Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Unsophisticated Prediction Contest 2007-08 #1

3 views
Skip to first unread message

milivella

unread,
Feb 1, 2008, 8:54:22 AM2/1/08
to
Quick and (I hope) entertaining game:
- Every (?) Friday, I'll propose a match from one of the "top 3"
leagues whose result is (I hope) difficult to predict. Obviously,
anyone can propose other matches, from different leagues as well.
- Before the match begins, everyone can tell which team will win.
- If your team wins, your prediction was correct.
- At the end of the season, the rsser who had, at some point of the
season, the longest streak of correct predictions will win. No
prediction or a draw doesn't break the streak; only the fact that your
team loses breaks the streak.

First match:
Wigan - West Ham

Who will win?

[My pick:West Ham.]

Jesus Petry

unread,
Feb 1, 2008, 11:37:33 AM2/1/08
to

West Ham.

Tchau!
Jesus Petry

Miguel

unread,
Feb 1, 2008, 12:06:09 PM2/1/08
to
On 1 feb, 14:54, milivella <milive...@gmail.com> wrote:

> First match:
> Wigan - West Ham
>
> Who will win?

Wigan

Sid

unread,
Feb 1, 2008, 12:10:33 PM2/1/08
to
milivella <mili...@gmail.com> wrote in news:baf8887b-4af8-461a-9c7c-
a7fedf...@b2g2000hsg.googlegroups.com:

Great idea!


> First match:
> Wigan - West Ham

> Who will win?

West Ham!

Sid

Futbolmetrix

unread,
Feb 1, 2008, 12:27:06 PM2/1/08
to

"milivella" <mili...@gmail.com> wrote in message
news:baf8887b-4af8-461a-9c7c->

> First match:
> Wigan - West Ham

Can you predict a draw?

If so, then: Draw

D


MH

unread,
Feb 1, 2008, 12:35:46 PM2/1/08
to

Nobody. I pick a draw.

>
> [My pick:West Ham.]

milivella

unread,
Feb 1, 2008, 1:00:04 PM2/1/08
to
> Can you predict a draw?

I think this would make the game less funny (and fun is the basic
effect I wanted to reach with the original rules, because it's clear
that it's not the fairest of the games, or the one that tell more
about your football knowledge). So, I wrote that you have two choices:
the home team or the away team, no draw.

But if the majority of people prefer to have this third choice, the
draw (MH predicted a draw too), it's OK to me.

I'd say: for this first week-end, no draw. Let's discuss this option
until the next Friday. :)

Cheers

milivella

Lléo

unread,
Feb 1, 2008, 3:33:35 PM2/1/08
to
"milivella" <mili...@gmail.com> escreveu na mensagem
news:baf8887b-4af8-461a...@b2g2000hsg.googlegroups.com...

I pick West Ham too.

--
Lléo


MH

unread,
Feb 1, 2008, 5:06:50 PM2/1/08
to

In that case I predict a Wigan win.

>
> Cheers
>
> milivella

Clément

unread,
Feb 1, 2008, 7:04:19 PM2/1/08
to
On Feb 1, 11:54 am, milivella wrote:
> First match:
> Wigan - West Ham
>
> Who will win?

Wigan.

Abraço,

Luiz Mello

Mark V.

unread,
Feb 2, 2008, 1:59:21 AM2/2/08
to

I'll pick West Ham since you won't allow me to pick a draw.

milivella

unread,
Feb 2, 2008, 6:07:36 AM2/2/08
to
> I'd say: for this first week-end, no draw. Let's discuss this option
> until the next Friday. :)

I'd like to explain why I chose the rules I proposed to you all, and
specially the "no draw prediction". My main objective, as I've wrote,
was to have an *entertaining* game. This explains some of my choices:
focusing on one match only per week; selecting a balanced match;
letting people that enter the competition later or don't always play
have good chances to win; making every prediction a possible first or
next step in a winning streak.

