Jump to content

How many points will it take to get through?


Lurker

Recommended Posts

To be on the safe side you would hope for a possible 9 from our 3 home matches, and maybe worst case, 7 if we drew vs Mexico.

For the away matches, I'd like to say we can realistically look to maybe getting 2 to maybe 4 points. A win and a draw, or maybe 2 draws wouldn't be so bad. We need to ensure a result at Honduras and try to draw Jamaica in their stadium at worst.

So overall, I think the magic number is probably 10 or 11 just to be safe. Any combination of 3 wins and 1 draw will probably be enough to get us through. Also, if we were to beat Jamaica 2x on both legs, I think that would almost cement our place through if we get results vs Honduras.

Link to comment
Share on other sites

I'm learning a new piece of statistical software and if I have time over the next few weeks, I'll try to statistically simulate how many points are needed to advance to the next stage. I'll do so in the following manner:

1) Baseline hypothesis--Assume that the probability of a win for each side is 1/3, and a tie is also 1/3.

2) Use historical records to determine the actual number of games in CONCACAF that are a) won by the home team; B) won by the away team; c) tied.

3) Build on 2 above by incorporating some measure of team quality, using most likely historical records and current FIFA ranking.

Can anyone think of any other data that could be used to determine winners (or tied) in the games coming up?

quote:Originally posted by theaub

Well there's a max of 36 points (12 games X 3), so it really all depends how many draws there are. If there's none, then 12 gets you in (or 9 with a good goal differential). If there's 4 ties (which I would say is about the average), 9-10 should be good enough. If there's a lot of ties even 7 or 8 may be good enough to squeak into 2nd.

Link to comment
Share on other sites

The first two games are so so crucial if we can get 6 points there we will be in the drivers seat handing our two closest rivals losses then we need to steal some points on the road from the 2 away fixures vs them (say 2 or 4pts) and I think we are in good shape barring Mexico coughing up a lot of points.

Link to comment
Share on other sites

quote:Originally posted by MediaGuy

FWIW:

Since 1994 the average amount of points needed to finish second in the semi final groups has been 10.36.

Did you count 3 points for a win for qualifying in 1994? I did count 3 points for a win in 94 and my calculator say it is an even 11.

I think 10 will get the second placed team through. We don`t have Barbados or Saint Vincent & Grenadines or other minnows. Jamaica, Honduras and Canada will probably all get points or steal points from each other.

Link to comment
Share on other sites

To be guaranteed progression we need 10 points (just like the champions league). 8-9 points makes things sticky. But 10 gives you that extra win over the third/fourth place teams thus putting separation in.

When you have problems is when everyone whips the bottom team in a 4-team group, and then team A beats B, and Team B beats C, but team C beats A.

10 points. We will get 11, I predict.

Link to comment
Share on other sites

quote:Originally posted by Kusch to the Corner

Mexico won't get 18 pts,

It would make things much easier if they did, actually. My biggest worry is a scenario that has Canada eliminated because Mexico drops unexpected points. I'd far rather they just run away and hide than start slow, or mail it in at the end (unless it's against us, of course).

It's hard not to get bitter thinking about this group...but..

We make the hex, we're going to South Africa.

Link to comment
Share on other sites

quote:Originally posted by MediaGuy

It would make things much easier if they did, actually. My biggest worry is a scenario that has Canada eliminated because Mexico drops unexpected points. I'd far rather they just run away and hide than start slow, or mail it in at the end (unless it's against us, of course).

It's hard not to get bitter thinking about this group...but..

We make the hex, we're going to South Africa.

So what you're saying is realistically we can't expect to control our own fate and Mexico is our friend. [8D]

corrected: for spelling

Link to comment
Share on other sites

Aug 20 - Jam @ Can, Hon @Mex

Sept 6 - Hon @ Can, Mex @ Jam

The first two weeks are absolutely key. We need six points. If Mexico does what it should through this stage and we hold serve:

Mex - 6

Can - 6

Hon - 0

Jam - 0

That, I like. It sets up a must win for both Honduras and Jamaica in match-day three:

Can @ Mex, Hon @ Jam

Let's assume we lose to Mexico in match day three. Ideally we want a tie, but let's go with a win for Jamaica (we're not going them enough credit)

Mex - 9

Can - 6

Jam - 3

Hon - 0

Match day four is, to me, our best chance at a road win.

Can @ Hon, Jam @ Mex

Mexico clinches with the win, we get a draw (We can't expect things to go exactly as hoped)

Mex - 12

Can - 7

Jam - 3

Hon - 1

We're now two points away from clinching. If Mexico doesn't mail it in, we're laughing.

