Discussion:
Fantasy Scout - Contest: post archive stat, win T-shirt
(too old to reply)
milivella
2024-05-27 02:02:13 UTC
Permalink
I added the data from all seasons in the archive pages:
http://fantasyscout.altervista.org/rankingsarchive.htm
http://fantasyscout.altervista.org/playersarchive.htm
http://fantasyscout.altervista.org/predictedheadtoheadsarchive.htm

In order to celebrate this fact, I am launching a contest: **whoever
posts the most interesting stat based on the data in the archive pages
will win a personalized T-shirt**.

Some details:

* This is the infamous May 27th announcement. But you already knew from
the hints, didn't you? :D

* The contest will end in a week.

* Anyone can enter the contest, even those who do not play Fantasy
Scout.

* You can post your stats either on
[rec.sport.soccer](https://www.novabbs.com/sport/thread.php?group=rec.sport.soccer)
or on [the Fantasy Scout
subreddit](https://www.reddit.com/r/fantasyscout/).

* Stats can have any form: a sentence, a table, a narrative, etc.

* You can post as many stats as you want whenever you want (before the
deadline, of course).

* I will be the only judge.

* If I will find two stats equally interesting, the one posted earlier
will win (so start posting as soon as you can!).

* If you have any question, ask!

--
Cheers
milivella
milivella
2024-05-27 02:07:10 UTC
Permalink
Post by milivella
* If you have any question, ask!
Let me anticipate a possible question: **what is the absolute minimum I
must know to use these pages?**

Here are the four basic features of the archive pages (the first two of
which are common to all tables on this website):

1. If you want to **display only the rows that contain a given value**,
write that value in the box above the column header and hit Enter. E.g.,
to display only the rows involving Jesus, write "Jesus" (without the
quotes) in the box above "Scout" and hit Enter.

2. If you want to **sort the rows on the basis of a column**, click on
the heading of that column. E.g., to check who had the highest predicted
score ever, click on "Prd" twice (to sort first in descending, then in
ascending order).

3. **The "+" columns** contain the difference between (a) the value on
the column on the left and (b) the value on the column on the left one
year ago. E.g., if you see -1 in the "+" column next to "% Win", it
means that in the last year that scout lost 1% chances of winning that
cycle; if you see 2, they gained 2% chances.

4. If you want to know **how many points a scout scored in a given
season**, consider that "points scored in a given season" is the same as
"points at the end of the season (= on July 31) minus points at at the
beginning of the season (= on July 31 of the previous year)", which is
the value you find in the "+" column next to the "pts" columns (see
previous feature). E.g., if you want to know how many points the players
picked by Alessandro in cycle 1 scored in the 2007-08 season, check the
"+" column next to "Player pts" in the Year 2008 - Cycle 1 - Scout
Alessandro row.

Again, these are just the basic features of the page. You can do more.
To know how, when you feel like it, read the questions and answers on
the archive pages (or ask me here!).

--
Cheers
milivella
milivella
2024-05-27 02:08:20 UTC
Permalink
Just to get the ball rolling, here are some hors concours stats.

(I will be a bit verbose, and I will tell you how to compute each stat.
You don't have to do either of those things.)



# Best player at each age

Who scored more points in the season in which they turned 17? and 18?
etc.

[How to compute the following stat: 1. Visit playersarchive.htm 2. Click
twice on the "+" next to "Pts" 3. Write the age above "Age" and press
Enter]

| Age | Player | Points |
|----:|:----------|-------:|
| 17 | Gavi | 11 |
| 18 | Gavi | 13 |
| 19 | Mbappe | 26 |
| 20 | Rashford | 19 |
| 21 | Neymar | 33 |
| 22 | Neymar | 24 |
| 23 | Schurrle | 24 |
| 24 | Mbappe | 26 |
| 25 | Griezmann | 24 |
| 26 | Pedro | 24 |
| 27 | Griezmann | 26 |
| 28 | Depay | 26 |
| 29 | Giroud | 26 |
| 30 | Griezmann | 25 |
| 31 | Giroud | 21 |
| 32 | Muller | 15 |
| 33 | Busquets | 11 |
| 34 | Giroud | 20 |
| 35 | Di Maria | 14 |
| 36 | Giroud | 18 |
| 37 | Neuer | 4 |
| 38 | Mandanda | 1 |

OK, but this is heavily influenced by WC/Euro/Copa seasons. How can we
account for that? One way is to divide the number of points of each
player by the number of matches his NT played that season. We have these
data in the table, in the "Pts gain / NT matches" column. Let's check
this score then. We'll have to exclude the 2019-20 season, though, as
with just 6 matches played because of COVID-19, it would distort the
ranking.