The entertainment factor is behind the "no draw prediction" rule as
well: the game as it is forces you to support a team, that is
something - I feel - far more involving than "supporting" a draw. Yes,
I know I've named the game "_prediction_ contest", and I wrote about
correct "predictions", but I wanted more a game about supporting
("situationist" supporting that lasts only a match, but still
supporting).

This said, I too have a doubt: choosing balanced matches, there could
be too many draws. In the top 3 leagues, draw is the outcome of less
than one third of the matches (acceptable to me), but this figure
includes unbalanced matches as well (e.g. United-Derby). Will there be
too many draws? Would we feel this as not-so-entertaining? (1)

Please, anyone interested, write down your opinions about this. Do you
want draw prediction too next week?

(1) An alternative could be to list every week some weak teams that go
against stronger ones: you should pick a weak team, hoping it will
beat the odds (drawing or winning). E.g. this week I would have
chosen:
Sunderland (Liverpool away)
Empoli (Inter away)
Cagliari (Juventus away)
Osasuna (Barcelona away)

Cheers

milivella

milivella

unread,
Feb 2, 2008, 3:36:02 PM2/2/08
to
> My main objective, as I've wrote,
> was to have an *entertaining* game.

Maybe "exciting" is more correct.

FairFootball

unread,
Feb 3, 2008, 7:32:02 PM2/3/08
to
I believe the no-draw option is OK. It doesn't favor anyone, since
nobody may pick it, and if it happens then nobody has either lost or
won. It's like asking "if somebody wins, who will it be ?" . As for
fearing there will be to many draws, I'd suggest you choose 3 or 4
games per week instead of 1. They have to be balanced though or else
it's like regular betting and you have to set quotas and it probably
gets to complicated.
My 2 pennies worth. I fear I won't participate, I'm hardly nearly as
good as others here at such things. But I wish you all good luck with
it.

FF

-------------------------------------------------------------------------
the best site all about ref errors
www.FairFootball.com

FairFootball

unread,
Feb 3, 2008, 7:35:19 PM2/3/08
to
I just thought: allowing a draw pick wouldn't be very good IMO because
since the games have to be balanced, a draw is a very likely outcome.
So, if somebody always takes the draw he has high chances of being the
best in the end, which is something you definitely don't want.

FF

milivella

unread,
Feb 4, 2008, 2:29:51 AM2/4/08
to
To FairFootball:

Thank you for your contribution!

> I believe the no-draw option is OK. It doesn't favor anyone, since
> nobody may pick it, and if it happens then nobody has either lost or
> won. It's like asking "if somebody wins, who will it be ?" .

You've expressed my thoughts better than I can.

> As for
> fearing there will be to many draws, I'd suggest you choose 3 or 4
> games per week instead of 1.

Great idea! I would go for 1 game per league: 1 from Premier League, 1
from La Liga, 1 from Serie A. I just hope that three matches are not
too many (i.e. nobody having correct predictions for two or more
weeks).

If nobody proposes different solutions, I would follow FairFootball's
proposal.

> They have to be balanced though or else
> it's like regular betting and you have to set quotas and it probably
> gets to complicated.

Exactly. E.g. last week, I would have chosen (together with Wigan-West
Ham) Fiorentina-Milan (a tough match, decided only by a great goal)
and Valladolid-Valencia (I read that it was "hard fought").

> I fear I won't participate, I'm hardly nearly as
> good as others here at such things.

:( What a pity. I hope you'll join the game, and Fantasy Scout too. If
I play - with my poor knowledge - play, anyone can! :)

Cheers

milivella

milivella

unread,
Feb 4, 2008, 2:33:02 AM2/4/08
to
To FairFootball:

Don't know about this. According to bookmakers, draw was not the most
likely outcome in the three matches I thought (see my previous post).
But maybe you're correct. In any case, to avoid such a scenario is
another reason to reject draw predictions.

milivella

unread,
Feb 4, 2008, 3:48:46 AM2/4/08
to
> I would go for 1 game per league: 1 from Premier League, 1
> from La Liga, 1 from Serie A. I just hope that three matches are not
> too many (i.e. nobody having correct predictions for two or more
> weeks).

