Monday, January 13, 2020

2019 Stay-At-Home Bowl: No Winner

With the conference championships now set, there is going to be no Stay-At-Home Bowl winner this season.

I.e., the teams that beat the Chiefs or Titans did not beat the Packers or 49ers, and vice versa.

Friday, January 03, 2020

2019 Stay-At-Home Bowl

I never did a final post about last year's Stay-At-Home Bowl, mainly because nobody won.

As a reminder, the Stay-At-Home Bowl is "the honor I bestow upon a team that has beaten both Super Bowl participants."

One final bit of business from last year's season, though, the 2017 winner, Kansas City, lost in the 2018 AFC Championship game. Now 12 of the 20 winners have made the playoffs the following year.

The possible winners this year are:
  • Baltimore Ravens
  • New Orleans Saints
  • Kansas City Chiefs
  • Green Bay Packers
  • Philadelphia Eagles
  • New England Patriots
  • Miami Dolphins 
The most unexpected entrant there is the Dolphins, who started 0-7 before going on to beat both the Eagles and the Patriots.

The Washington Post wrote an article about the most likely Super Bowl matchups. You have to get to the 9th-most-likely game (Patriots/49ers) before you have a winner (Ravens). The 10th-most-likely game (Ravens/Vikings) also has a winner (Chiefs).

Sunday, December 22, 2019

Geocaching: Fall 2019

This fall, I found 17 caches, the same number as I did in summer. Most (13) of those were on a trip to the Monterey, California area, which makes this my busiest December ever.

In October, I found 1 cache on a camping trip to Culpeper County (County 99).

Asilomar Whistle Stop
In November, I found three caches, including my first March 2011 one (Month 192).

My first day in the Monterey area, I found four caches. The first was convenient to my conference and right near an old railroad whistle stop. The next three were in downtown Monterey, including two virtuals and a traditional from October 2019 (Month 193).
Monterey Virtual

On my free afternoon, I found seven caches. First I found two challenge caches. One was for finding caches in two states in one day (I once did three), and the other was for finding caches in a total of seven states, five countries, and three continents. (I'm at 19, 12 and 4.)

Then I drove to a nice spot on the beach where there was a December 2019 cache (Month 194) and another nearby. Continuing north, I found a highly-favorited one and a virtual at a Giant Artichoke.
Giant Artichoke Virtual

Finally, I drove up to Santa Cruz County and found a cache there to give me my 100th county!

My last day at the conference, I found a January 2015 (Month 195) near the conference by a nice coastal view, and then a September 2018 one near my hotel (Month 196).

That drops me to 40 unfound months. I think that's a new low. I now have all of the months consecutively from June 2013 to September 2018, which is a nice streak of 64. I have 7 months missing May 2013 or later, so there are maybe 33 "hard" months. (They're not really that hard; 31 of the 40 are available in Fairfax County alone.)
Months Found

Counties I've found geocaches in

Monday, September 23, 2019

Geocaching: Summer 2019

This summer, I found 17 caches in 5 states and 2 foreign countries, which made it my busiest season than last summer (which was more than twice as busy, so I'm still not at my old pace. First, I found 3 caches on a June visit to England, including my first from June 2009 (185th hidden month). No caches in July, but on a drive from California in August, I found caches in Imperial County, California, Yuma County, Arizona, and Maricopa County, Arizona (Counties 91, 92 and 93).

Returning to the East Coast, I was first-to-find on one in Maryland (FTF #21, 186th month).
The blank log on my FTF.

I also found one in Virginia for International Geocaching Day (63rd souvenir).

I found two in Canada, including one from June 2019 (187th hidden month).
Me at Canadian Virtual Geocache

I found another one in Maryland, then headed to Wisconsin for my last six finds of the summer. I added Columbia, Sauk, Iowa, Lafayette, and Green Counties for Counties 94-98. I also added May 2004,  August 2010, April 2019 and July 2019 for Months 188-191.
So I added 7 months, while 3 new ones appeared, for a net +4, and down to 42 missing. I was at 41 last summer, so I haven't been keeping pace for the last year.

Saturday, September 07, 2019

Scotland's UEFA Coefficient: 2019/20 Group Stage Preview

We have group stage! For two clubs!
Just like last year!

I included Greece and Cyprus in this, not because they are in Scotland's "neighborhood", but because under the new 2021/22 system, 15th place is what is needed to get 2 clubs in the Champions League and the other clubs in the Europa League (instead of the new "Europa League 2"). That's pretty much out of reach. I guess it matters if they get to 17th to see if the Premiership champions enter in the first or second round of the Champions League, but that seems small potatoes.

Playoff-Round Review

Celtic won the home leg 2-0 before dominating AIK in the away leg 4-1 for a convincing 0.5 points for the coefficient.

