Monday, July 24, 2017

Scotland's UEFA Coefficient: Third-Round Preview


Well, that went about as well as could be expected.

Second-Round Review


Celtic dispatched Linfield as handily as you would think -- 2-0 in the away leg, and 4-0 in the home leg. I guess some might have expected even more lopsided scores, but a win's a win, and this was two wins and 0.5 for the coefficient. Also, this guarantees Celtic a place in the playoff round of either the Champions League or Europa League.

Aberdeen unexpectedly struggled to a 1-1 draw at home, meaning they needed at least one away goal to make it through. Midway through the second half, they had none, but they came up with two in rapid succession to progress. 0.375 for the coefficient, and they progress. We could quibble with the dropped 0.125 at home, but let's not.

Third-Round Preview


Celtic face the Norwegian champions, Rosenborg. I have to admit, I like the qualifying better than the group-stage, where 15 out of the 32 teams are not actually league champions. Here, however, you have the Scottish champions facing the Norwegian champions. The Scottish champions got there after defeating the Northern Irish champions, who defeated the Sammarinese champions. The Norwegian champions defeated the Irish champions.

Celtic's Elo is 1596, Rosenborg's 1501. It shouldn't be a blowout, but Celtic should win comfortably. Not getting the full 0.5 for the coefficient would be a little bit of a disappointment, but the main thing is making it through. (Clubelo gives Celtic a 65.3% chance of advancing, but that seems a little low.)

Aberdeen is up against Cypriot side Apollon Limassol, the 3rd place Cypriot club (and cup winners). Aberdeen has an Elo of 1393, Apollon Limassol 1444. Clubelo gives this a 57.5% chance to go to the Cypriots. I have hope that this is Aberdeen's year to break through to the playoff round. So let's say any result that sees Aberdeen advancing is good, and exiting without picking up any points is bad.

Where We Stand


Hey, Scotland moved up to 26th from 27th, so that's something. (Not a good couple of weeks for Kazakhstan.) That was my "best realistic scenario". My best realistic scenario for this round is keeping 26th, which should be even easier since Celtic should deny Norway points by beating Rosenborg. 24th is theoretically possible, but very unlikely. On the other side, a Celtic collapse would be the only thing that would let Scotland slide to 27th or 28th.

The good news is that nobody below 21st place has more than 2 teams left in European competition, which puts things on an even playing field moving forward. Sweden lost its Champions League entrant, so they may be out of Europe sooner rather than later. So I think the best case scenario is having both Scottish sides advance with 0.75-1 point, and a lot of the nations above them having only 1 side left in Europe.

How bad has this year been for Scotland's coefficient? On the one hand, the 1.125 is worse than any comparable side ranked higher than Iceland (33rd). On the other hand, last year (not a bad year!) at this stage, they had 1.875. If they picked up 1.875 this year, they would be...in 26th place (although significantly closer to 25th).

The question going forward...will this be a struggle to keep the second-round Europa League entrance due to the Scottish Cup winner (or league runner-up). (This threshhold is 24th or 26th depending on...stuff.) Or can Scotland put two teams in the group stages and move closer to 18th (which would put the Cup winner in the third round) and higher?

Monday, July 10, 2017

Scotland's UEFA Coefficient: Second-Round Preview

First-Round Review


Well, that was a disaster. I was embarrassed by it, until I realized that I am not really Scottish, so I am not tainted by association.

Let's start with St. Johnstone. They should have been able to beat the Lithuanian runners-up at home, but they lost 2-1. When FK Trakai went down to 10 men in Lithuania, St. Johnstone should have been able to get two goals to win the tie. But they didn't. They should have been able to get one goal, to get 0.25 for the coefficient before exiting. But they didn't. They should have been able to prevent the shorthanded goal to salvage the minimum 0.125. But they didn't. And their manager probably shouldn't have described the Lithuanians as "technically better" after the first leg. But he did.

