About time to start counting magic and tragic numbers now, right? I do believe that this past weekend's results have officially eliminated KC from the playoffs. With 5 games left, their maximum point total is 11+15=26, which would put them in a three-way-tie with the current totals of WAS and HOU for the sixth and final playoff spot - but keeping both WAS and HOU on 26 points means WAS loses a future match to NJY, putting NJY on 28pts and officially blocking KC out of the playoffs. LOU is far from eliminated, even after losing to HOU this weekend, but they have been eliminated from a first-round bye. With the middle of the table as crammed as it is, there's no way to tell who else on track for a first-round home match or not, but it's worth starting to consider how close POR (and RGN) are to clinching at least 4th and maybe even at least 2nd. I believe I heard talk that POR could have clinched a top 4 spot this past weekend, but CHI thankfully put a short pause on that. POR has at least already clinched a playoff spot, and RGN's magic number is 2 I believe.
Okay, so I actually misheard the broadcast this past weekend. POR's clinching scenario that CHI spoiled was not for a home playoff game, but for playoffs in general - you can actually see on SportsClubStats in their unweighted simulation that there is a miniscule but nonzero chance that POR misses the playoffs still! It requires them to go 0-5 for the rest of the season, of course, and I assume RGN beating them this weekend (to have the H2H) and then also losing out so that the middle of the table can properly catch up... Of course, this is all still talking about very, very minute possibilities, specifically 4-in-1-billion per SCS - even going 0-5 would still leave POR a 96% chance of ending in the top 4 at this point. Being the numbers guy that I am, though, I just wanted to be thorough and make the very minor correction on my previous post! I may try to suss out exactly what that catastrophic scenario might be, but it'd just be academic curiosity. I'm a bit more interested in doing a strength-of-schedule look for CHI through NJY to see which of those five are best poised coming down the line.
Alright, I think I found the scenario that might get POR eliminated: -POR loses out -LOU and KC lose out (except KC beats POR) -RGN loses out (except for beating POR and KC) -ORL loses to NJY and CHIx2 (it's impossible for both ORL and CHI to tie/pass POR) -okay... WAS beats NC, WAS beats NJY, NC beats HOU, NJY beats NC, HOU beats WAS -This leaves RGN+CHI on 41pts, NJY on 40pts, and a four-way tie on 38pts between POR, NC, HOU, and WAS. I'm not sure how all the tiebreaks play out in this scenario, but it's the only way I could find a set of results that didn't have four teams outright below POR = = = = = Anyway, here's the other thing I wanted to look at - strength of schedule (S.o.S.) down the line for the middle six teams! Looking at who everyone faces down the line, I calculated the average PPG of opponents to face, with the added nuance of splitting home/away PPG. Now, this method of calculation S.o.S. includes a natural bias, specifically in that you can't face yourself, so a team with a very low PPG will naturally have a higher S.o.S. in this method than a team with a high PPG. For comparing the six teams in the middle of the table, this won't have a large effect since they all have very similar PPGs of their own, but just for thoroughness I did a second calculation by taking the original S.o.S. and averaging it with the team's own PPG for a somewhat-normalized S.o.S. score. No matter how you cut it, NJY has the easiest schedule by S.o.S. (though also the busiest schedule left) due to facing the expansion teams 3x but the Cascadia teams 0x, giving them a natural S.o.S. PPG of 1.211 and a normalized S.o.S. PPG of 1.300. HOU and NC have the hardest schedules, both facing POR at least once each and also facing each other, giving S.o.S. PPGs from 1.44 to 1.55 between them between the two calculations. And on facing each other, CHI-ORL happens twice down the stretch as mentioned above, so while they have the 2nd- and 3rd-easiest schedules down the stretch, they have a "12-point series" between them, which could easily doom one or the other - so it would not be surprising for one of them to miss the playoffs despite the "easy" schedule as their fates could easily reflect each other.
Looking at these before bed, and I'm realizing that NJY could just draw NC, in which case NC would be at 39pts and NJY would instead be involved in the 4-way tie at 38pts. I'd have to really dig into the tiebreak rules to see which of the two scenarios would be more likely to eliminate POR. NJY has 1pt in 3g against POR, while NC would have 6pts in 3g if they win the last match of the season, so I'm guessing my original setup is the correct one as it greatly lowers POR's PPG in the theoretical 4-way mini-table.
And now, STT, do it all again based on the possibility that the five games not being played this weekend simply will be cancelled.
No need for big recalculations - if they're outright cancelled, those are 3pts each for WAS, HOU, and NJY in the scenario outlined that then vanish into the ether. Any of those three could get 3pts "back" by beating NC, but then NC joins the crew unable to hit 38pts. That makes six total teams (three from the cancellation + the three already out in the scenario outlined) below POR; POR is qualified not just to the playoffs but to a home game in that case!
Even though POR lost last night, the draw between WAS & NJY means that POR has officially qualified for the playoffs. NC's victory over LOU also means that LOU has been eliminated from a home playoff game - I'd have to look more closely to see if they've been eliminated from the playoffs outright.
