Can You Outpredict Paul the Octopus in the 2014 PF World Cup League?

[micromass]

26th June:
Portugal - Ghana: 0 - 1
USA - Germany: 1 - 3
Korea Republic - Belgium: 0 - 3
Algeria - Russia: 1 - 0

 

[D H]

June 26th:

Portugal v Ghana: 1 - 2
USA v Germany: 1 - 1
Korea Republic - Belgium: 0 - 2
Algeria - Russia: 0 - 1

 

[D H]

26th June:
Portugal - Ghana: 0 - 1
USA - Germany: 1 - 3

Booo!

 

[DennisN]

Some tough decisions for tomorrow, but I'll go for:

26th June:
Portugal - Ghana: 1 - 2
USA - Germany: 1 - 2 (sorry - please note I really hope for US )
Korea Republic - Belgium: 0 - 2
Algeria - Russia: 2 - 1

Matterwave gained +2 against us all for the Greece - Côte D'Ivoire match, that's fun .

 

[D H]

Portugal - Ghana: 1 - 2
USA - Germany: 1 - 2 (sorry - please note I really hope for US )

That's an acceptable outcome (just barely) for the US. That would put the US and Ghana tied in total points (4), tied in goal differential (0), tied in goals for (5). The next tie breaker is head-to-head, and since the US beat Ghana, the US advances to the round of 16.

My boo! to micromass a couple of posts back was his prediction would have Ghana advancing instead of the US.

Ghana advances if

• Germany defeats the USA, AND
• Ghana defeats Portugal, AND
  • The margin by which Germany defeats USA plus the margin by which Ghana defeats Portugal is three or more OR
  • The margin by which Germany defeats USA plus the margin by which Ghana defeats Portugal is two AND
    • Ghana scores at least two more goals against Portugal than the USA scores against Germany.

Portugal advances if

• Germany defeats the USA, AND
• Portugal defeats Ghana, AND
  • The margin by which Germany defeats USA plus the margin by which Portugal defeats Ghana is six or more OR
  • The margin by which Germany defeats USA plus the margin by which Portugal defeats Ghana is five AND
    • Portugal scores at least three more goals against Ghana than the USA scores against Germany OR
    • Portugal scores two more goals against Ghana than the USA scores against Germany AND
      • Portugal wins the coin toss against the USA.

The odds of the US advancing are pretty good. Nate Silverman (who normally predicts elections with amazing accuracy) currently puts the US as having a 76% chance of advancing.

There are some bizarre scenarios by which Germany goes home. For example, if the US defeats Germany 2-0 and Ghana defeats Portugal 4-0 then it's bye-bye, Germany. FiveThirtyEight puts the odds of this and similar outcomes as an extremely unlikely 0.3%.

 

[trollcast]

26th june
Portugal v Ghana: 1-2
USA v Germany: 1-3
Korea Republic v Belgium: 0-2
Algeria - Russia: 2-1

 

[Matterwave]

26th june
Portugal v Ghana: 3-2
USA v Germany: 3-1
Korea Republic v Belgium: 0-3
Algeria - Russia: 1-1

 

[Intrastellar]

26th june

Portugal v Ghana: 2-3
USA v Germany: 1-4
Korea Republic v Belgium: 1-2
Algeria - Russia: 2-1

I hope I made it on time, but I doubt that

 

[Logger]

26th june
Portugal v Ghana: 2-1
USA v Germany: 1-1
Korea Republic v Belgium: 0-3
Algeria - Russia: 1-1

 

[DennisN]

Algeria - Russia (26 June): 1 - 1
cristo: 1 - 0
micromass: 1 - 0
DH: 0 - 1
DennisN: 2 - 1
trollcast: 2 - 1
Matterwave: 1 - 1 +2
montadhar: 2 - 1

+2 against us all again, congratulations Matterwave!

 

[Matterwave]

I'm going to catch up! :D

 

[cristo]

Last 16 predictions:

Brazil v Chile: 4-1
Colombia v Uruguay: 1-0
Netherlands v Mexico: 2-0
Costa Rica v Greece: 1-0
France v Nigeria: 3-1
Germany v Algeria: 2-0
Argentina v Switzerland: 3-0
Belgium v USA: 2-1

 

[Borek]

1-0 in all remaining matches, assuming the team names are sorted alphabetically.

 

[D H]

June 28th:
Brazil v Chile: 2-1
Colombia v Uruguay: 1-0

June 29th:
Netherlands v Mexico: 1-0
Costa Rica v Greece: 2-0

 

[micromass]

Congratulations to DennisN for being the best one in the group stages!

Important!
In the following games, each of the following will get you 1 point:

• Identifying correctly who goes through to the next round
• Identifying the outcome correctly (win for Team A, win for Team B, draw), penalties do not count
• Identifying the score correctly (for example 2 - 1), penalties do not count

So it is possible to gain 3 points for a game this round!

Note that if you predict a draw, you have to predict a winner too.

 

[micromass]

June 28th:
Brazil v Chile: 3 - 1
Colombia v Uruguay: 1 - 1 (Uruguay wins)

June 29th:
Netherlands v Mexico: 2 - 1
Costa Rica v Greece: 2 - 0

 

[Intrastellar]

Congratulations to DennisN for being the best one in the group stages!

