Herd wisdom exists?

including formats, draws, seedings, etc.
User avatar
Dragonite
Senior Member
Posts: 12029
Joined: Fri Dec 14, 2007 19:42
Location: Lisboa, Portugal
Contact:

Post by Dragonite »

Do you know what subset means?
A is a subset of B if any member of A is also a member of B.

So if A is “the correct predictions of Adam” and it has 10 elements and B is “the correct predictions of Bethany” and it has 13 elements then A isn’t necessarily a subset of B.

A=(Krasnodar, Rennes, Slavia Prague, Zenit, Arsenal, Benfica, Sevilla, Dynamo Kiev, Inter, Valencia)
B=(Zenit, Arsenal, Benfica, Sevilla, Dynamo Kiev, Inter, Valencia, Salzburg, Chelsea, Frankfurt, Villarreal, Dinamo Zagreb, Napoli)

A isn’t a subset of B here, A has 3 elements that B hasn’t (Krasnodar, Rennes, Slavia Prague) and B has 6 elements that A hasn’t (Salzburg, Chelsea, Frankfurt, Villarreal, Dinamo Zagreb, Napoli)

From these 9 matches, Adam won 50€, Bethany lost 91€. In the remaining 7 they got the same score.
But sure, “10” < “13”… :roll:

Your “10” is a subset of Bethany’s “13”. Adam’s may not be. :nono:

og2002gr wrote:To me Bethany is an idiot.
Why bet on odds around 1.00? There is no value there!
Bethany isn’t an idiot; she’s a devoted follower of her #Godds. :mrgreen: (And this time her #Godds haven’t disappointed her.)
Records and Statistics:
:arrow: Champions League (all 141 participants - 1992/1993 to 2019/2020)
:arrow: Europa League (all 215 participants - 2009/2010 to 2019/2020)
:arrow: UEFA Youth League (all 162 participants - 2013/2014 to 2019/2020)
User avatar
Overgame
Senior Member
Posts: 4545
Joined: Wed May 02, 2007 23:58

Post by Overgame »

Dragonite, this topic had a nice turn. Go troll elsewhere.
User avatar
og2002gr
Senior Member
Posts: 8255
Joined: Fri Feb 19, 2010 13:51
Location: Greece

Post by og2002gr »

War of #GODDS
CL last 16

correct predictions (out of 8 )

Code: Select all

6   - manager CL pts
5   - Euro Club Index
4,5 - bet companies
4   - market value (transfermarkt)
4   - squad CL experience
3   - points CL+EL (5 years)
3   - UEFA Ranking
3   - times going further (10 years)

6 - og predictions
percentages

Code: Select all

 299%   - manager CL pts
 109%   - Euro Club Index
 102%   - bet companies
  32%   - market value (transfermarkt)
 14%    - squad CL experience
-113%   - points CL+EL (5 years)
-123%   - UEFA Ranking
-126%   - times going further (10 years)

 266%   - og predictions
CL round was full of surprises.
Still "manager CL points" topped this ranking, followed by "Euro Club Index' and "bet companies".
Looking at the 2nd table, the other 5 rankings hardly passed or didn't pass the 50-50 limit at all.
Past results (UEFA Ranking, 10y record, 5 years results) not only failed to predict the most winners, but on the contrary they did worse than having a blind "coin toss" prediction!

My instinct was very good this time topping the first table, and finishing 2nd just behind "manager CL points". This time I beat #Godds 8)
fewer spots :arrow: weaker leagues :arrow: fewer spots :arrow: weaker leagues :arrow: fewer spots :arrow: weaker leagues

http://kassiesa.net/uefa/forum2/viewtop ... 52#p359852
User avatar
og2002gr
Senior Member
Posts: 8255
Joined: Fri Feb 19, 2010 13:51
Location: Greece

Post by og2002gr »

War of #GODDS
EL last 32

correct predicitons (out of 8 )

