I guess I just don't see it as much of an issue. If your predictive model indicates that a team only made the playoffs in 5 of 10,000 attempts, then they effectively have a zero percent chance of making the playoffs based on your model. Your model could certainly be wrong, but you're reporting what your model output. I think so long as it's explained that zero is not actual zero but effective zero, as the Fangraphs post does, there's no problem. Is it really practical to write out dozens of decimal places just to show how small the actual chance is? We're talking about playoff predictions, here. Of course, now I'm wondering under what circumstances a team could have an actual zero percent chance of making the playoffs. I can only think of a team being punished by being banned from playoffs for some reason, or if a team ceases to exist and thus no longer plays any games in the league at all (which you wouldn't bother putting into your predictions for obvious reasons).
On the other hand, I don't see the need to run ten thousand simulations to tell me what I already suspected intuitively: the Quakes are vanishingly unlikely to win the Wooden Spoon for 2019 -- but it's not impossible. I've run zero simulations (and I do mean "zero" ) and yet I arrive at the same place, albeit without a number attached. What's the point of crunching numbers to validate what we already knew? A false sense of certitude?
Matthew DoyleVerified account @MattDoyle76 Through 4 games San Jose are on track to concede 119 goals this season. #SJvLAFC 2:58 PM - 30 Mar 2019
worst to fifth - let's not get ahead of ourselves. Too much hype. Not even Almeyda believes we're as good as everyone says we are right now.
Not really. Neither the Seattle Sounders , the LA Galaxy or Minnesota United lost their first four games. If the Quakes won or tied one of their 7 games they lost, things will have been much different. Currently they are two wins from Seattle. Things can always go bad but the way this team has been playing, I don't think there is any turning back. I hope they continue their trek.
All the hype is based on the same idea - chop out results from March and then extrapolate the trend from April. The idea is that the games in March don’t count because we had to learn a new system and new identity of playing. There are some problems with this though. At what point did the team learn this new system and identity? By the above logic it would be convenient to say the new system was learned in between our loss to LAFC and our first win against Portland. Although common sense tells us improvements come much more gradually than that. But that would mean we weren’t as bad as we thought we were in March and we’re not as good as we think we are now. That would mean we’re a fifth place team and not a second place team. Yes we’re improving, but the improvements are happening gradually and we shouldn’t think we can ignore the first four games (those were all tough games btw). https://sbisoccer.com/2019/07/from-...eyda-has-earthquakes-on-remarkable-turnaround https://www.sjearthquakes.com/post/2019/07/16/recap-quakes-vs-real-valladolid-cf
March was Stahres team. April until now is Almeyda’s. The difference is like night and day. I haven’t seen a transformation like this since Dom’s 2005 team but I honestly think this team is better... The league is a lot tougher as well....
You're working backwards, though. At the beginning of the season, the Quakes had a much higher chance than zero of repeating wooden spoon recipients. I'm not sure how many people genuinely believed they would be the worst team in the league again, but it was looking pretty bad based on what we knew. And If you look at any single week, I think you could reasonably guess the odds if you pay attention to the league pretty closely. But it's about tracking that prediction over time, not taking a single week, or even a few weeks, in isolation. You look at where you start, and ultimately where you end up with the actual results to see how accurately you could predict the outcomes. I will say that FiveThirtyEight's model does seem to have some potential gaps, at least when it comes to MLS. There is no account for coaching, as far as I could tell. Based on observational evidence, it would seem that coaching would be an important factor in how many games a team wins. So the question then becomes how much does coaching affect a game's outcome, and how do you evaluate coaches? Is their model still accurate despite not factoring in coaching? My suspicion is that coaching is more important in MLS than it is in other leagues, but before I could be more certain, I would need to test it. I'll level with you, too. I don't really care much about playoff predictions in particular. I think predictions and how they relate to true talent level is interesting, but the focal point of whether a team will qualify for the playoffs is not interesting. The league has shifted so much in such a short amount of time that making the playoffs doesn't really mean the same thing year to year.
I agree that the playoffs mean nothing unless you qualify , make or win in the finals stages. Just like I thought backing into the postseason and losing 0-5 to the Caps in 2017 was a failure, making the postseason in this league is not a great achievement. This team is playing differently though and all they needed was a boost. No one across the league or on this board believed this type of transformation could take place. Last year or even in March looking at the LAFC game specifically, they weren’t playing with any will , confidence , tactic , game plan or system in place. This year is a totally different story as they are now among the best and most talked about teams in the league. Other than Vega and Espinoza and a few new part timers, they have virtually the same team intact. I never believed these players would be able to play as they have been. I’m optimistic that there is only room for growth and improvement. Keep it going Matias & go Quakes!
Weave a better goal difference than the Gals and beat them twice head to head so shouldn't the Quakes be in fourth place overall? & third in the west?
https://www.mlssoccer.com/standings/scenarios Looking at the playoff brackets in the west, it is set up so that the 4th and 5th teams play each other, and the winner of that match would play against the 1st place team. If playoffs started today it would be EQ vs Loons, the winner then goes against LAFC. Meanwhile seed 2 (Seattle) plays against Seed 7 (Portland), while 3 (Galaxy) plays 6 (Dallas). The winners from these two games would then play each other. The winner of that game would go up against the winner of the LAFC vs EQ/Minn game. I am not going to predict any playoff games other than to say I think the seeded teams will not be in the same order at the end of the season as they are now.
Least bold prediction ever. Even if the gaps below LAFC at the top were not 1 point, 0, 2, 2, 1, 0, 1, with a crapload of games left to play....
Honestly , the way this team has been playing, it will be very difficult to stop them. LAFC beat us 5-0 at home in March but that was a different story. Since then , the team has improved and if these newcomers can add anything, I don’t foresee too many teams that can beat the Quakes...