Concacaf Qualifying For World Cup 2022 Prediction Contest: Round 3 (Octagonal) [R]

Here are olephill2's standings predictions with records as Wins-Draws-Losses:

1. Mexico: 36 (12-0-2), 23 goals scored, 5 goals allowed, +18
2. USA: 29 (9-2-3), 24 goals scored, 9 goals allowed, +15
3. Honduras: 26 (8-2-4), 21 goals scored, 17 goals allowed, +4
4. Costa Rica: 21 (6-3-5), 18 goals scored, 16 goals allowed, +2
5. Canada: 19 (5-4-5), 19 goals scored, 18 goals allowed, +1
6. Jamaica: 11 (2-5-7), 15 goals scored, 23 goals allowed, -8
7. El Salvador: 7 (1-4-9), 12 goals scored, 25 goals allowed, -13, above Panama on goal differential
8. Panama: 7 (1-4-9), 5 goals scored, 24 goals allowed, -19

Draw percentage: 12/56 = 21.4 percent
Goals per team per game = 137/112 = 1.22

Here are the mean of the predictions for each team:

1. Mexico: Rank of 1.125, 32.625 points (10.000-2.625-1.375), 28.875 goals scored, 9.625 goals allowed, +19.250
2. USA: Rank of 1.875, 30.125 points (8.875-3.500-1.625), 27.375 goals scored, 12.250 goals allowed, +15.125
3. Canada: Rank of 3.875, 20.000 points (5.125-4.625-4.250), 20.125 goals scored, 17.250 goals allowed, +2.875
4. Costa Rica: Rank of 4.500, 18.500 points (5.000-3.500-5.500), 16.500 goals scored, 18.875 goals allowed, -2.375
5. Honduras: Rank of 1.875, 17.750 points (4.625-3.875-5.500), 16.375 goals scored, 18.750 goals allowed, -2.375
6. Jamaica: Rank of 5.500, 15.500 points (4.000-3.500-6.500), 15.625 goals scored, 20.875 goals allowed, -5.250
7. Panama: Rank of 7.000, 10.500 points (2.375-3.375-8.250), 11.125 goals scored, 23.625 goals allowed, -12.500
8. El Salvador: Rank of 7.250, 8.750 points (1.750-3.500-8.750), 10.625 goals scored, 25.375 goals allowed, -14.750

 

Your scores added up to 19 scored and 18 allowed.

Because USA and Canada didn't win, the score predictions only gave 35 points, for an average of 4.375. An exact score is worth 4, so an exact matchday would be 16. The score predictions get converted into points out of 27 after the Octagonal, so I will post cumulative score predictions, but not anything cumulative for all rounds combined. I lead after Round 2. If everybody's score predictions for every matchday are the same as they were for Matchday 1 in terms of how many points they are worth, and I add Octagonal score predictions to Rounds 1 and 2 but not Octagonal standings predictions, Dr. Gamera would win. After Matchday 3 I will post score prediction results. devioustrevor's prediction of Panama 0:0 Costa Rica was the only exact score.

 

Seven of twelve games had two or fewer correct result predictions out of eight people. devioustrevor was the only person with three correct results today, and he leads the score predictions with 18. I'm in last with 9. I'll post the standings later. Canada 3-0 El Salvador was the only game where the most common prediction was the exact score. The average person has 13.625 points. A person who predicted the most common prediction in every game would have 15 points. 16 games had a tie for the most common prediction. The only one of those games that happened was Costa Rica vs. Jamaica, where Costa Rica winning 1-0, 2-0, and 2-1 were predicted by two people each. Since it was 1-1, none of the most common predictions got any points. There are four games with a tie for the most common prediction with different outcomes.