Code: Select all

7   - Club Elo
6   - market value (trasnfermarkt)		
5,5 - bet company
5,5 - going further (10y)
5   - UEFA ranking	
5   - squad EL experience	
5   - manager EL pts
3,5 - points (5y)	

5,5 - og predictions
percentages

Code: Select all

395% - Club Elo
310% - market value (trasnfermarkt)		
238% - bet company
238% - going further (10y)
140% - UEFA ranking	
138% - squad EL experience	
128% - manager EL pts
-56% - points (5y)	

219% - og predictions
Things were more expected tonight for EL last 32 round.
I used ELO for the first time instead of EuroCLubIndex and they didn't disappoint, topping the rankings.
Market value and bet companies followed just like previous EL round.
This time manager and squad experience didnt help.
Once again previous performance of the clubs was a bad predictive value, even a negative one!
My performance was average - sub-top.
fewer spots :arrow: weaker leagues :arrow: fewer spots :arrow: weaker leagues :arrow: fewer spots :arrow: weaker leagues

http://kassiesa.net/uefa/forum2/viewtop ... 52#p359852
User avatar
og2002gr
Senior Member
Posts: 8255
Joined: Fri Feb 19, 2010 13:51
Location: Greece

Post by og2002gr »

These are the tables with all results so far:
Image

As (I) expected "bookmakers" are the best so far, predicting correctly around 2/3 winners, while leading "percentages" table too.
ECI/ELO and "market value" are doing good too.

It is ..impressive how weak UEFA ranking (similar to last 5 seasons points) are at predicting the current strength of teams.

If you have any better ideas of how to use data above please make your suggestions!
Last edited by og2002gr on Fri Mar 15, 2019 00:20, edited 1 time in total.
fewer spots :arrow: weaker leagues :arrow: fewer spots :arrow: weaker leagues :arrow: fewer spots :arrow: weaker leagues

http://kassiesa.net/uefa/forum2/viewtop ... 52#p359852
User avatar
Dragonite
Senior Member
Posts: 12029
Joined: Fri Dec 14, 2007 19:42
Location: Lisboa, Portugal
Contact:

Post by Dragonite »

Dragonite wrote:Image
Interesting thing about “following the herd” strategy in the EL last 16:
With the same amount played in each of the 8 clashes, the profit from the 6 hits (1.28+1.17+1.57+1.25+1.40+1.33) is equal to the loss from the 2 misses.

Congratulations to the people that predicted Eintracht Frankfurt and mostly Slavia Prague to advance!! :clapping:
(I wasn’t one of them… and I also missed Napoli advancing).
Dragonite wrote:Image
In the CL, “following the herd” was disastrous, only 4 hits, not enough to compensate the 4 misses, and a loss of 31.375% (-251€ if the bet was 100€ per clash). :fingal:
Records and Statistics:
:arrow: Champions League (all 141 participants - 1992/1993 to 2019/2020)
:arrow: Europa League (all 215 participants - 2009/2010 to 2019/2020)
:arrow: UEFA Youth League (all 162 participants - 2013/2014 to 2019/2020)
User avatar
Dragonite
Senior Member
Posts: 12029
Joined: Fri Dec 14, 2007 19:42
Location: Lisboa, Portugal
Contact:

Post by Dragonite »

og2002gr,

Can I ask you a question?

Why have you predicted Frankfurt to advance in Lyonnais’ prediction game (March 5th 11:50) and then in your table posted a few hours later (March 5th 19:49) predicted Inter to advance? :confused:
og2002gr wrote:Champions League

Dortmund - Tottenham 2-1
Real Madrid - Ajax 2-1
Paris - Man. United 2-1
Porto - Roma 2-1


Europa League

Zenit - Villarreal 2-1
Rennes - Arsenal 1-2
Sevilla - Slavia Praha 2-1
Dinamo Zagreb - Benfica 1-2
Frankfurt - Internazionale 2-1
Valencia - Krasnodar 1-0
Napoli - Salzburg 2-1
Chelsea - Dynamo Kyiv 2-1
og2002gr wrote:Image