So, on to Rangers. I was excited when before the match, Rangers' manager explicitly mentioned the importance of doing well in order to raise up Scotland's UEFA coefficient. And then they beat Luxembourg's fourth-best team 1-0 at home. You'd think they could do better, but still, 0.25 for the coefficient and all that. Then they went to Luxembourg and not only lost, but lost 2-0, which knocked them out.

The optimistic (less pessimistic) way to look at this would be to say that Rangers, while fielding a team good enough to get third place, knew that wouldn't be good enough, so they had to blow up the roster as if they had just been promoted. The new group of players didn't have time to gel and got upset by a more cohesive squad.

The pessimistic (more pessimistic) way to look at this is that a perfectly-good third-place squad got blown up for no good reason. While that squad certainly would have been good enough to beat the fourth-best side in Luxembourg, this group of overpaid imports is not going to cut it.

Time will tell, but not in the Europa League, because that door is closed for another year.

Second-Round Preview

That leaves Aberdeen and Celtic.

Aberdeen opens things up on Thursday against Bosnia-Herzegovina's Siroki Brijeg. Aberdeen's Elo is now 1381, and Siroki Brijeg's is now 1236. Siroki Brijeg's went up for advancing from the first round, and Aberdeen's tumbled based on the St. Johnstone and Rangers results pulling down everybody in Scotland. clubelo.com lists Aberdeen as 72.7% favorites to advance. So let's say that a win and a draw, worth 0.375 for the coefficient is the target.

Celtic starts on Friday against Northern Ireland Champions Linfield. Friday is an unusual date for a Champions League match -- most will be played on Tuesday (hence why I'm rushing to get this post up). But the match was rescheduled to avoid "marching season" in Northern Ireland, which has its own set of sectarian tensions that will cause the police to have their hands full.

Speaking of sectarian tensions, Linfield is a traditionally Protestant club, and Celtic is traditionally Catholic. So while it seemed like Linfield might get a big payday from Celtic fans traveling over to see their club, Celtic told their fans to stay away. This is the sort of thing that fascinates me about European soccer that you don't get by following American sports.

Anyway, clubelo.com puts Celtic as 98.1% favorites to advance. Anything less than the full 0.5 will be an embarrassment. Though Scottish clubs are getting used to embarrassment, I think there will be enough continuity to manhandle Linfield.

Where We Stand


I said that 27th place would be the "extreme negative scenario".  Here were are in 27th. A good showing by Norway could drop Scotland down to 28th, even if the Scottish teams are perfect in this round. At this point, we have to hope that both Celtic and Aberdeen stay alive into the group stage and climb back up as other countries inevitably have some teams knocked out.

If everything breaks Scotland's way, they could be in 23rd at the end of the second round, but 26th with both clubs alive is the best realistic scenario.

Wednesday, July 05, 2017

Google Local: Rocketing to 500K

Well, it took about two more weeks to hit 500K photos. Nothing magic, just the bunch of pictures I've accumulated over the past few months starting to gain traction.

This Dollar Tree (Chantilly, VA) from April cracked 30K photos and is now #4 overall. Fun fact: I had a raging ear infection when I took that picture.

Four more photos topped 10K.
This Dollar Tree (Crofton, MD) from May is now #12.
This Harris Teeter (Bowie, MD) from June is now #13.
This Dollar Tree (Bowie, MD) from June is now #14. If I cared more about this project, I think I would spend a day driving around and taking pictures of Dollar Trees.
This Michael's (Vienna, VA) from June is now #15. The fact that I am getting pictures to 10K views in about a month tells me I shouldn't have much problem breaking the million view mark. Halfway there!

Also, those latest ones aren't as bad as some of the other popular photos.



Monday, June 26, 2017

Google Local: 400K Photos and Some Free Music

I seem to be getting the hang of increasing my photo views on Google Local (take lots of interior photos, especially of Dollar Trees), so I have hit 400K already. I was pleased that Google recognized this by give me "points" for each photo, and the number of points I had achieved got me to a level where I got three free months of Google Play Music. I was more pleased before I discovered that last month they had a promo of four free months for doing nothing.