 

Here are the score predictions out of 48 points with everybody's points, correct results, amount exact scores, and exact scores named if you had any:

1. devioustrevor: 18, 6, 2 (exact scores for PAN 0-0 CR and CAN 3-0 ES)
2. pipilesunidos: 17, 6, 0
Tied 3. Dr. Gamera: 15, 6, 0
Tied 3. olephill2: 15, 5, 1 (exact score for CR 0-1 MEX)
5. barnabyjonez: 13, 4, 2 (exact scores for USA 1-1 CAN and CAN 3-0 ES)
6. IASocFan: 12, 4, 1 (exact score for PAN 1-1 MEX)
7. jagum: 10, 4, 0
8. EvanJ: 9, 3, 1 (exact score for CAN 3-0 ES)

The means were 13.625 points, 4.75 correct results, and 0.875 exact scores. The games with the best predictions were CAN 3-0 ES with 25, CR 0-1 MEX with 18, PAN 0-0 CR with 16, and MEX 2-1 JAM with 16. Half of the games gave 4 points or fewer, four games and one correct result prediction, and everybody predicted Canada to win hosting Honduras and got no points from that 1-1 draw. Of Canada's first four games including their upcoming game in Mexico, everybody predicted the same outcome in each game (win hosting HON, loss at USA, win hosting ES, and loss at MEX) except for barnabyjonez who predicted the exact score of their 1-1 in USA. CAN 3-0 ES was the only game where the most common prediction was the exact score.

 

It's strange watching the end of a game rooting for no goals, but that's what I did last night as I was the only one to predict El Salvador 1-0 Panama. Nobody predicted Mexico to not beat Canada. I'll post scores after Matchday 6. devioustrevor has the best score predictions through Matchday 4.

 

Nobody predicted Panama to beat USA. Dr. Gamera's predictions of MEX 3-0 and CR 2-1 ES were the only exact scores today, and he won the matchday with 11 points. With 35, 35, 39, 53, and 53 points, every matchday has had at least as many points as the previous matchday.

 

Here are the score predictions for Matchdays 4 through 6 out of 48 points with everybody's points, correct results, exact scores, and exact scores named if you had any:

Tied 1. Dr. Gamera: 25, 8, 2 (exact scores for MEX 3-0 HON and CR 2-1 ES)
Tied 1. olephill2: 25, 8, 1 (exact score for USA 2-0 JAM)
3. barnabyjonez: 24, 8, 2 (exact scores for USA 2-0 JAM and ES 0-2 MEX)
4. IASocFan: 23, 8, 0
5. EvanJ: 22, 8, 2 (exact scores for ES 1-0 PAN and USA 2-1 CR)
6. devioustrevor: 20, 7, 0
7. pipilesunidos: 16, 6, 1 (exact score for USA 2-0 JAM)
8. jagum: 10, 4, 0

The means were 20.625 points, 7.125 correct results, and 1 exact score.

Here are the cumulative score predictions through Matchday 6 out of 96 points with everybody's points, correct results, and exact scores:

Dr. Gamera: 40, 14, 2
olephill2: 40, 13, 2
devioustrevor: 38, 13, 2
barnabyjonez: 37, 12, 4
IASocFan: 35, 12, 1
pipilesunidos: 33, 12, 1
EvanJ: 31, 11, 3
jagum: 20, 8, 0

The means were 34.25 points, 11.825 correct results, and 1.875 exact scores. The second set of games has 60.2 percent of the points, meaning that the mean prediction for the second set of games was over 50 percent more points than the mean prediction for the first set of games. By "points" I mean the way I evaluate the predictions, not the team standings.

 

Here are the Matchday 7 score prediction results with everybody's points out of 16, correct results, and exact scores:

Tied 1. barnabyjonez: 10, 3, 1 (exact score for ES 1-1 JAM)
Tied 1. olephill2: (exact score for CAN 1-0 CR)
3. IASocFan: 8, 2, 2 (exact scores for ES 1-1 JAM and USA 2-0 MEX)
4. Dr. Gamera: 7, 2, 1 (exact score for ES 1-1 JAM)
5. devioustrevor: 6, 2, 0
6. pipilesunidos: 5, 2, 0
7. jagum: 3, 1, 0
8. EvanJ: 2, 0, 0

The means were 6.375 points, 2 correct results, and 0.625 exact scores. Nobody predicted Panama to beat Honduras, and in their last three games Panama has two wins that nobody predicted them to get, with the prior one hosting USA. I'll most the cumulative score predictions after Matchday 8. My score predictions continue to be bad, but I calculated the points off in the standings if every team finishes with double their current points, and I would do well if that's what happens. Here are how many points off in the standings everybody would be if every team finishes with double their current points:

1. barnabyjonez: 42
2. EvanJ: 44
3. jagum: 48
4. devioustrevor: 52
Mean: 53.75
5. Dr. Gamera: 54
6. pipilesunidos: 56
7. olephill2: 66
8. IASocFan: 68

 

The whole is greater than the sum of the parts. The plurality (most common) score prediction for each game has 61 points, and olephill2 leads with 60. When there was a tie for the plurality, I scored both of them and took the mean, which is different from taking the mean of the predictions before scoring them. Taking the mean of the predictions first doesn't make sense because, for example, if the pluralities were 1-1 and 2-1, and I evaluated 1.5-1, it would be at a disadvantage because it couldn't be the exact score. The pluralities have something-point-5 in two places, so it adds up to a whole number.

Nobody predicted Canada to beat Mexico, which is the fifth time in 32 games that everybody predicted the wrong outcome. The other three games gave about the same amount of points. CR 2-1 HON and JAM 1-1 USA gave 19 points, and PAN 2-1 ES gave 17 points.

I'll post score prediction results later.

 

Here are the Matchday 8 score prediction results with everybody's points out of 16, correct results, and exact scores:

1. barnabyjonez: 11, 3, 2 (exact scores for PAN 2-1 ES and JAM 1-1 USA)
Tied 2. EvanJ: 10, 3, 1 (exact score for PAN 2-1 ES)
Tied 2. IASocFan: 10, 3, 1 (exact score for JAM 1-1 USA)
Tied 2. olephill2: 10, 3, 1 (exact score for JAM 1-1 USA)
Tied 5. Dr. Gamera: 4, 1, 1 (exact score for CR 2-1 HON)
Tied 5. pipilesunidos: 4, 1, 1 (exact score for JAM 1-1 USA)
Tied 7. jagum: 3, 1, 0
Tied 7. devioustrevor: 3, 1, 0

JAM 1-1 USA was the first game to have four people predict the score.

Here are score predictions through Matchday 8 with everybody's points out of 128, correct results, and exact score:

1. olephill2: 60, 19, 4
2. barnabyjonez: 58, 18, 7
3. IASocFan: 53, 17, 4
4. Dr. Gamera: 51, 17, 4
5. devioustrevor: 47, 16, 2
6. EvanJ: 43, 15, 4
7. pipilesunidos: 42, 15, 2
8. jagum: 26, 10, 0

The means were 47.5 points, 15.875 correct results, and 3.375 exact scores. 49.6 percent of predictions had the correct result, and 10.5 percent of predictions had the exact score.

On Matchday 7, I was happy that USA beat Mexico, but my score predictions were bad. Yesterday I was disappointed that USA didn't beat Jamaica, but my score predictions were good. I'm happy with my standings predictions. Panama has exceeded six people's predictions, and Mexico can't get up to four people's predictions. jagum's score predictions are the worst, but if every team finishes with their current points per game, his points off in the standings would be good. Since score predictions scores want to be high and points off in the standings want to be low, if the same people were doing better at both or worse at both, there would be a negative correlation. The correlation is 0.271. It's impossible for me to say how people are doing for the components combined and what outcomes each person wants going into the last matchday without doing way more calculations than I will do. In the Hexgonal, each matchday had 3*3*3 = 81 possible outcomes in terms of points. Now it is 3*3*3*3 = 243 outcomes. What I might do before the last matchday is take each person's score predictions for that day and calculate who will win if all of those scores happened.