[How to compute the following stat: 1. Click twice on "Pts gain / NT
matches" 2. Write the age above "Age" and press Enter]

| Age | Player | Points |
|----:|:-----------|-------:|
| 17 | Gavi | 0.85 |
| 18 | Gavi | 1.18 |
| 19 | Mbappe | 1.44 |
| 20 | Mbappe* | 1.45 |
| 21 | Neymar | 1.65 |
| 22 | Neymar | 1.41 |
| 23 | Mbappe* | 1.46 |
| 24 | Mbappe | 2.00 |
| 25 | Depay* | 1.70 |
| 26 | Pedro | 1.50 |
| 27 | Depay* | 1.47 |
| 28 | Depay | 2.00 |
| 29 | Kane* | 1.85 |
| 30 | Kane* | 1.73 |
| 31 | Griezmann* | 1.31 |
| 32 | Giroud* | 1.27 |
| 33 | Blind* | 0.91 |
| 34 | Giroud | 1.18 |
| 35 | Di Maria | 1.00 |
| 36 | Giroud | 1.38 |
| 37 | Neuer | 0.36 |
| 38 | Mandanda | 0.08 |

* = different from the previous table



# Best scout performance in each relative year

Which were the best performances at the beginning of a cycle? And in
later seasons? To answer this kind of questions, we can use the concept
of relative years. Relative year = years since the end of the picking
phase of that cycle. So -1 = 1 year before the end, 3 = 3 years after
the end, and so on.

OK, but what measure of performance should we use?

* Scout points? No, as it is measured in different ways in cycle 1.

* Then points scored by the scout's player? No, as in some seasons more
matches are played, so the performances in those seasons would be
unfairly favorited (compare e.g. a World Cup season vs. 2019-20, with
the COVID-19 stop).

* Then, we should divide those points by the number of matches played by
NTs.

Summing up, here "performance" is:

points score by the scout's player / average number of matches played by
that cycle's NTs

I.e., for each round of international matches, how many points did that
scout's players score? That's the figure in the "Player pts gain / avg.
no. of NT matches" column.

[How to compute the following stat: 1. Visit rankingsarchive.htm 2.
Click twice on "Player pts gain / avg. no. of NT matches" 3. Write the
equals sign and the relative year above "Rel year" and press Enter]

| Rel year | Scout | Cycle | Performance |
|---------:|:-------------|------:|------------:|
| -3 | Andrea V. | 1 | 3.90 |
| -2 | Daniele | 1 | 1.50 |
| -1 | Leon | 8 | 3.64 |
| 0 | Michael H. | 1 | 5.42 |
| 1 | Jackson | 5 | 7.19 |
| 2 | Daniele | 4 | 7.21 |
| 3 | Jackson | 6 | 6.94 |
| 4 | Daniele | 4 | 7.33 |
| 5 | Jackson | 5 | 5.70 |
| 6 | Andrea V. | 3 | 4.50 |
| 7 | Andrea V. | 3 | 5.24 |
| 8 | Andrea V. | 3 | 6.17 |
| 9 | Andrea V. | 3 | 3.78 |
| 10 | Tom Two | 2 | 2.13 |
| 11 | Michael H. | 2 | 2.43 |
| 12 | Sid Debgupta | 1 | 1.60 |
| 13 | Sid Debgupta | 1 | 1.71 |



# The turning point of the 3rd cycle

Don't know about you, but something in that last table caught my eye...
Let's delve a bit into the story of the 3rd cycle.

Let's go back to July 31st, 2016.

[How to compute the following stats: Write "2016" above "Year", "Andrea
V." above "Cycle", and press Enter]

Picks for this cycle have ended two years ago. Andrea V.'s players
scored just 2.05 points per match in the seasons that has just ended,
4th best performance in the cycle, a little more than *half* the points
scored by the players of leader Daniele. Moreover, Andrea V.'s players
scored 1.13 points *fewer* than the season before---the worst
deceleration of the cycle this season. On this basis, Andrea V.'s
predicted score being the 4th best looks fair, maybe even optimistic
(parenthetically, let's note that the ranking for the top 5 performances
of the year is the same as for the top 5 predicted scores).