Rennes - Arsenal is the only game where everything goes one way.
Benfica is a strong favorite in all rankings except manager EL pts.
Inter has the advantage in all dimensions, but Club Elo considers Eintracht as the favorite!
Zenit-Villareal looks like the most balanced pair.
It’s impossible to determine how good your predictive skills are when you’re predicting contradictory things even during the same day!! :oops:
Records and Statistics:
:arrow: Champions League (all 141 participants - 1992/1993 to 2019/2020)
:arrow: Europa League (all 215 participants - 2009/2010 to 2019/2020)
:arrow: UEFA Youth League (all 162 participants - 2013/2014 to 2019/2020)
User avatar
max tre
Posts: 483
Joined: Wed Dec 11, 2013 20:22
Location: Veneto, Italy

Post by max tre »

og2002gr wrote:Another ranking type I used for more detailed analysis was adding units according to the percentage given when the prediction was correct minus the units according to the percentage when the prediciton was wrong.
ex. if you predict 80-20 you win 80 units when you are right and you lose 80 when you are wrong. Maximum would be 100*16=1600.
Wait. I know it's old, but just caught up with this.
Did you mean 80-20 = 60 when "I'm right" and 20-80 = -60 when "I'm wrong"? That's what I would do at first glance, at least.
User avatar
og2002gr
Senior Member
Posts: 8255
Joined: Fri Feb 19, 2010 13:51
Location: Greece

Post by og2002gr »

@max tre
When the prediction is correct you win the value of the favorite.
When the prediction is wrong you lose the value of the favorite.
So that when you feel confident and choose a high percentage you gain the maximum profit if you were right, or you lose the maximum if you were that wrong.
I guess your method would give the same results.

@Dragonite
Different "questions", different "answers".

Any comment about the very poor predictability value of past result (so far)?
I knew it was weak but I was surprised to see that they are close to "coin toss" method (50-50).
fewer spots :arrow: weaker leagues :arrow: fewer spots :arrow: weaker leagues :arrow: fewer spots :arrow: weaker leagues

http://kassiesa.net/uefa/forum2/viewtop ... 52#p359852
User avatar
Dragonite
Senior Member
Posts: 12029
Joined: Fri Dec 14, 2007 19:42
Location: Lisboa, Portugal
Contact:

Post by Dragonite »

Answer the question, please.
It’s a simple question, why are you predicting Frankfurt to advance in March 5th 11:50 and then Inter to advance in March 5th 19:49, less than 8 hours later?

You say that this time you defeated the #Godds, it’s impossible to check that when your predictions are contradictory (not to mention the 50%-50% in Zenit vs. Villarreal that’s no prediction at all, pick a side!!)
Records and Statistics:
:arrow: Champions League (all 141 participants - 1992/1993 to 2019/2020)
:arrow: Europa League (all 215 participants - 2009/2010 to 2019/2020)
:arrow: UEFA Youth League (all 162 participants - 2013/2014 to 2019/2020)
User avatar
Overgame
Senior Member
Posts: 4545
Joined: Wed May 02, 2007 23:58

Post by Overgame »

Dragonite, could you go troll elsewhere ?
User avatar
og2002gr
Senior Member
Posts: 8255
Joined: Fri Feb 19, 2010 13:51
Location: Greece

Post by og2002gr »

Dragonite wrote:Answer the question, please.
It’s a simple question, why are you predicting Frankfurt to advance in March 5th 11:50 and then Inter to advance in March 5th 19:49, less than 8 hours later?

You say that this time you defeated the #Godds, it’s impossible to check that when your predictions are contradictory (not to mention the 50%-50% in Zenit vs. Villarreal that’s no prediction at all, pick a side!!)
Already answered. Different competitions, different rules, different tactics.