 

Here are the Matchday 9 score prediction results with everybody's points out of 16, correct results, and exact scores:

Tied 1. IASocFan: 9, 3, 1 (exact score for CR 1-0 PAN)
Tied 1. Dr. Gamera: 9, 3, 1 (exact score for JAM 1-2 MEX)
Tied 1. jagum: 9, 3, 1 (exact score for JAM 1-2 MEX)
4. EvanJ: 8, 3, 1 (exact score for JAM 1-2 MEX)
5. olephill2: 7, 3, 0
6 barnabyjonez: 6, 2, 1 (exact score for CR 1-0 PAN)
Tied 7. devioustrevor: 5, 2, 0
Tied 7. pipilesunidos: 5, 2, 0

The averages were 7.25 points, 2.625 correct results, and 0.625 exact scores. Nobody predicted Canada to beat Honduras. It was the sixth game that nobody got right, which is one-sixth of the 36 games. Those six games include four of Canada's games including both of Canada's games against Honduras, the previous one being CAN 1-1 HON on Matchday 1. 23 of the 58 points came from CR 1-0 PAN, which was tied for the second most points from the 36 games. CAN 3-0 ES on Matchday 3 gave people 25 points.

 

Three of today's four games had everybody be wrong, and PAN 3-2 JAM had three correct results and no exact scores. I got 3 points along with barnabyjonez, and IASocFan got 2. People's points by matchday are 35, 35, 39, 53, 53, 59, 51, 55, 58, and 8. Today ended a streak of six matchdays where people earned 51 to 59 points. barnabyjonez and olephill2 are tied for the lead with 67 points, and I'l post the cumulative standings after Wednesday. Eight people predicting eight teams is 64, and 34 point totals are impossible. Panama exceeded everybody's predictions. El Salvador tied my prediction of 9 points and exceeded everybody else's. Canada will exceed everybody's predictions if they win at El Salvador, so exactly one of them will exceed everybody's predictions after Wednesday. Six of eight people predicted Mexico to get more points than they can. Nobody predicted Canada, USA, Mexico, and Panama to be the top four in any order.

 

Here are the Matchday 11 score prediction results (see my previous post for the only people who got points from Matchday 11) with everybody's points out of 16, correct results, and exact scores:

1. barnabyjonez: 10, 3, 1 (exact score for USA 3-0 HON)
2. jagum: 8, 3, 0
Tied 3. EvanJ: 6, 2, 1 (exact score for USA 3-0 HON)
Tied 3. IASocFan: 6, 2, 0
Tied 3. devioustrevor: 6, 2, 0
Tied 3. olephill2: 6, 2, 1
7. Dr. Gamera: 5, 2, 0
8. pipilesunidos: 4, 2, 0

The averages were 6.375 points, 2.25 correct results, and 0.375 exact scores.

Here are the cumulative score prediction results with everybody's points out of 176, correct results, and exact scores:

1. barnabyjonez: 77, 23, 9
2. olephill2: 73, 24, 5
3. IASocFan: 70, 22, 5
4. Dr. Gamera: 65, 22, 5
5. EvanJ: 60, 20, 6
6. devioustrevor: 58, 20, 3
7. pipilesunidos: 51, 19, 2
8. jagum: 43, 16, 1

The averages were 62.125 points, 21.125 correct results, and 4.375 exact scores.

Eight people multiplied by eight teams is 64. 18 of the standings predictions can be exactly right about the amount of points a team gets. Nobody will be right about Canada. Seven people could be right about Costa Rica, but not all of them because they didn't predict the same amount of points.

 

Keep in mind that the Octagonal score predictions are only 27 percent of the total, and I can't evaluate standings predictions until the round is over. Since you only predicted the last matchday of Round 1, you didn't have a chance to win the total. When the round ends, I will post all the components of the Octagonal, the Octagonal, and the total, so it is possible to do below average but still win a component. It's possible for a game to be good for your score predictions and bad for your standings predictions, which would happen if the standings are different from what you predicted at that point so being wrong on future games would make your standings predictions better, or vice-versa, and I have no way of saying if you should ever root against your score predictions.

Each matchday has 3*3*3*3 = 81 outcomes. Before Matchday 14, I will try to post standings predictions scenarios for the last matchday. I won't do 81 scenarios. If I assume that Mexico will win hosting El Salvador, there are 27 scenarios for the other games, and with only eight people having predicted, I should be able to make time for that. What would make it take longer is if standings positions could be decided by goal differential that is close.

