The winner of CONCACAF'S third place match on Wednesday will qualify for their first World Cup, who wins?
I do not see two things everyone seem to always say Jamaica has. 1. Jamaica is fast - To me watching them play Jamaica seems to have average speed in active play. They seem, even against poor teams, to be thinking and acting well behind the play on the field. They may be fast in a straight line but that is usually negated by how slow they react to situations. 2. Jamaica has good defense - That is just not the case. They are one of the most disorganization teams I have ever seen. The thing that mainly seems to keep Jamaica in matches is size and their opponents are always over estimating their speed and playing them soft. As far a Panama goes I think as long as their coach does not over think the lineup and plays their best players, particularly their best keeper, they are currently well in form and playing very well. Panama beats Jamaica by three or four unless Panama plays the keeper they played in the last match in which case it is a toss up.
Unluckily, I missed all of the CONCACAF Championship full matches (they were mostly happening in the late evening or even in the middle of the night for me in Europe) and the one-minute-length official highlights I later managed to see weren't enough to actually gauge the respective levels of the teams. Anyway, judging by the results and by the posts on the various threads here on BigSoccer (that I have read quite extensively), I'd say that Panama look like a slightly better team than Jamaica. I could be wrong, but I'd say that upsetting Mexico by 2-0 was more difficult, on paper, than upsetting Costa Rica by 1-0: so Panama look a little more solid to me than Jamaica, whose win vs Costa Rica could even be seen as quite random, since Las Ticas had basically equalized by scoring a goal that was probably unfairly disallowed. Yes, it's true, Mexico could have scored too, but a missed PK is just part of the game and it's something very different from a regular goal that was disallowed: having a potential PK-stopper as a GK is a strong point for a team, not just random luck. And anyway Panama won by a 2 goals margin (although the second one was at the end of the game, when Mexico was desperately pushing for a goal). So, I'd say that Panama, on paper, should be marginally better than Jamaica, but I guess the levels are close enough that a Jamaica's win shouldn't be considered as an actual upset.
Jamaica is ranked #64 by FIFA's system and Panama #66. A total of 11 points separate them, which is an insignificant amount. The game is at a neutral site. So, from a statistics perspective, it's a toss up. As a stats person, I follow the stats. So, having to go for one team or the other, I went for Jamaica. A tie probably would be a fair result, but predicting ties is a statistically losing proposition.
So you're basically saying that what tips the scales for you is that slight advantage in rankings for Jamaica that you called yourself an "insignificant amount"? But now tell me: what if Panama, having upset Mexico (that was sitting at #24) while Jamaica have upsetted Costa Rica (that was #34-ranked), would have gained enough points to be virtually ahead of Jamaica in the rankings? FIFA Women's rankings are updated each three months, but teams actually gain and lose points in-between updates also. Wouldn't it be more statistically relevant to calculate where Panama and Costa Rica are ranked right now, instead of basing on the update of 28th of September?
Yes, you're right the current rankings would be better, but unlike for college Division I women's soccer, I can't do current rankings for international teams. (I could do them, if I wanted to spend a lot of time programming FIFA's formula for women and entering years' worth of results for games, but that's a little much for me.) So maybe I should say I'll go with whichever team is higher in the rankings, which so far as I know is Jamaica although they might not be. I'm not a fan of using my own judgment based on having watched them play and I'm also not a fan of "current form" as a basis for decisions. That's why I stay with the stats. Not that they're good predictors -- probably right about 70 to 75% of the time. Rather, they're better than the alternatives.