Rangers had a scoreless draw away, and was heading into extra time with the same scoreline until they pulled out a 1-0 win over Legia Warsaw. 0.375 for the coefficient.

Scotland stays in 19th place, but more importantly has two clubs alive.

Group-Stage Preview

Celtic (Elo 1595) got drawn with Lazio (Elo 1672), Rennes (Elo 1641) and Cluj (Elo 1574). Remember, Cluj knocked them out of the Champions League.

Rangers (Elo 1542) ended up in a group with Porto (Elo 1809), Young Boys (Elo 1669) and Feyenoord (Elo 1587).

Here we switch to the Fivethirtyeight ratings, because they give match-by-match predictions for the group stages. For reference, the ratings are:
  • Celtic 66.1
  • Lazio 69.9
  • Rennes 60.9
  • Cluj 45.3
  • Rangers 61.6
  • Porto 78.4
  • Young Boys 72.0
  • Feyenoord 57.8

So clearly the Fivethirtyeight model likes the Scottish clubs more than Clubelo does.

In fact, they give Celtic a 66% chance of making it out of the group stage, and Rangers a 35% shot.

What we can do, though is translate the match-by-match win/draw probabilities into an expected number of points for Scotland's coefficient. The expected number of wins is about 4.6, with about 2.9 draws, giving roughly 3 more points for the coefficient, bringing it up to 7.5. That is usually a top 10 coefficient, which would demonstrate that Scotland is truly on its way back.

In fact, we can do something similar for all the nations in the 15-22 range to project the post-group stage standings. (No countries are expected to drop into or out of this range. The projections are:
  • 15. Cyprus 26.67
  • 16. Greece 26.1
  • 17. Scotland 25.62
  • 18. Serbia 25.33
  • 19. Croatia 24.98
  • 20. Switzerland 24.05
  • 21. Norway 22.33
  • 22. Sweden 22.17

2020/21 Sneak Peak

Since Scotland is unlikely to move into 15th place this year, let's look at the race as it is likely to stand next year (which, I remind you, affects the 2022/23 leagues.)

Scotland starts barely below 15th. If we go a step further and add the expected points from the group stage, we see:
  • 14. Scotland 22.62
  • 15. Denmark 22.03
  • 16. Serbia 21.13
  • 17. Greece 20.7
  • 18. Croatia 20.48
  • 19. Czechia 20.17
(I skipped Ukraine because they will probably pick up a bunch of points in the group stage.)

So over five years ago, I started wondering about the then-Celtic manager's complaint that, "It is stupid the Scottish coefficient is not regarded as good enough." And next year, the Scottish clubs should be perfectly placed to do something about take effect in 2022.

    Monday, August 19, 2019

    Scotland's UEFA Coefficient: 2019/20 Playoff-Round Preview

    Could have gone better, but definitely could have gone worse.

    Third-Round Review

    Celtic picked up 0.125 for the coefficient in an away draw to Cluj and seemed poised to progress as late as the second half of the home tie, where they were leading 3-2. Then they dropped the match 4-3, meaning only 0.125 total and falling to the Europa League.

    Rangers looked absolutely dominant, winning both legs against Midtjylland, 4-2 and 3-1. 0.5 for the coefficient, and they should have an easier time the next round.

    Aberdeen lost both legs 2-0. Absolutely dreadful, and another third-round exit for Derek McInnes. They depart having contributed 0.875 to the coefficient.

    That adds up to 0.625 points, staying in 19th place, and still having more points from this year than any other nation.

    It's more points than they had last year at this time, and last year was spectacular. So it's hard to argue with success, though there are some "if only"s in there.

    Fourth-Round Preview

    Celtic (Elo 1568) is up against Swedish champions AIK (Elo 1525). Clubelo gives Celtic a 56.2% chance of advancing.

    Rangers (Elo 1549) faces Legia Warsaw (Elo 1420). Clubelo gives Rangers a 69.8% chance of advancing.

    I think there's a strong hope for two teams in the group stages again, and the bounty of points that is possible there, but there is a slight chance of crashing and burning and seeing nobody progress.

    In particular, if you believe the odds (and Rangers in particular has a history of overperforming odds in the past year), there is about a 40% chance of having both progress, and a 13% chance of neither progressing.

    Where We Stand

    Not much changed in Scotland's neighborhood. Norway moved ahead of Switzerland. Everybody is down two teams, so there's no real advantage there (though having 3 of 5 left is a slight advantage, particularly if one is entering the Champions League now). Scotland could easily pass Serbia, and hypothetically Croatia. In turn, a bad round would see them drop below Norway, and there's a mathematical possibility of dropping below Switzerland.

    But the most important thing is to get teams through to the group stage (and see their rivals not manage the same).