 

Here's a summary of predictions:

Costa Rica vs. Canada:
6 predicted Costa Rica to win
1 predicted a draw
1 predicted Canada to win
The mean was 1.875-1.
The most common prediction was 2-1
I predicted 3-2

Panama vs. Honduras:
5 predicted Panama to win
3 predicted a draw
The mean was 1.25-0.5.
The most common predictions were a tie between 2-1, 1-0, and 1-1.
I predicted 2-1.

Jamaica vs. El Salvador:
All 8 predicted Jamaica to win.
The mean was 2.375-0.5.
The most common prediction was 1-0.
I predicted 3-1.

Mexico vs. USA:
6 predicted Mexico to win
2 predicted a draw
The mean was 1.875-0.625.
The most common prediction was 2-0.
I predicted 2-0.

USA qualifying is more important than my predictions being good, so I'm rooting against Costa Rica and Panama.

After COVID after other confederations started qualifying, Concacaf's format change, how many games were packed together in June 2021, and three qualifiers in a week four times, it is special to realize that a week from now qualifying will be over except for playoffs, and I'll be scoring predictions. FIFA is acting quickly. Normally rankings come out on Thursday eight days after qualifiers, but this time they have qualifying (except for playoffs) end on Wednesday March 30, FIFA Rankings on March 31, and the World Cup draw on April 1.

 

Here are the Matchday 12 score prediction results with everybody's points out of 16, correct results, and exact scores:

1. IASocFan: 7, 2, 1 (exact score for PAN 1-1 HON)
2. olephill2: 6, 2, 0
3. devioustrevor: 5, 2, 0
Tied 4. Dr. Gamera: 4, 1, 1 (exact score for PAN 1-1 HON)
Tied 4. pipilesunidos: 4, 1, 1 (exact score for MEX 0-0 USA)
Tied 6. barnabyjonez: 3, 1, 0
Tied 6. jagum: 3, 1, 0
8. EvanJ: 2, 1, 0

The means were 4.25 points, 1.375 correct results, and 0.375 exact scores.

Here are the Matchday 13 score prediction results with everybody's points out of 16, correct results, and exact scores:

1. IASocFan: 8, 3, 1 (exact score for ES 1-2 CR)
2. Dr. Gamera: 7, 3, 0
Tied 3. EvanJ: 5, 2, 0
Tied 3. barnabyjonez: 5, 2, 0
Tied 3. devioustrevor: 5, 2, 0
Tied 3. olephill2: 5, 2, 0
Tied 7. jagum: 4, 2, 0
Tied 7. pipilesunidoes: 4, 2, 0

The means were 5.375 points, 2.25 correct results, and 0.125 exact scores.

Here are the cumulative score preidctions through Matchday 13 with everybody's points out of 208, correct results out of 52, and exact scores out of 52:

Tied 1. IASocFan: 85, 28, 7
Tied 1. barnabyjonez: 85, 27, 9
3. olephill2: 84, 28, 5
4. Dr. Gamera: 76, 26, 6
5. devioustrevor: 68, 24, 2
6. EvanJ: 67, 24, 6
7. pipilesunidos: 59, 22, 3
8. jagum: 50, 19, 1

The means were 71.75 points, 24.75 correct results, and 4.875 exact scores.

My standings predictions look good. The question is if my standings predictions will be good enough to make up for my below average score predictions.

 

I just finished scoring. There are a lot of components, and I'm not posting final scores yet. My score predictions were not good enough for me to win after I led after Round 2. Here are the Matchday 14 score prediction results with everybody's points out of 16, correct results, and exact scores:

1. jagum: 9, 3, 0
2. EvanJ: 8, 3, 0
Tied 3. pipilesunidos: 7, 2, 1 (exact score for JAM 2-1 HON)
Tied 3. olephill2: 7, 2, 1 (exact score for MEX 2-0 ES)
Tied 5. IASocFan: 6, 2, 0
Tied 5. barnabyjonez: 6, 2, 0
7. Dr. Gamera: 3, 1, 0
8. devioustrevor: 2, 1, 0

The means were 6 points, 2 correct results, and 0.25 exact scores.