Match day 5:

Mex @ Can, Jam @ Hon

The Mexicans beat us at home, causing several Vs to jump off a bridge, but Hondouras saves our ass with a win at home.

Mex - 15

Can - 7

Hon - 4

Jam - 3

So November 19 dawns with us playing an eliminated Jamaica in Kingston and already through Mexico on the road against a team they have struggled with on the road. We're all going to need a stiff drink and a calculator in our hand to figure out the tie-breakers.

I say we clinch it with a late goal that ties are game at 1-1. Good thing too, as Mexico crapped the bed against the Hondurans. Too bad, too say for them. It's someone else's turn to cry over the injustice of CONCACAF football.

Final

Mex - 15

Can - 8

Hon - 7

Jam - 4

Link to comment
Share on other sites

^ MD3 is Jamaica at Honduras and MD5 is Honduras at Jamaica - you have them reversed.

Honduras start with 2 road games then play at home to JAM and us, before going to Kingston and finish at home to the Mexicans.

Jamaica start off with 3 road games in their first 4 games. Their only home game is MD2 against Mexico. Then they finish at home to Honduras and Canada MD's 5 & 6.

Link to comment
Share on other sites

As many have already stated, the opening two games for us are crucial. 6 points taken from Jamaica and Honduras that early in a group where any of the four teams can beat the others is very important.

I see this group group being very close, and it might depend on whether Eriksson can gel with the Mexicans or if they stumble and drop points to the rest of the group. Either way, I think we will need at least 10 points to go on... Preferably 18 points though.

Can't wait for the 20th!

Link to comment
Share on other sites

quote:Originally posted by bettermirror

When you have problems is when everyone whips the bottom team in a 4-team group, and then team A beats B, and Team B beats C, but team C beats A.

10 points. We will get 11, I predict.

Agreed, if I am reading you right, I think this is kind of what happened to us in 2000. Everyone was beating Panama but in our very first game we came out and took the game to T&T. we couldn't finish worth lick and T&T had dwight Yorke. Also, in retrospct, it wasn't Forrest's best game. we lost to TNT at home and were essentially eliminated after two games because TNT went and got a draw against Mex.

Its not good to be in a tough group, but it also has its benefits. take 2004 for example, we were in a very tough group. But by the fifth game in vancouver, we still had a chance.

Link to comment
Share on other sites

quote:Originally posted by Free kick

Everyone was beating Panama but in our very first game we came out and took the game to T&T. we couldn't finish worth lick and T&T had dwight Yorke. Also, in retrospct, it wasn't Forrest's best game. we lost to TNT at home and were essentially eliminated after two games because TNT went and got a draw against Mex.

Its not good to be in a tough group, but it also has its benefits. take 2004 for example, we were in a very tough group. But by the fifth game in vancouver, we still had a chance.

Onstad was the GK for the T&T match in 2000. You point however is

well founded. Luckily, Pat seemed to play much better since

then, including the last GC.

Link to comment
Share on other sites

What I find the most puzzling with this group is the question of who the weakest team is. At first glance I would say Jamaica, but then I wonder if Honduras is really all they're cracked up to be. Is it realistic to think that Jamaica could draw or even win against Honduras? Is it possible for Mexico to be drawn or beaten by Honduras? I think the Hondurans are going to be the key to this group, if they suck balls against Mexico but play strong against Jamaica, we could be in trouble (depending on how we show against them).

Link to comment
Share on other sites

quote:Originally posted by brownbear

I'm learning a new piece of statistical software and if I have time over the next few weeks, I'll try to statistically simulate how many points are needed to advance to the next stage. I'll do so in the following manner:

1) Baseline hypothesis--Assume that the probability of a win for each side is 1/3, and a tie is also 1/3.

2) Use historical records to determine the actual number of games in CONCACAF that are a) won by the home team; B) won by the away team; c) tied.

3) Build on 2 above by incorporating some measure of team quality, using most likely historical records and current FIFA ranking.

Can anyone think of any other data that could be used to determine winners (or tied) in the games coming up?

Well, I was able to bang out a simulation quicker than I thought. So, this is a quick and dirty simulation, which assumed that the average rate of ties was 1 every 3 matches (note, that is "average rate" and this was allowed to vary over the 10,000 simulations).

In addition, I assumed that for each match in which there was a victory, each team had an equal chance of winning that match. Of course, this isn't plausible (given home-field advantage), but it works out in the end since this is an evenly matched group and any home-field advantage (for a victory probability-wise) will be about equally offset by an away-field disadvantage in the return match.