CORRECTION: Portland is actually still not qualified to the playoffs! I found one last scenario that potentially eliminates them: ---POR, ORL, and LOU all lose out, except for the the ORL-LOU H2H game ---KC almost loses out except being TBD against CHI and RGN* ---WAS wins out, as does CHI except for TBD against KC and RGN* ---Finally, NC loses to HOU but beats NJY This leaves WAS on 39pts, CHI and RGN on 38-41pts*, and the quartet of {POR, NC, HOU, NJ} all on 38pts. *The TBDs between CHI, RGN, and KC can in theory leave either CHI or RGN beneath 38pts if KC does too well - but if KC loses out, then the CHI-RGN H2H isn't important I have no clue how the mini-table for the teams on 38pts total would actually play out... but it's shockingly still possible for POR to miss out on some incredibly wacky tiebreak scenario. Their magic number is 1, or getting some extra help. = = = = = As for LOU, here's what I'm looking at right now: ---LOU's max is 29pts ---one of CHI/ORL will pass 29pts outright based on their remaining H2H games ---one of HOU/DC will pass 29pts outright based on their remaining H2H game ---Three other teams {POR, RGN, NC} are already outright passed 29pts, for a total of 5 teams, i.e. LOU can't let anyone else outright pass 29pts, so we let those three teams beat mid-table teams ---NJY needs 4pts from vsORL and vsKC to get to 30pts, so if they don't do that, LOU has a 29pt-tiebreak for 6th place that still gives them a chance. But their tragic number is 1.
NWSL just Tweeted that the ORL-NJY result has clinched playoffs for POR. I don't believe that's true, based on the scenario outlined in the previous post. Once the DC-LOU result is final, I'm going to let SportsClubStats run again and see if POR-in-7th still is nonzero. If it *does* go to zero, then there was still some weird scenario that had ORL as one of the playoff team, contract to the interplay between ORL/CHI/NJY outlined above. Either that, or POR qualified a while ago (i.e. the big tiebreak scenario I found never would have tie-broken properly anyway) and none of us realized it... because it sure as hell doesn't look like it should have come down to ORL dropping points to me.
Okay, yes, the unweighted SportsClubStats simulation still finds at least one scenario where POR does not qualify, even after tonight's results. NWSL has jumped the gun in declaring POR's qualification. http://www.sportsclubstats.com/USA/f/NWSL2.html (LOU, of course, has been eliminated)
I mentioned this in the weekly results thread, but the Shield is coming down to a two-horse race: POR and RGN have a 99.2% chance of winning the Shield between them, with only HOU, NC, and NJY having a nonzero chance still (DC, CHI, and ORL are out of Shield contention). Unsurprisingly, RGN's win last night also clinched a playoff berth for them, while CHI and ORL have been blocked off from a first-round bye. At the moment, 42pts guarantees a first-round bye, while 41pts can with other results, so POR and RGN are both basically one win away from that bye. The last interesting things are seeing who qualifies for the playoffs, and who among them earns a first-round home game. I'll play around with some of the numbers and see what scenarios give what. Interestingly, at first glance, it looks like most teams need to climb to 39pts to guarantee a playoff spot, but HOU and NJY can get away with 38 and 37pts respectively, likely because most of their remaining schedule is against other mid-table teams, so them earning points means denying six-point-ers to their competitors. Also, since these "guarantee" numbers depend on complicated tiebreaks, those numbers will ultimately be more likely home playoff game numbers and not simple qualification numbers.
Okay, after some playing around, here are the scenarios I found for Wednesday specifically: POR will clinch a first-round bye IF... ......they win against RGN POR will clinch a home playoff match IF... ......they draw again RGN and one of HOU+NC drops points ......they lose to RGN and one of HOU+NC loses RGN will clinch a home playoff match IF... ......they win against POR and one of HOU+NC drops points ......they do not lose against POR and one of HOU+NC loses ORL will finish no higher than 5th (i.e. eliminated from a home playoff match) IF... ......they lose against CHI and HOU+NC both win their matches There are no other clinching scenarios, but I did notice that the target number for qualifying for the playoffs is probably closer to 35pts than it is to 39pts mentioned in the previous post. This is because CHI+ORL play each other twice down the stretch, which at this point means at least one of them will finish the season below 35pts, meaning that reaching 35pts guarantees you have at least three teams beneath you in the standings. In fact, if CHI beats ORL Wednesday, then ORL's max falls to 34pts - and that's what leads to that final scenario listed above, since both HOU+NC can themselves reach 35pts on Wednesday.
Okay, so I'm slightly dumb - the scenarios above are not the *only* clinching scenarios - if I included ORL failing to make 4th, I should have stated whether anyone couldn't make 2nd as well. Even if POR or RGN do not clinch the first-round bye, it's possible that individual chasers are eliminated from the first-round bye. It's easy for individual chasers to miss out, but POR and RGN just can't clinch unless enough chasers miss out (or they get out of reach themselves). Quick rundown of how to be eliminated from the bye, by team: HOU - loss and RGN at least draws NC - loss and RGN at least draws WAS - loss, draw, or win with RGN win NJ - (not playing Wednesday)
I am mildly following your calculations, @SiberianThunderT, but I must admit that I am not really aware of this season's play-off formula: it sounds complicated.
haha it's less that the "playoff formula" is complicated, and more just that the schedule and standing have worked out to have a right mess with most of the league clustered together in the *middle* of the standings. (Things would be more cut-and-dried if the cluster was closer to the top or two the bottom.)
But what you mean by a "first-round bye"? Aren't the play-off just two semi-finals and a final for the top 4 teams as in the past?
Big night tonight. I'm really looking forward to seeing as much as I can of all four matches. It's been a mundane season (on the playing field) in my opinion, but there have been some really good matches the last two weeks. So, hopefully we'll have a classic or two tonight. I don't think there's a clear leader now in the MVP and Golden Boot awards and Trinity Rodman looks to me right now like the rookie of the year, but we'll see which players come through when it matters most.
Not really. If POR win, they'll be 4pts ahead with 2g left, which is fairly significant, but not insurmountable, especially considering their last two games are much tougher than RGN's last two games. And if RGN win, they'll be just 2pts ahead of POR with 2g left to play, which is even farther from settled.