Maybe two matches overall (not three) is better. Assuming that 1 match
in 3 is a draw, the chance of having two draws in two matches is 1 in
9: not so probable. And it's not too difficult to correctly predict 4
matches (2 in a week, 2 in the next one).

Anyone opinions about this?

Cheers

milivella

Jesus Petry

unread,
Feb 4, 2008, 10:31:30 AM2/4/08
to

Yes, I guess you should pick one Saturday match and one Sunday match,
so the streak can be counted timewise. If you get both matches right,
you add 2 to your current streak; if you get both wrong, you go back
to zero. If you get Saturday wrong but Sunday right, you go to 1
point; but if you get Sat right and Sun wrong, you go to zero again.

And if draws are a concern, you could add a rule that draws don't
zero your streak, but also don't add to it.

Tchau!
Jesus Petry

milivella

unread,
Feb 4, 2008, 11:11:13 AM2/4/08
to
To Jesus Petry:

> I guess you should pick one Saturday match and one Sunday match,
> so the streak can be counted timewise.

This is interesting.

> And if draws are a concern, you could add a rule that draws don't
> zero your streak, but also don't add to it.

This is the actual rule! Probably I was not clear in my original post.

So? If nobody has different proposals, I would go for Jesus's one (one
Saturday match, then one Sunday match) because he showed interest,
because it's exciting and because I'm in debt with this gentleman :)
I'd say: you can submit the prediction for Sunday match even after the
Saturday one.

Cheers

milivella

FairFootball

unread,
Feb 4, 2008, 9:30:47 PM2/4/08
to
milivella wrote:
> To FairFootball:
>
> Thank you for your contribution!

You're most welcome.

> Great idea! I would go for 1 game per league: 1 from Premier League, 1
> from La Liga, 1 from Serie A.

Yes, this would be also good since it would give people from at least
these 3 countries a good chance.

> I just hope that three matches are not
> too many (i.e. nobody having correct predictions for two or more
> weeks).

I don't think that's a problem. If somebody isn't confident enough on
a certain game's winner he'll just pass. And if he gets 2 or 3 games
right they all are added to his streak. It's more or less like having
one game per week but a 3-times longer tournament. It just gives you
more room to express yourselves and make a difference, if I'm making
myself clear.
One difference though - what happens on wrong predictions. I'd say the
simplest way to go is count the week's hits, if any, to the current
streak, which is then broken of course. The order of the games doesn't
count.

Jesus's proposal is also OK except I find it a little more
complicated, and it allows for only 2 games per week.

> If I play - with my poor knowledge -, anyone can! :)

Believe me, you are way better than me. :) I hardly watch any football
except the Champions League and national team tournaments. Pick 3
balanced games from 3 different leagues - this is far beyond my
wildest dreams, if I may say so. :)

FF

FairFootball

unread,
Feb 4, 2008, 9:39:37 PM2/4/08
to
Correction: after reading Jesus's proposal again, it's not more
complicated, it's actually having 2 rounds every week, in a given
sequence. So either proposal is fine; if you want 3 guesses per week
you can go with mine or pick a 3rd game and assign it a sequence.

thethi...@gmail.com

unread,
Feb 5, 2008, 3:09:15 AM2/5/08
to
On Feb 5, 12:39 pm, FairFootball <FAIRFOOTBALL....@domainsbyproxy.com>
wrote:

I just checked our your website, and there isn't another as full of
bollocks as yours that I've ever seen. Hope you didn't pay much for
that domain name, or the monthly hosting, lol.

milivella

unread,
Feb 5, 2008, 3:53:20 PM2/5/08
to
To FairFootball:

> So either proposal is fine

I agree. your proposal is more flexible (it's possible to have more
than two games a week), but I find Jesus' one more exciting, because
every match is "death or glory" (with your rules, it can happen that
the first match turns to be a wrong prediction, and the next ones can
be meaningless to you). So, I myself prefer Jesus' proposal. But,
again, if more people prefer FairFootball's one, it will become the
official rule. Please anyone interested express yourself.