So, what is the absolute minimum number of points that will potentially get you through as the second place team?

EDIT: It's actually 4; I accidentally cut off the histogram at 5, which is not correct! (with advantageous goal-differential) The chart below (click on the link below) shows what percentage of times--in 10,000 simulations (that is, I programmed the computer to simulate the 12 games of the group 10,000 times and recorded the results)--a specific number of points is needed to finish second. The average is exactly 11, while the standard deviation is 2.3 (the normal curve is in red).

Note that the mode--the number of points that will get you 2nd place more than any other--is 11 (this occurred in just over 25% of the 10,000 simulations). 10 points will get you to 2nd place just under 7% of the time, while 9 points will get you to 2nd place 16% of the time. This may seem strange, but it isn't given the nature of how points are accumulated--i.e. 3 points for a win, and 1 for a tie.

So, in order to have a better than 50% chance of going through--as the 2nd place team--we'll have to get 11 points. 10 points will leave us with only a 48% chance of getting through, once again as the 2nd place team. There is also a very slight chance (which I'll simulate later) of winning the group with 10 points, but I'd guess it's less then 3%.

Enjoy!

photo#5220469688863100434

Link to comment
Share on other sites

quote:Originally posted by redhat

Onstad was the GK for the T&T match in 2000. You point however is

well founded. Luckily, Pat seemed to play much better since

then, including the last GC.

Your right. It was Onstad. and TNT actually beat Mexico in that round

http://www.fifa.com/worldcup/archive/edition=4395/preliminaries/preliminary=3881/matches/match=20294/report.html

Link to comment
Share on other sites

quote:Originally posted by brownbear

Can anyone think of any other data that could be used to determine winners (or tied) in the games coming up?

Depends who you can bug to help. If you're just doing statistics (which really don't help that much these days), the historical home and road records of each team (e.g. Mexico at Azteca, Mexico at Puebla) as well as any head-to-head records (which are very deceptive because of how infrequently they play) would factor in. The home&road records I think are very pertinent.

Now if you can grab a game theorist, you can properly simulate qualifying and come up with an answer. Two unique challenges that game theory solves that WCQ faces that would be less significant in other sports:

1. The scoring margin that you develop via statistics is typically less than one goal, which means that your statistical analysis has a massive beta (and therefore less statistical relevance). It's pointless to predict a 1.7-to-1.3 score, and the relevance of a 58% chance of victory, 27% draw, 15% loss against similar statistics is limited without game theory plotting the forward-looking outcomes against those probabilities.

2. With such a short qualifying draw, with two of four teams qualifying for the next round, the variance that will occur as teams approach clinching a qualifying spot is a substantial part of the equation. Mexico effectively clinches its spot after 3 games, certainly after 4 (home vs. Jamaica...which is an automatic win according to the stats), and that skews two road results. Game theory will address that, statistics won't. Conversely, if Mexico draws either of its first two games, they know that their last two games are on the road against their most difficult opponents. That affects all four of their remaining games.

Throw in the concept of foreign-based players being unable to join (something that Mexico is encountering almost for the first time), maybe try to factor injuries, weather, travel distances, and a myriad of other relevant factors...and no game theorist I know would dare take this project on. Not without a PhD thesis at the end of the effort.... :)

Link to comment
Share on other sites

quote:Originally posted by Daniel

The most extensive simulation for the SF round: http://www.bigsoccer.com/forum/showthread.php?t=712784

Mexico 13

Honduras 9

Canada 6

Jamaica 4

Although Honduras is at 9.3 and Canada at 7.6 (projected).

Probability of advancing:

Team Prob

MEX 87%

HON 59%

CAN 41%

JAM 13%

An interesting demonstration of the (small) deficiency of statistical analysis. His same statistics showed the most probable point total for Canada as 9 (15% probability). There are only two ways to achieve 9 points - 2 wins and 3 draws, and 3 wins. For the third-placed team (according to his stats) it seems unlikely that either of those scenarios would occur, let alone in sufficient probability to make it the most likely outcome (by percentage). 5 results out of 6 (presumably with a road loss to #1 Mexico) seems rather high for a team that doesn't qualify. And while 3 results is realistic, a pure three (home?) wins and three (road?) losses, especially with a (statistically) poor Jamaica squad and a (statistically) strong Mexico squad in the mix seems unusual. Then look at the results from the last three WCQ group stages - 9 points happened only once in 9 groups (and 36 teams, or 18 relevant opportunities).

The statistics might be 100% valid, but the means used to get to those numbers shows a lack of quality. Anyways...everyone here knows that Canada will qualify, so it's obvious his numbers are broken! [:P]

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...