what it takes to consistently hit top 5k?

If you were Ville coming in here all arrogant and giving it all Bertie Big Bollocks, then I'd lap up like the fanboy I am.
But your posting is of questionable quality and your playing record is worse so I'm certainly not inclined to take it from you.
Perhaps your ceiling is 5K, I'd don't think I'd be inclined to accept your ceiling when talking about accomplished players however.
Carlos Kickaball wrote:
The Dazzler wrote:What these guys do isn't magic, it is basically common sense. And yet, we can't match it.
If you think what these guys do is basically common sense, then I'd suggest that what separates players with very good records and the best records is often luck.

I have to say Dazzler, that with your own very good record, and that of some of the others on the forum we'd do better talking about our own reasons, teams, and decisions, especially as I imagine you do anyway. It would be good to see you on more debates about the merits of players, decisions and RMTs.
So it's all luck, luck, luck.
And let's stop talking about these lucky managers. Derail, derail, derail.
Carlos Kickaball wrote:The most amusing thing is that some people who have played the game for years, would deny that luck can affect performance, especially when there is clear variance over their record. :roll: :lol:
I don't know if this eye rolling bewilderment was aimed at me? It's hard to know what with the, "some people..." part. You know the way you keep doing that? If so, my poor years were as a result of poor play, not poor luck. So sorry, I don't help your little theory. But this is yet another example of a strawman, as I never said that luck didn't affect performance.
Carlos Kickaball wrote:I do actually think that most good managers on here are not particularly different in ability to Ville, and that for a very good manager finishing in the top 5-10k is probably achievable with around average luck over a season.
Okay, here we have FPL and mathematical expert Carlos stating that he thinks "that most good managers on here are not particularly different in ability to Ville".
"Most good managers on here" is probably a goodly number. Is that 20 people? Roughly as good as Ville?
This is despite the fact that I, BigMon, Stemania, SutterKane, Gooberman, Silkworms (sorry for any I've left out) have all stated we're definitely not as good as him. I can only think Hancock and Spiderm4tt can have any claim. And that wouldn't be "most good managers on here", would it? Actually now that I think of it, you're probably including yourself in this group of "most good managers on here are not particularly different in ability to Ville"?
You are, aren''t you? You think you're comparable to Ville, right? :shock:

Next Carlos makes this ball park claim that very good managers can expect 5-10k finishes with average luck. Which leads us to the following comedy goldmine.
Carlos Kickaball wrote:It is worth remembering, that to avoid bad finishes, you need to simply avoid being unlucky, as these seasons will come with quite large ranks, as due to the distribution of points scores in the game, bad luck seasons will add much more to your rank, than good luck seasons do.

You don't need to be lucky as a top 10k manager to finish in the top 5k, but you do need to not be unlucky, (with a 50% chance). If we see that over 6 years Ville has finished twice around this position and four times higher (which for each of those years I'll say he was in the top 10% for luck) then the probability of that happening is:

(0.5^2)*(0.1^4)*6C2 = 0.000375

[Where 6C2 means 6 choose 2, or 6!/(2!*((6-2)!))]

Anyhow that is roughly 4 in 10k, so in essence we can see that it is likely that even if there are 10k managers as good as Ville, a record as good as his is can and is even highly likely to exist. The point is Ville's record is excellent and great evidence of his personal ability as a manager, and even claim to be one of the best in the world, but it doesn't pass as evidence that certain managers are much better than almost anyone else.

I'm not saying that you need to be lucky to do well at all, of course good managers will accrue good records, and bad managers won't, a real casual will struggle to have multiple top 10k finishes as many on this site do. Any individual with a good record can rightly be proud of it, without having to evidence better records, and insisting their is little luck involved.

"It is worth remembering, that to avoid bad finishes, you need to simply avoid being unlucky".
What? We need....I don't even......what? How about picking good players?

"You don't need to be lucky as a top 10k manager to finish in the top 5k, but you do need to not be unlucky"
This presupposes that the 10,000th player* is roughly as skilled as the 1st ranked player.

"If we see that over 6 years Ville has finished twice around this position and four times higher (which for each of those years I'll say he was in the top 10% for luck) then the probability of that happening is:"
Ok this strawman says, now listen up as this will blow your mind.
There are 10,000 players as good as Ville. Yes, 10,000 of them. Ummmmm, okay.
Therefore Ville has a 50% chance of top 5,000. He has done roughly this twice (although a bit higher than 5k so even more lucky) but he has totally flukily smashed the speed of light luck barrier 4 times due to being in the top 10% of luckboxes those 4 seasons. Obviously this will have to be updated due to Ville being an incredible luckbox yet again this year for the 5th time in 7 seasons.
And the highest ceiling a top player can achieve is equal top 10,000th. Therefore, anyone finishing higher than that, even the equal strength top 10,000 players, is experiencing luck. And that's a fact. Apparently. Here's some jazzy looking maths, which 'proves' it.
Ville has finished in the top 5k 7 years in a row and well exceeded it 5 times so now his luck factor is probably something like, "super duper lucksack 5 star general" rating. Which I believe is as high as the highly scientific rating system goes.
So because Carlos believes his ceiling is 5k, he believes all players ceiling is 5k. He is unwilling to accept that Ville figures are non luck related. Therefore he suggests that anytime Ville gets better than 5k, he is in the top 10% of luck, which is his randomly assigned luck number he pulled out his ass. Or something.
People, this guy actually thinks this stuff he is saying is scientific! Wait, there's more :)