Let's look at some photos that have reached milestones rather than going through all of the top photos.

This photo of Michaels (Bowie, MD) from December 2016 hit 50,000 views and remains my number one photo for some reason.

Four new photos have cracked 10,000 views.

This photo of a Safeway (Reston, VA) from May 2017 is now #8.
This photo of a Wawa (Gambrills, MD) from May 2017 is now #9. This is actually one of my better photos (not that that seems to help with views). Also, I had a feeling when I took it that I would get lots of views.
This photo of a Safeway (Herndon, VA) from May 2017 is now #10.
This photo of a Dollar Tree (Sterling, VA) from April 2017 is now #11.

I guess I can add Safeway to Dollar Tree as a reliable source of views. (I have two Safeway pictures from this month that haven't taken off, though.)


I had decided to try taking pictures of food to see if they would get views. The answer seems to be (based on two examples), not more than pictures of interiors.

I'll leave you with a final mystery. Here are three pictures of Targets.


All are from May 2017. Despite the fact that the one on top is the most recent, it has received 7,624 views, while the other two have received 158 and 231. I don't think it's in any way better or more representative than the other two






Monday, June 19, 2017

Scotland's UEFA Coefficient: First Round Preview

It's a very exciting time of year here at the Android's Dungeon, as the Champions League draw...and, more to the point, the Europa League draw has been announced for the first round. Four Scottish teams are alive with potential contributions to Scotland's UEFA coefficient.

Where We Start

So let's see where we are heading into the first round of competition.

Scotland sits at #24, tightly packed in with a group of countries that I have come to view as its peers. So tightly packed, in fact, that I can't imagine them sitting in the same slot most of qualifying, like they clung to #23 in the previous year's rankings.

Recall that in qualifying, each win translates into 0.25 added to the coefficient over the next five years and each draw contributes 0.125.

First-Round Matches

Two of Scotland's clubs enter in this round, so there's a potential full point to be gained from four wins in home-and-away competition. On the other hand, Bulgaria and Kazakhstan have three teams enter in the first round, so there's a potential to lose more ground to teams from those countries. All in all, the extreme positive scenario vaults Scotland to 21st, and the extreme negative scenario drops them to 27th.

I was disappointed to see low-budget St. Johnstone grab 4th place in the league, given their lack of success in European competition in the past. I made my peace with it by realizing that this year there weren't four solid squads in the Scottish top flight. At least St. Johnstone has an excellent manager, even if they don't have a decent transfer budget.

St. Johnstone faces Lithuania's FK Trakai. Here is where I point out that Lithuania is currently rated the 50th-best league, so even though St. Johnstone finished 4th and FK Trakai finished 2nd, one would hope St. Johnstone would be more than a match for the Baltic squad.

I am going to try to use Club Elo to preview these match-ups. Trakai is ranked at an Elo of 1055. St. Johnstone is 1327. The 272-point difference between the clubs is roughly the distance between a club in the lower half of Italy's Serie A and St. Johnstone. At some point, the Club Elo web site will translate this into a win probability, but not yet.

So, anyway, I would put St. Johnstone as the heavy favorites to advance, despite their loss in the first round (on away goals) two years ago. I will say the expected result would be at least a win and a draw, netting 0.375 for the coefficient.

The other Scottish club in the first round is Rangers, one of the two biggest in Scotland. After emerging from bankruptcy, this is their first European tilt since losing in the Europa League play-off round six years ago.

Rangers face Progres Niederkor, the fourth-place Luxembourg side who have been outscored 40-1 lifetime in European competition. Rangers, who finished third, have been going on a bit of a transfer spending spree, so...this should be a cakewalk. Expect the full 0.5 towards the coefficient, and be embarrassed should even a draw sneak in there. Progres Niederkorn does not even merit a Club Elo ranking.

Second-Round Sneak Peak

The draw for the second round happened at the same time. So...