Concerning my predictions, each way I beat the GODDS for once (so far).
And no. I don't have to pick a side in this game. When I have to choose (Lyonnais game) I do.
fewer spots :arrow: weaker leagues :arrow: fewer spots :arrow: weaker leagues :arrow: fewer spots :arrow: weaker leagues

http://kassiesa.net/uefa/forum2/viewtop ... 52#p359852
User avatar
Dragonite
Senior Member
Posts: 12029
Joined: Fri Dec 14, 2007 19:42
Location: Lisboa, Portugal
Contact:

Post by Dragonite »

“Competition”?!

Lyonnais’ game is a competition, you’re trying to defeat an opponent (that’s why you did it, you weren’t sure so you copied your opponent’s choice?) :oops:

This here is no “competition”
You choose 16 teams to advance (8 CL quarterfinalists+ 8 EL quarterfinalists), then you’ll hit some and you’ll miss some. And then we see if your hits compensate your misses using your beloved #Godds.
When you say 50%-50% that’s nothing, you have to pick a side. :evil:
And the side that you pick should be the same for the Lyonnais’ game at 11:50 and for these boards that you’re having fun with at 19:49, there’s no reason for it to be different. :sigh:
Records and Statistics:
:arrow: Champions League (all 141 participants - 1992/1993 to 2019/2020)
:arrow: Europa League (all 215 participants - 2009/2010 to 2019/2020)
:arrow: UEFA Youth League (all 162 participants - 2013/2014 to 2019/2020)
User avatar
og2002gr
Senior Member
Posts: 8255
Joined: Fri Feb 19, 2010 13:51
Location: Greece

Post by og2002gr »

So the one is competition, the 2nd is not.
You already had your answer. Why asking thrice?

Since the value is positive then there is compensation. This should be enough.

As long as some teams can be 50-50 in every other ranking (look at YOUR tables), I can do the same too since I expect it to be so.
fewer spots :arrow: weaker leagues :arrow: fewer spots :arrow: weaker leagues :arrow: fewer spots :arrow: weaker leagues

http://kassiesa.net/uefa/forum2/viewtop ... 52#p359852
User avatar
Dragonite
Senior Member
Posts: 12029
Joined: Fri Dec 14, 2007 19:42
Location: Lisboa, Portugal
Contact:

Post by Dragonite »

So, just to be clear, the answer is that in Lyonnais’ prediction game you also consider your opponent’s prediction when you’re posting yours, and if you disagree but you’re not “sure”, you just copy your opponent’s choice? :oops:

The “value is positive”… What are you talking about? :?
In case you haven’t noticed, “Following the herd” in the CL+EL last 16 with 100€ per clash would get you a LOSS of 251€. 10/16 may beat your simple “better than coin flip” test, but it’s not enough to have a profit here. :nono:

I have an advice for you since you’re such a fan of fence-sitting. From now onwards, why don’t you ALWAYS “predict” 50%-50% in every clash?
Then you’re always right, you can always say that the winner was “your 50.1%” and the loser was “your 49.9%” and you were just rounding it to the nearest integer, you can even blame the Excel sheet for that. :roll:

And then if this is still contradictory with your Lyonnais prediction game choices, you can always say that there you play by different rules and instead of predicting what you think that will happen, you just copy what others do sometimes.

This is hopeless… :upset:
I created this with the goal to discredit your #Godds #FollowTheHerd theories, hopefully by now you’d have abandoned them already and start focusing your attention on “everything else”.
Instead the opposite happened, you abandoned “everything else” and are getting more and more obcessed with finding #Godds or some other “method” that can tell you in just a few seconds what will happen in the future, as if such thing existed!! :sigh:
Records and Statistics:
:arrow: Champions League (all 141 participants - 1992/1993 to 2019/2020)
:arrow: Europa League (all 215 participants - 2009/2010 to 2019/2020)
:arrow: UEFA Youth League (all 162 participants - 2013/2014 to 2019/2020)
Post Reply