"even if there are 10k managers as good as Ville, a record as good as his is can and is even highly likely to exist."
That's right. If there are even 10,000 managers as good as Ville. That's all. Only 10,000 players as good as Ville. That's what he said. It's right there. You're reading it correctly.
Do you know 10,000 Villes? Go out and find them. They have got to be all over the place.
They are as skilled as Ville but some of them are just very, very unlucky. You know the guy who was "going to transfer in N'Doye as captain" this week but wasn't able to as his internet went down? That guy. Yeah, stuff like that happens to him every, single week. For the last 7 years. Worlds unluckiest guy. Brilliant at FPL though. He'll crack 500K one of these years.

Okay, back to the real world. There isn't 10,000 Villes. There just isn't. There might be 100 of them, maybe even 200. But I don't think so.
*Here is the HOFs 10,000th ranked manager btw;
http://fantasy.premierleague.com/entry/630404/history/
Having a blinder of a season and on for a personal best but I don't think he's quite in Ville territory!
Carlos Kickaball wrote:To be honest I just state what I believe to be true, and back it up with reasoning, you'll see my first post on the initial page on Ville's record.
Carlos Kickaball wrote:It's a good record, but if you look at enough player's records, you'll eventually find one that good.
And that was made before the thread strayed into guessing managers motivations for their decisions in a hope to learn about the optimal FPL style, and people comparing their personal lives with hypothesised ones of Ville. It is quite amusing that someone would make the drive of the thread to learn of Ville's decisions, and then tell us that Ville has made a mistake when he veers from his own personal preferences as a manager. :lol:

Firstly, I didn't make "the drive of the thread" anything. I posted what I thought would be valuable and that was a commentary of Villes moves.
If people saw that as the drive of the thread, then perhaps that was because it was well written commentary from an accomplished FPL player and it was deemed of some value to the community.
As opposed to the poorly thought out and argued, agenda driven, mantra that FPL excellence was more a by product of luck than of skill. Perhaps people found that line boring and unconvincing after the 30th time they heard it.
Perhaps, but others would have to be the judge of that.

A well respected poster pm'ed me about "someone" once and said;
"if he could cut out all the points-scoring, arguing, I-told-you-so's, trying to look correct all the time, refusing to back down, he said it first mentality, condescending tone, and ego type stuff, then I think there's something there of value."
Quite so.

Carlos Kickaball wrote:Great post Sween, good to have some very constructive ones, and I would agree with most of what you've said, and have expressed similar in another topic for a while.

This is essentially the central point (to me anyway):
sween wrote:So when it comes to seperating say the top 1000 or so players in the FPL world, over a season the definitive factor will be standard deviation (or luck).
The real questions comes in how large this group can get until the managers we start to include are much worse, and how different in skill are say the 1000th, 10000th, and 50000th best managers, to the very best.

When evaluating the effect of luck it's clearly erroneous to selectively chose one of the most uniform records, as some seem to have done on here (for by example by choosing the best), and it would be much better done looking at a larger sample of records, perhaps even just the ones of active forum members.

"When evaluating the effect of luck it's clearly erroneous to selectively chose one of the most uniform records, as some seem to have done on here (for by example by choosing the best)...."
Ummmmm, what!!!!!!!! Yes dear reader, it was he himself who did so.
Stemania, quick as a flash, pounces on the error.
Stemania wrote: The posters who picked out Ville's record were not trying to evaluate the effect of luck.
Back comes Carlos.
Carlos Kickaball wrote:It was selected due to how good it was, and then used in this way.
Stemania doesn't let go of his prey and closes in for the kill.
Stemania wrote:As you well know, this thread was a splitting off from the Discussion of genius players thread. People brought up Ville's record there due to it's unusually high standard. A discussion of luck framed around Ville's individual record was essentially initiated there by you, so to now critisise others for including Ville's record in the discussion seemed a bit cheap to me, 'tis all. ;)
Well that's pretty conclusive I think.
Carlos Kickaball wrote:Nope I didn't solely use Ville's record to make my conclusions about the amount of luck involved in the game in general. I looked at the variability across many records from good managers. A very cheap shot, both are misrepresentations of what I have said. Just as some facts and interesting debate was starting to happen too. I'd be happy for the last 4 posts to be deleted as they offer nothing to the debate.