Cheers

milivella

FairFootball

unread,
Feb 5, 2008, 9:00:12 PM2/5/08
to
milivella wrote:
> To FairFootball:
>
> > So either proposal is fine
>
> I agree. your proposal is more flexible (it's possible to have more
> than two games a week), but I find Jesus' one more exciting, because
> every match is "death or glory" (with your rules, it can happen that
> the first match turns to be a wrong prediction, and the next ones can
> be meaningless to you).

OK, don't take this to mean that I'm somehow trying to promote my
proposal against that of Jesus, as I said I think they're very
similar, and his one also can easily be made to accept more than 2
games a week. I just think I wasn't clear enough so it seems there is
some misunderstanding left. So, my rules would be:

1. Anybody can pass on any number of games in a week, including all of
them.
2. Any number of hits in a week's prediction are added to the current
streak.
3. Any number of misses in a week's prediction break the current
streak.

The rules are applied in this order; thus, if a week's prediction
contains both hits and misses, first the hits are added to the current
streak, then the streak is broken. (To be very very clear, this
doesn't depend on the order of the games. So, if the first game is a
hit ant the next 2 are misses, these are still added to the current
streak, then it is broken.)

So, you see it's not correct that if the first match is a miss, the
other ones don't count any more. As I think I've already said, in my
proposal the order of the games in a week is totally unimportant.

That said, I repeat, both proposals are fine and essentially alike.

BTW, personally I'm not sure this idea of having the winner given by
the longest streak is a very good one. It essentially leaves very much
room for luck. So, if say Jesus has 10 correct guesses in a season but
the longest streak happens to be 3 (very likely outcome), whilst
milivella has 5, of which 4 are in a row (just a hypothetical
example ;) ), who's the winner ? It may be somewhat more exciting this
way but it's hardly fair. I'd go with the old-fashioned approach of
hits minus misses.

FF

FairFootball

unread,
Feb 5, 2008, 9:06:39 PM2/5/08
to
FairFootball wrote:
> ... So, if the first game is a

> hit ant the next 2 are misses, these are still added to the current
> streak, then it is broken.

I don't know where my head was when I wrote this (meanwhile I checked
and it's back at its old place). It should read, of course: if the
first game is a miss and the next 2 are hits, these are still added to
the current streak, then it is broken. Sorry.

FF

milivella

unread,
Feb 6, 2008, 2:36:27 AM2/6/08
to
To FairFootball:

> OK, don't take this to mean that I'm somehow trying to promote my
> proposal against that of Jesus

No, we're just discussing to make the game better :)

> So, you see it's not correct that if the first match is a miss, the
> other ones don't count any more.

You were clear about this in your previous post already. It was me:
_I_ haven't explained my assertion. I wanted to say: assume that the
best streak, at a given point of a season, is 5, and you've a streak
of 1 hit before the three matches of the weekend start. If the first
match is a miss, you already know that, even if the two next matches
are hits, you can't go further than 3 hits in your current streak,
i.e. you can't reach the best seasonal streak so far (5 hits). This is
the reason I wrote "the first match turns to be a wrong prediction,
and the next ones _can_ be meaningless to you": it's a chance, not
every first-match-miss make the next matches meaningless.

> That said, I repeat, both proposals are fine and essentially alike.

As I've said, I agree. I hope anyone interested will show his
preference, so we'll be able to prefer one proposal or the other one.