Here are the final score predictions with everybody's points out of 224, correct results out of 56, and exact scores out of 56:

Tied 1. IASocFan: 91, 30, 7
Tied 1. barnabyjonez: 91, 29, 9
Tied 1. olephill2: 91, 30, 6
4. Dr. Gamera: 79, 27, 6
5. EvanJ: 75, 27, 6
6. devioustrevor: 70, 25, 2
7. pipilesunidos: 66, 24, 4
8. jagum: 59, 22, 1

The means were 77.75 points, 26.75 correct results, and 5.125 exact scores. Those scores out of 224 are converted to a score out of 27 that I am not posting yet.

 

I converted the score predictions to a score out of 27 with the best people getting 27, the worst person getting 0, and everybody else distributed proportionately in between:

Tied 1. IASocFan: 27.000
Tied 1. barnabyjonez: 27.000
Tied 1. olephill2: 27.000
4. Dr. Gamera: 16.875
5. EvanJ: 13.500
6. devioustrevor: 9.281
7. pipilesunidos: 5.906
8. jagum: 0.000

The mean was 15.820.

Here is how many points each person was off in the standings for the teams combined:

1. jagum: 40
2. EvanJ: 42
3. barnabyjonez: 50
4. devioustrevor: 60
5. Dr. Gamera: 62
Tied 6. IASocFan: 64
Tied 6. olephill2: 64
8. pipilesunidos: 78

The mean was 57.5.

I converted the standings predictions to a score out of 14 with the best people getting 14, the worst person getting 0, and everybody else distributed proportionately in between:

1. jagum: 14.000
2. EvanJ: 13.263
3. barnabyjonez: 10.316
4. devioustrevor: 6.632
5. Dr. Gamera: 5.895
Tied 6. IASocFan: 5.158
Tied 6. olephill2: 5.158
8. pipilesunidos: 0.000

The mean was 7.553.

I gave 1 point per team for being correct about whether that team was in the top three, fourth, or bottom four:

Tied 1. EvanJ: 8
Tied 1. Dr. Gamera: 8
Tied 3. barnabyjonez: 6
Tied 3. jagum: 6
Tied 3. olephill2: 6
Tied 3. devioustrevor: 6
Tied 7. IASocFan: 5
Tied 7. pipilesunidos: 5

The mean was 6.25. It was impossible to get 7 because if you were wrong about a team, it had to be switched with another team or had more than one other team wrong.

For positions within the top three, I gave 1.5 points for having a team in the correct position, and 0.75 for having a team one position off. Everybody got 1.5. Seven people had Mexico first and USA second for 0.75 each. devioustrevor had USA first and Mexico second for nothing from USA and 1.5 from Mexico.

For positions within the bottom four, I gave 1.5 points for having a team in the correct position, 0.9 for having a team one position off, and 0.3 for having a team two positions off:

1. jagum: 3.6
2. olephill2: 3.0
Tied 3. EvanJ: 2.7
Tied 3. Dr. Gamera: 2.7
Tied 5. barnabyjonez: 2.1
Tied 5. IASocFan: 2.1
7. devioustrevor: 1.5
8. pipilesunidos: 1.2

The mean was 2.3625.

The Octagonal was out 59.500. Here are the score for the five components of the Octagonal combined:

1. barnabyjonez: 46.916
2. olephill2: 42.658
3. IASocFan: 40.758
4. EvanJ: 38.963
5. Dr. Gamera: 34.970
6. jagum: 25.100
7. devioustrevor: 24.913
8. pipilesunidos: 13.606

The mean was 33.4855.

Here are the final standings. Only the top five people predicted every round:

1. barnabyjonez: 77.681
2. EvanJ: 74.625
3. IASocFan: 74.230
4. Dr. Gamera: 69.427
5. jagum: 57.945
6. olephill2: 42.658
7. italiancbr: 29.871
8. Paul Caliste: 29.143
9. devioustrevor: 27.909
10. edcalvi: 14.639
11. pipilesunidos: 13.606

The mean was 46.612. The mean for the people who predicted every round was 70.982.

Thank you for predicting. I wonder what qualifying will be like next time if all three hosts do not have to qualify.

IASocFan, please unsticky this topic.