Should St. Johnstone defeat the Lithuanian runners-up, they face the winner of  Sweden's Norrkoping (Elo: 1468) and Kosovo's Prishtina (Elo: not calculated). I have to think that St. Johnstone's run ends here with a loss to Norrkoping.

If Rangers overcome the Luxembourgers, they face either Cyprus' AEL Limassol or Gibraltar's St. Joseph's. Neither has a Club Elo rating, but I have to think the fourth-placed Cypriots have the advantage in the first round. Rangers may get the better of them in the second.

Entering the Europa League second round of qualifying is Aberdeen. They dodged a bullet when their red-hot manager, Derek McInnes, chose to stay rather than ply his trade down in the English Championship with Sunderland. They have not been big players in the transfer market (unless you count losing players), but hopefully not entering until the second round will give them the time they need to make some key moves.

By the way, I had previously had Aberdeen entering in the first round, due to Scotland's #25 ranking from 2015/16. However, with Manchester United winning the Europa League, this opened up another slot higher up, and bumped a several teams to enter in later rounds. On the one hand, it feels like this makes the coefficient less important, since even dropping to #25 didn't matter. On the other hand -- good thing they didn't drop down to #27.

Aberdeen (Elo: 1413) faces the winner of Kazakhstan's Ordabasy (Elo: 1269) and Bosnia-Herzegovina's Siroki Brijeg (Elo: 1197). McInnes will chase his dream of the Europa League group stage at least into the third round.

In the less-exciting Champions League (am I the only one who feels this way?), Celtic (Elo: 1623) debut against Northern Ireland's Linfield (Elo: 1041) or San Marino's SP La Fiorita (Elo: 689). Look for Linfield to collect quite the windfall in gate receipts to have Celtic come across the Irish Sea and destroy them. Brendan Rodgers' boys got embarrassed in Gibraltar last year, but they are too experienced to suffer a repeat.

Saturday, June 03, 2017

Google Local: 300K Views, More Photo Diversity

I am now over 30% of my way to my goal of one million views for my photos on Google Maps. I figured out a few things.

I was paging through other people's photos on Google Maps, and I realized something, I was seeing the photos Google wanted to show me, not the ones I necessarily wanted to see. So how could this help me figure out why my interior photos seem to be doing so much better? Well, it hit me -- Google has exterior photos from its own mapping. I might argue that they're better off with a well-framed photo that shows exactly what you're looking for, but the algorithm doesn't reward that.

At the 100K and 200K thresholds, about 70% of my views came from photos with 10K or more views. That's down to 63% as I have only one new photo with 10K views. I think that's because I have a lot of recent, popular photos that have not yet crossed that point -- I have four photos between 8K and 10K. Still, let's look at my star performers.
#7 (was #6) Rubio's, San Diego, CA, 17,695  views, August 2016. (Was 17,375 last time.)
#6 (new) Dollar Tree, Chantilly, VA, 20,138 views, April 2017. This has garnered an astounding number of views in a short period of time. I have other pictures of Dollar Trees that look almost identical, and have been popular -- but not this popular.

#5 Wendy's, Chantilly, VA, 23,646 views, August 2016. (Was 22,665 last time.)
#4 (was #3) Jumping Jack Sports, Ashburn, VA, 26,142 views, January 2017. (Was 24,891 last time.)
#3 Office Depot, Bowie, MD, 31,281 views, December 2016. (Was 22,854 last time.) This one continues to get a lot of hits, unlike  #2, #4, #5 and #7. I can only assume pictures continue to get hits until more recent ones taken by other users bump them. Perhaps I should re-take ones if I revisit formerly popular locations?
#2 (was #1) Arby's, Beltsville, MD, 34,414 views, January 2017. (Was 32,053 last time.)
#1 Michaels, Bowie, MD, 42,040 views, December 2016.  (Was 30,952 last time.)

Every week Google e-mails me my top six photos -- the Dollar Tree one was the only one that ranked in May, so my road to 400K is in the new photos. It would be nice to take interesting pictures that people wanted to see -- I am thinking that next time I have a sit-down meal, taking a picture of the food could be worth trying.