> BTW, personally I'm not sure this idea of having the winner given by
> the longest streak is a very good one. It essentially leaves very much
> room for luck. So, if say Jesus has 10 correct guesses in a season but
> the longest streak happens to be 3 (very likely outcome), whilst
> milivella has 5, of which 4 are in a row (just a hypothetical
> example ;) ), who's the winner ? It may be somewhat more exciting this
> way but it's hardly fair.

You're completely right. It's absolutely not fair [and your nickname
is enough to know how much fairness is important to you ;) ]. But it's
more exciting, because anyone (even someone who haven't submitted a
single prediction) can hope to win until the last weekends of the
season. And - I've stated it explicitly - I proposed the rules having
fun in mind above any other factor, such as fairness.

But, again, with more people preferring your proposal (global count of
hits minus misses), the rule will be changed :)

Cheers

milivella

FairFootball

unread,
Feb 6, 2008, 7:52:24 PM2/6/08
to
milivella wrote:
> ... I wanted to say: assume that the

> best streak, at a given point of a season, is 5, and you've a streak
> of 1 hit before the three matches of the weekend start. If the first
> match is a miss, you already know that, even if the two next matches
> are hits, you can't go further than 3 hits in your current streak,

OK, so you could let the players choose if they want to add the hits
to the current streak or the next one. This way it might even get more
exciting. :)

> > BTW, personally I'm not sure this idea of having the winner given by

> > the longest streak is a very good one. ... It may be somewhat more


> > exciting this way but it's hardly fair.
>
> You're completely right. It's absolutely not fair [and your nickname
> is enough to know how much fairness is important to you ;) ].

Well, actually it depends on what kind of game you want. With streaks,
I'd say it's about 10% skill 90% luck. Especially when the games are
truly balanced. Which is fine, as long as the rules are clear in
advance.
Actually, I don't see why you couldn't have both. Award 2 titles at
the end, "best predictor" and "luckiest predictor".

milivella

unread,
Feb 6, 2008, 8:47:11 PM2/6/08
to
To FairFootball:

Thanks again for the thought reply :)

> you could let the players choose if they want to add the hits
> to the current streak or the next one. This way it might even get more
> exciting. :)

I don't know: isn't more exciting something that happens to you (once
you've chosen the team to support)? isn't it, in any case, something
that makes the rules slightly more complicated? But, really, I don't
know.

> Well, actually it depends on what kind of game you want. With streaks,
> I'd say it's about 10% skill 90% luck. Especially when the games are
> truly balanced. Which is fine, as long as the rules are clear in
> advance.

I agree.

> Actually, I don't see why you couldn't have both. Award 2 titles at
> the end, "best predictor" and "luckiest predictor".

Good idea! I go for it.

Regarding "best predictor": I see many ways to implement this. E.g.
1. Hits minus misses (FairFootball's proposal)
2. Hits squared divided by misses: this way, you could win with few
predictions submitted, if they are all hits; but if someone gains the
same ratio hits/misses as you (or even slightly worse) submitting more
predictions, he'll have more points than you.
Any other idea? As for now, I'd go for FairFootball's proposal (hits
minus misses) because it's the simplest.

Cheers

milivella

FairFootball

unread,
Feb 6, 2008, 9:31:21 PM2/6/08
to
milivella wrote:
> I don't know: isn't more exciting something that happens to you (once
> you've chosen the team to support)?

Very complicated question. I pass on this one. :)

> isn't it, in any case, something
> that makes the rules slightly more complicated?

That it certainly is. But only slightly.

> Regarding "best predictor": I see many ways to implement this. E.g.
> 1. Hits minus misses (FairFootball's proposal)

> 2. Hits squared divided by misses: ...

Beware, divide-by-zero looming. :)

milivella

unread,
Feb 6, 2008, 10:10:31 PM2/6/08
to
> > 2. Hits squared divided by misses: ...
>
> Beware, divide-by-zero looming. :)

Sorry: divided by hits+misses! Thanks for the warning :)

0 new messages