Except yes, it was you that initiated the process that you now say is "clearly erroneous" and you can't just hold your hands up and say, "haha, guilty as charged. I made an error there."
There was the opportunity. A window where you can own the mistake. Even Nigel Pearson, who is a blithering idiot, took that window.
But no, you attack the poster that exposed you for a "cheap shot" and ask the mods to delete all the evidence. Un-blooming-believable!
Completely incapable of just saying, "I was wrong. I apologise." When you were wrong. And should apologise.


Carlos Kickaball wrote:It comes back to:
Red Eye wrote:1) The environment is sufficiently regular to be predictable.
Some of these posts seem to imply that vast quantities of statistical data, being able to watch every minute of every match, being a football mastermind, being a superb statistician, will give you a big prediction edge over in these close calls, over someone who is an avid Premier League football fan, that watches some games and highlights, and has access to all the information in the online online stats, odds, forums and media. I just don't think that is true, once you get to a certain point the extra data becomes quite superfluous, especially with how many random factors there are. We just don't have models and football experts who can really tell you who will get a clean sheet, we just have rough ideas.

In something like poker or chess it is very different, players have to make decisions themselves, and the calculations and skills needed to make the best decisions are much higher. There will be things that seem more obvious to a grandmaster, than a committed club player, and the grandmaster can explain and demonstrate why. But I don't think this is true for FPL, the guess of a well informed committed reasonable FPL player who has access to the same things we do, is not going to be much worse than a expert football obsessed statistician who watches every game and has UEFA coaching badges. If you go onto a chess or poker forum, you'll probably observe a consequence of this, which is that the ideas being shared in comparison to an FPL forum will be much more complicated.
So this is yet again another "some of these posts seem to imply....", followed by some hyperbole that nobody ever implied, followed by a conclusion.
This conclusion? FPL is less complicated than chess and poker, therefore consistently doing well in FPL is luck.


So what is this agenda to continously and relentlessly insist that anyone with a record as good as Ville MUST be lucky?
What's that about? Right from the word go, when Villes record was first posted, Carlos started with his, "yeah but lucky", routine. And he has never stopped, it's an obsession. What does that achieve?
Mav3rick wrote: Is the real reason for labelling Ville as lucky to be able to say that you don't have to have a good record to be seen as a good manager? That's a sentiment I agree with anyway, but 21 pages to get there seems superfluous! Anyway I have to pretend to do some work now but will check back in later on.
Bingo!
There it is.
Because "some posters" might have a very, very poor record. And if all good records can be diminished by suggesting that they have a high liklihood of positive variance, then the poor records can be similarly enhanced by suggesting they may have a possibility of negative variance. Or rather, it dilutes the value of all players records altogether.
And who does that serve? He who wishes to build a reputation not on the quality of his performance but on his ability to shout loudest and longest.







Have a look at these 2 random playing histories I've just pulled up completely randomly by chance. Randomly.
Player A
2008/09 2078 4266
2009/10 2466 569
2010/11 2225 362
2011/12 2243 1631
2012/13 2248 4668
2013/14 2534 90
2014/15 2069 439

Player B
2011/12 1510 1956011
2012/13 1737 1278367
2013/14 2035 902216
2014/15 1978 6460

Just based on their previous years rankings, which player do you think is the most likely to be experiencing extreme positive variance in their ranking this year?
Which guy is likely to have been super lucky this year, player A or player B?
Last edited by The Dazzler on 02 May 2015, 23:16, edited 1 time in total.

User avatar
Carlos Kickaball
Dumbledore
Posts: 7801
Joined: 04 Sep 2013, 18:02

Re: FPL Skill vs. Luck: An Epic Debate

Post by Carlos Kickaball »

And I thought I was doing okay this season. :lol:

I was actually a dead team for all of the previous years of my record, managing for about 4 months of a season once. If I knew it would be held against me I would have set up a new account this year. :D

User avatar
thekrumcake
FISOhead
Posts: 835
Joined: 25 Sep 2012, 16:15

Re: FPL Skill vs. Luck: An Epic Debate

Post by thekrumcake »

The Dazzler wrote: long post
This is the most awesome thing I've ever seen. How long did it take you to write that?!

:shock: seriously...

User avatar
Vid
Head Moderator
Posts: 21728
Joined: 13 Oct 2005, 18:33
FS Record: winning is a distant memory

Re: FPL Skill vs. Luck: An Epic Debate

Post by Vid »

Had a couple more reads through, the bulk of it is within acceptable limits of attacking the post(s) rather than the poster but there are a few areas that cross that line, I'll be requesting The Dazzler via pm to make a few changes so that it is all within (or at least close to) forum boundaries.

Notned
FISO Knight
Posts: 11198
Joined: 13 Sep 2013, 12:30

Re: FPL Skill vs. Luck: An Epic Debate

Post by Notned »

Bearing in mind I had 5 points today prior to Myhill against United, and Schlupp coming in off my bench, I think it's fair to say which side of this particular fence I currently sit on...

View Latest: 1 Day View Your posts
Post Reply

Return to “Fantasy PremierLeague.com (FPL)”