Well, I can do a quick a back-of-the-envelope guess for where JAM and PAN are right now... I'm gonna assume their opponent's ratings are static (from the previous release), just to give the rough idea of changes to date. JAM: 1386 v CAN2014, exp. 0.026, act. 0.080, K=45 :: +2 1388 v CRC1658, exp, 0.174, act. 0.850, K=45 :: +30 1418 v CUB1238, exp. 0.738, act. 0.990, K=45 :: +11 1429 v USA*2214, exp. 0.011, act. 0.010, K=45 :: 0 1429 PAN: 1376 v TRI1443, exp. 0.405, act. 0.960, K=45 :: +25 1401 v USA*2214, exp. 0.009, act. 0.020, K=45 :: 0 1401 v MEX1732, exp. 0.129, act. 0.920, K=45 :: +36 1437 v CAN2014, exp. 0.035, act. 0.010, K=45 :: -1 1436 So while I'd expect to be a few points off in either direction due to using static values for the opposition, it certainly looks like PAN should have caught up to and probably passed JAM in "real-time rankings". The big difference here is that T&T were rated above PAN while CUB were rated well below JAM, so PAN made up a ton more points from that game. JAM did "much" better against CAN than PAN did in terms of expected-v-actual results, but the points difference isn't nearly as significant - and was likely offset entirely by the respective upsets over CRC and MEX. (PAN did better against the USA than JAM did, too, though you'd only see it in the next decimal point.)
Thanks, @SiberianThunderT. I had the feeling Panama could have passed Jamaica, but your calculations comfort my hypothesis. So, maybe, @cpthomas should have made a different prediction, not just on a "current form" basis, but on actual stats correctly updated to the day of the match.
Right, too bad I can't change my prediction. So, it's Panama over Jamaica since a tie wasn't a choice in the poll. (A game that goes to KFTM would count as a tie.) SiberianThunderT, is there a table that shows expected win/loss/tie percentages based on teams' rating differences, for the FIFA rating system? It looks to me like for the 7 point rating difference between Panama and Jamaica, if I consider a tie as a not possible result, the likelihood of a Panama win must be in the vicinity of 50.1%. Or, if I consider a tie as a possible result, the likelihood of a Panama win is less than 50% but marginally greater than the likelihood of a Jamaica win, with both of those likelihoods greater than the likelihood of a tie.
Not a table, but an equation: take the difference in ratings D, and the expected results (out of 1) is then 1 / (1 + 10^(-D/400) ) It's basically a logistic function, so the expected result slides away from 0.50 before leveling off, in either direction. Still, for a D of just 7 points, you're not gonna get too far away from 50%. You'd get the expected result for PAN of 51.01% and for JAM of 48.99%. Sadly the "result" is basically binary, i.e. it's only win %, so it doesn't predict the likelihood of a draw. But what FIFA considers the narrowest of wins - a high-scoring game decided by a single goal, the actual result percentages are 80% for the winning team and 20% for the losing team, while a draw earns anywhere from 47% to 53% for both teams. As such, expected results inside that range almost certainly mean a draw is predicted, while an expected result somewhere between, say, 53% and 80% is murkier on what it's "predicting". Actually making a table (or maybe a scatter plot) of expected-versus-actual results would be a fascinating exercise, IMO. I've considered it in the past - around the same time I considered making a functional website to play around with the FIFA women's rankings - but grad school takes most of your focus. X-D
Interesting article on Bailey, Panama GK sensation: https://www.goal.com/en/news/panama...ut-star-at-concacaf/39degofahbcv1p42mtbgwjc0g "'Yes, that’s true. I’ve been a goalkeeper for a year,' Bailey told Goal."
Very good article. I was impressed by the part where it says that Bailey was supposed to play through a rotator cuff injury. I had an injury at the rotator cuff when I was younger and I used to swim regularly: not only I couldn't actually swim through it for months, but it was painful even to just lift my arm!
I guess one of the benefits of being an oldster is I get to play around with stuff like that, although I do it using the NCAA's RPI as used for Division I women's soccer. Based on work I've done comparing that system to other systems, although academically trained statisticians hate it, its expected versus actual results are on a par with theoretically correct systems. This table shows the result probabilities for the RPI based on rating differences between opponents. For example, after dividing the games into 5% groups based on the teams' rating differences, the group with the greatest rating differences has a median difference of 0.1795; the higher rated team wins 98.6% of the time, ties 1.2%, and loses 0.2%. One of the things the table shows is that you never should "expect" a tie. It also shows that for the 15% most closely rated games, you really shouldn't "expect" a particular result at all -- the result of a tie or the other result from the one you expected has a higher likelihood of occurring than the one you expected. (Thus "expecting" a Panama or Jamaica win, or a tie, is more likely to be wrong than right!) And, just to be fancy, here's a chart based on the table:
Jamaica, 3 votes Panama, 8 votes Well, that settles that. Minor footnote: Jamaica won the game BELIEVE IN REGGAE Concacaf seems to be claiming the Reggae Girlz won it 42-22, like it's rugby or something, but that can't be right
I think a lot of people were banking on Bailey preventing Jamaica from doing much damage. And to be entirely fair, she did just about all she could to keep PAN competitive - she just didn't get enough of the help she needed from her field players.