What happens next:

[How to compute the following stat: 1. Click on the "Clear filters"
button on top right 2. Click twice on the "++" on the right of "Player
pts gain / avg. no. of NT matches"]

* In 2016-17, Andrea V. scores 6.62 points per match. This is 4.57
points more than the season before, the strongest acceleration *ever*.

[How to compute the following stat: 1. Write "3" above "Cycle" and press
Enter 2. Click twice on "Player pts gain / avg. no. of NT matches"]

* In the seasons from 2016-17 to 2021-22, Andrea V. has *the six best
performances in the history of the cycle*. (Fun fact: the best 9
performances are all from 9 different seasons, indeed from successive
seasons: Daniele 2014-16, Andrea V. 2017-23.)

The result is that Daniele will win the cycle, but not without getting
scared beforehand.
Futbolmetrix
2024-05-27 09:04:50 UTC
Permalink
Post by milivella
Just to get the ball rolling, here are some hors concours stats.
I.e., for each round of international matches, how many points did that
scout's players score? That's the figure in the "Player pts gain / avg.
no. of NT matches" column.
Needlessly complicated analysis snipped. Why don't we just report which
scout/cycle had the highest Player pts gain/ avg. no. of NT matches"?

The answer:

Year Cycle Scout Pts gain/match
2020 4 Daniele 7.33
2018 4 Daniele 7.21
2019 5 Jackson 7.19


But here is the question that I would really like to ask (and
unfortunately your archive can't answer): who wins the Santi Cazorla
award? That is, which player has the highest number of FS points
relative to NT matches in which he *was in the starting lineup*? (let's
say, minimum 10 starts?)
milivella
2024-05-27 10:56:08 UTC
Permalink
Post by Futbolmetrix
Post by milivella
Just to get the ball rolling, here are some hors concours stats.
I.e., for each round of international matches, how many points did that
scout's players score? That's the figure in the "Player pts gain / avg.
no. of NT matches" column.
Needlessly complicated analysis snipped.
Translation for those who are not fluent in futbolmetrish: "needlessly
complicated" = "showing I had just three good seasons in the only cycle
I managed to pick semi-competently"...
Post by Futbolmetrix
Why don't we just report which
scout/cycle had the highest Player pts gain/ avg. no. of NT matches"?
Year Cycle Scout Pts gain/match
2020 4 Daniele 7.33
2018 4 Daniele 7.21
2019 5 Jackson 7.19
How magnanimous of you to include the 3rd place!

Now this is a good example of minimum effort stat. Practically any
attempt will beat it! :D

(But thank you for breaking the ice!)
Post by Futbolmetrix
But here is the question that I would really like to ask (and
unfortunately your archive can't answer): who wins the Santi Cazorla
award? That is, which player has the highest number of FS points
relative to NT matches in which he *was in the starting lineup*? (let's
say, minimum 10 starts?)
I recognize nerd sniping when I see it... Let's say that if you win, I
_may_ get you both the T-shirt and an answer to this question. :)

--
Cheers
milivella
milivella
2024-06-01 02:32:17 UTC
Permalink
Post by Futbolmetrix
Year Cycle Scout Pts gain/match
2020 4 Daniele 7.33
2018 4 Daniele 7.21
2019 5 Jackson 7.19
Futbolmetrix made the economist/game theorist thing and made the minimum
bid. Do you *really* want to live in a world where this kind of moves
wins contests?!

--
Cheers
milivella
milivella
2024-06-03 02:56:18 UTC
Permalink
Post by Futbolmetrix
Why don't we just report which
scout/cycle had the highest Player pts gain/ avg. no. of NT matches"?
Year Cycle Scout Pts gain/match
2020 4 Daniele 7.33
2018 4 Daniele 7.21
2019 5 Jackson 7.19
Congratulations (?!), you won. Tell me your T-shirt size and home
address privately, and I'll send you the prize. I non-ironically think
you'll enjoy it!

--
Cheers
milivella

Loading...