Top 300 European Clubs

For comparison, the Swedish blogger Spelare12.com publishes a Top 100 European clubs list in January every year. New list out now.

 

You either need way more than just one year of data, or you need to really retool all the constants you've put into your Elo rankings (since there is no "one true" Elo system) - there are far too many eye-raising rankings here to put much faith in the "top 300" outside of the top five or six. You don't have a single Spanish club in the top 10, nor a single Russian club in the top 30, and Kazakhstan's #1 above Italy's #1.

The big problem with a ranking like this - at least in the current WoSo climate - is that the UWCL simply doesn't provide enough games per season to be a reliable indicator of interleague strength. The UWCL only tells you things about the clubs that consistently qualify for the UWCL, and even then, applying an Elo system to just one season unduly rewards minnows that do well in the qualifying phase more than it rewards stronger clubs that get bad R32 draws.

If you get the chance, see about including results from as many preseason friendly tournaments as you can. Aside from including more seasons of the UWCL, that'll be the only way to get a somewhat-better picture of how the leagues actually compare to each other outside of the top 1 or 2 teams in each league.

 

Ah, well I think that if you're comparing teams within one country, using just recent results might be okay, but when you start trying to compare between leagues, just recent results definitely won't cut it. No matter what Elo system you use, there's essentially a latency period, and unless you play enough games more than that number, the comparison won't be good. So in a domestic league that plays 20+ games per year, great, but when you only get 3-9 UWCL games per year, that's definitely not enough to make a good comparison.

There's a lot to unpack here, but I think the lack of interleague play is really evident here, and maybe there's some cherry-picking going on. Since BIIK is the only Kazakh team to play in Europe and usually makes the R32, it make the rest of its league look much better than they probably are - that is, your comment about winning against inferior competition is true, but I think it's underrepresented in the numbers you have. You have Kazakh's #2 team between Florentina and Umea in the ranking - that's pretty hard to believe.

The fact that your choice of Elo system doesn't let winning teams lose points for nevertheless "underperforming" (or losing teams gain points for "overperforming") is a fair enough choice, but I think for the level of parity (or disparity) in WoSo right now, it's probably a better idea to allow for that level of nudging.

Also, just adding a second year of UWCL results would make the Barca/BIIK comparison much better. Yes, their lone 2 matches against each other this year were pretty evenly split, but look at what happens when you add last year's results: BIIK gains 1 win and 3 losses, including getting absolutely creamed by Lyon, while Barca gains 4 wins and 2 losses, the 2 losses to Lyon being by just 1 goal each. Two of Barca's wins were against Avaldsnes, which I think is also greatly underranked at the moment even if they went out at the R32 in the past two years - simply because they played Barca and Lyon! Teams don't change that much from one year to the next, so having a second (or even third) year, especially for UWCL, seems necessary to me.

Finally, Atletico getting creamed by the #2 overall team two years in a row shouldn't hurt them as much as I think you've let it. Touching back on the over/underperforming concept, I'd be interested to see what your GD adjustments are. Once you get much past a GD of 5, I don't think they should matter at all. Also, I would guess that their draw+win over last year's semifinalists Man City should help them more than it looks like it has...

Yes, but I'd still chalk it up to recency bias. And look at those last four opponents - three of the four were traditionally strong UWCL teams themselves.

Again, I think just looking at one year of UWCL results isn't enough to draw a good comparison. All three of the leagues you mention have a good history of making it past the R32 - so Italian teams losing to those others shouldn't hurt them enough to have the Italian leaders fall beneath BIIK, which has only made it past the R32 once.

Actually I like the look of those K values! You could nuance the system a bit more if you wanted - e.g. have CL knockout games at 60, CL qualifying round games at 40, domestic games for the top leagues at 40, and domestic games for weaker leagues at 30 - but the numbers you have sound pretty good to me already. Aside from the temporal issue preventing sufficient interleague data, I think the biggest thing I worry about with your system is how you handle GD.

 

I am not an expert in ELO rankings at all. I just chime in to say that, as an observer of Norway's Toppserien (actually, more of an avid reader of the Norway thread, from our excellent @Bauser), I can tell you that if it's true that Avaldsnes has been unlucky with WCL draws in the last two seasons, it's also true that in 2018 the club actually spiraled down in flames in the local league, compared to the previous season. If in 2017 they were fighting with LSK-Kvinner for the title (it wasn't much close, but it was at least a fight), in 2018 they had abysmal results and at times were even sucked into the relegation brawl at the bottom of the rankings, while LSK-Kvinner was winning the title undisturbed with an almost perfect record, OL-like.

Just to say that Women's football, at the moment, keeps being so erratic that, indeed, it can sometimes happen that "teams change a lot from one year to the next".

 

Avaldsnes are a club with an international profile. They get rid of 8-9 players every year and replace them with equally many, mostly untested non-European foreigners. A big number of Brazilians, Australians and Americans have been there over the years - and done well. For some reason, they also have a new head coach every year. It is hard to build a backbone with that policy. Most times it has worked anyway, but not last year. They came up short in the transfer window for once. Now they are rebuilding again for 2019. I don't know what to expect from them in 2019. Could be Top 3 or lower half of the table.

 

Well, there's the FIFA women's rankings style, which I know you were kinda iffy on, which assigns the "actual result" with a GD-based lookup table instead of a straight 0/0.5/1 split. The other possibility I could think of would be to take the "expected result" and use it (maybe with another logistic function) to assign an "expected GD" so that you can include a new multiplier of (actual GD - expected GD).

 

Memories of Cecilie Pedersen, who I've always really enjoyed. One of the true enigmas of Norwegian fotball.

And Tom Nordlie. And Arne Utvik, the Jean-Michel Aulas of Norge.

When Nathalie Utvik retires, does Avaldsnes retire back to anonymity?

 

Very injury-prone, but still part of the squad. Just extended her contract for another year.

I seriously doubt it. She was mostly a rotation player. With last season as the yardstick, Avaldsnes are back in anonymity right now. Time to fight their way back into the limelight again. The best thing they did in 2018 was winning the qualifying group in the Champions League and enter the main competition while struggling around the bottom in Toppserien.

 

Yeah, FIFA uses a lookup table instead of a direct formula to see how GD determines "actual result", so that's probably the issue.

 

... And Arne Utvik, the Jean-Michel Aulas of Norge.

When Nathalie Utvik retires, does Avaldsnes retire back to anonymity?

Arne Utvik invested a princely sum around the time Nathalie joined the team. And while he didn't pull the plug when she took a brief hiatus and went to study and play in Oslo, it wouldn't be unnatural to wonder how strong the connection is when she hangs up her boots.