Important!
In the following games, each of the following will get you 1 point:

• Identifying correctly who goes through to the next round
• Identifying the outcome correctly (win for Team A, win for Team B, draw), penalties do not count
• Identifying the score correctly (for example 2 - 1), penalties do not count

So it is possible to gain 3 points for a game this round!

Note that if you predict a draw, you have to predict a winner too.

This is a better idea than simply multiplying by 2 :)

How about a point for identifying each team's goals ? This way the number of points will be doubled from the group stage, and predicting the goals will become more significant.

 

[Matterwave]

Round of 16 predictions:

Brazil v Chile: 4-3
Colombia v Uruguay: 2-0
Netherlands v Mexico: 1-1 Netherlands advances
Costa Rica v Greece: 1-1 Greece advances
France v Nigeria: 2-1
Germany v Algeria: 4-1
Argentina v Switzerland: 2-1
Belgium v USA: 2-2 USA advances

 

[micromass]

This is a better idea than simply multiplying by 2 :)

How about a point for identifying each team's goals ? This way the number of points will be doubled from the group stage, and predicting the goals will become more significant.

There might be too much emphasis on the goals then. It's not a bad idea, but let's keep it simple.

 

[D H]

Congratulations to DennisN for being the best one in the group stages!

Congrats, DennisN!

Important!
In the following games, each of the following will get you 1 point:

• Identifying correctly who goes through to the next round
• Identifying the outcome correctly (win for Team A, win for Team B, draw), penalties do not count
• Identifying the score correctly (for example 2 - 1), penalties do not count

So it is possible to gain 3 points for a game this round!

Note that if you predict a draw, you have to predict a winner too.

I like it!
This gives the matches in the knockout stage a bit more weight than those in the group stage, it accounts for kicks from the mark to resolve ties, and at the same time it keeps things simple.

 

[phion]

I like it!

Me too.

 

[trollcast]

June 28th:
Brazil v Chile: 3 - 2
Colombia v Uruguay: 1 - 1 Uruguay win on penalties
June 29th:
Netherlands v Mexico: 2 - 1
Costa Rica v Greece: 3 -1

 

[Intrastellar]

June 28th:
Brazil v Chile: 1-2
Colombia v Uruguay: 2-0

 

[DennisN]

Congratulations to DennisN for being the best one in the group stages!

Thanks, micromass and DH!

Important!
In the following games, each of the following will get you 1 point:

• Identifying correctly who goes through to the next round
• Identifying the outcome correctly (win for Team A, win for Team B, draw), penalties do not count
• Identifying the score correctly (for example 2 - 1), penalties do not count

So it is possible to gain 3 points for a game this round!

Note that if you predict a draw, you have to predict a winner too.

That's cool. Regretfully I missed to make predictions for today, and I will post predictions for tomorrow later.

 

[DennisN]

Sport commentators often seem to have a love of statistics, so I thought we could indulge ourselves with our own result table for the group stages:

I hope the table is understandable, and that neither I nor Excel did make any mistakes . The outcome accuracies and result accuracies are calculated with respect to the number of actual votes (that is, blank votes are excluded from the calculation). A fun thing is that PhysicsGente has the top accuracies for both outcomes and results. If anyone finds any error, please PM me. I also attach the Excel sheet and picture to this post.

Another big thanks to micromass for managing this thread, and a good luck to everybody for the rest of the cup!

 

[micromass]

Wow! Very nice and beautiful work. I always love those statistics

Too bad PhysicsGente made so few predictions, he would have been a worthy opponent for DennisN!

 

[DennisN]

And here are my predictions for tomorrow (I found it particularly hard to decide on the Netherlands - Mexico match):

June 29th:
Netherlands - Mexico: 2 - 2 (Netherlands wins)
Costa Rica - Greece: 2 - 1 (Costa Rica wins)

 

[Matterwave]

At least I'm over 50% so...slightly better than random guessing...

 

[DennisN]

At least I'm over 50% so...slightly better than random guessing...

I hope people don't take my table too seriously . There's a lot of luck involved too IMO (and I never expected to be so fortunate as I have been).

 

[Intrastellar]

June 29th:
Netherlands v Mexico: 3 - 1
Costa Rica v Greece: 2 - 0

 

[micromass]

I hope people don't take my table too seriously . There's a lot of luck involved too IMO (and I never expected to be so fortunate as I have been).

I am exactly 50%, which means that I am basically a perfect random number generator.

 

[Intrastellar]

Wow, our outcome accuracy is way too low! I thought it would be much higher than that.
If someone picked the underdogs of every single match, they would still have very comparable (or higher) number of points than ours

 

[micromass]

Wow, our outcome accuracy is way too low! I thought it would be much higher than that.
If someone picked the underdogs of every single match, they would still have very comparable (or higher) number of points than ours

But draws are also possible. Draws do not happen as much as a team winning, but it definitely does skew the results.

 

[Intrastellar]

But draws are also possible. Draws do not happen as much as a team winning, but it definitely does skew the results.

Ahh yea, that is right. If outcomes were picked at random, and I predicted no draws for any match, and for a big enough sample, is the maximum I can get about 66% or 50% of the predictions ?

My apologies for making this into a maths thread

 

[micromass]

My apologies for making this into a maths thread

One should never apologize for that!