And anyway, the field players managed to equalize (twice!!!) when it looked like Jamaica could have run away with the match (both in regulation and extra-time). In the end, as @cpthomas had predicted, it was a written-on-the-wall draw if there ever was one, and in fact it had to be decided at penalty shootout.
If Bailey had had even basic defense in front of her I think Panama wins pretty easily. But the Panamanian defense seems to have successfully avoided the area in front of the goal for much of the match. They left Jamaican players alone in front of the goal so often that I thought, at times, that Panama had chosen to play without central defense. Once they managed to get into the shootout Bailey did make some mistakes by guessing rather than reacting to the shot. She saves at least one if she just stays still. But again the Panamanian field players let her down by passing the ball to the Jamaican keeper. Jamaica was stronger, by a pretty good margin, in the field and even great play from a keeper cannot make up for the poor play on the field. After all keepers should touch the ball much less than anybody else on the field. Now Panama must focus on the home and home with Argentina that is upcoming and I think they can win that but it will be a challenge due to lack of experience both in games and with travel. I guess the exact location and time of the matches are not yet set but does anyone know which country is given the disadvantage of being hosts first and if any TV or streaming agency is planing on having the matches available in the US?
The Panamanian football federation has taken the play-off very seriously and they have decided the game to be played at Rommel Fernandez Stadium the biggest facility in the whole country. Apparently the first game in Argentina will be on November 9, while the second leg will be played on November 13th here in Panama City and there is already a lot of radio and TV advertising for the girls. https://www.prensa.com/deportes/Partido-repechaje-seleccion-femenina-Rommel_0_5148235163.html https://www.panamaamerica.com.pa/de...ya-tiene-fecha-para-recibir-argentina-1118923
I wasn't actually aware of the individual awards in this torunament, but apparently "la portera panameña Yenith Bailey se ganó el Guante de Oro del Premundial Femenino de la Concacaf". So, Golden Glove for Bailey, and well-deserved, according to everyone who saw her in action and wrote on these boards.
I think the voting, largely based on Bailey, helps illustrate a difficulty for humans in making predictions. She's one player who made a big impression -- as shown by the Golden Glove award -- and it's a recent impression. But, in the bigger picture, she had 10 other teammates on the field. Given her great performance, it seems like for predictors the significance of her teammates had a tendency to recede into the background. One of the reasons I like statistical history as a basis for predictions is that the stats over time aren't affected by big and/or recent impressions. That doesn't mean humans shouldn't make predictions. I think a good approach to predictions is to see what the stats over time say is likely to happen and then evaluate whether there's new information that says something else is likely to happen. In the "new information" phase of decision-making, however, it's important to be aware that humans have a tendency to overweight the significance of things that have made big impressions and things that are recent.
This article about Jamaica was published after the Concacaf qualifiers – one of the Reggae Girlz' sponsors is the Bob Marley Foundation, and the team was revived in 2014 with help from Cedella Marley, Bob Marley's daughter. The Jamaican FA (JFF) had shut down the team in 2010. - https://www.bbc.com/sport/football/45899523 Jamaica are playing a friendly at Nottingham Forest in England today, to celebrate the 70th anniversary of the ship the Empire Windrush. Also, a useful football fact for any commentators or other chancers who want to sound knowledgeable at the 2019 World Cup: the striker Khadija 'Bunny' Shaw got her nickname "due to her love of carrots".
Notts Forest 0 Jamaica 3 attendance 3,914 more than any FA WSL or FA WC game on the day. Forest play in the third division in England.