Ranking European Football Clubs

I have been trying get out a ranking of clubs across various leagues in Europe (only top divisions of say top 5 nations to start with), and wanted to get your thoughts/suggestions on this.

My key critiques of UEFA ranking is that it does not consider domestic performances, and 5 years is way too long. Also, you cannot have the same wieghtage applied to performances this season and those 4-5 years ago. And finally, they do not give additional points for winning silverware. So I am attempting to come up with a system which does all that, with 3 different ways clubs can score points.

1. Points from European Competitions:
I plan to use UEFA Cofficients scoring system, now that they have corrected it to ensure you get more points for CL than EL. But with the following change:
- Only last 3 seasons points count. Current seasons points have a triple weightage, previous seasons have double weightage, and the season before only 1. For eg, if a club has scored 10 points this season, 5 last season, and 15 the season before that, then their points will be = 10*3 + 5*2 + 15*1 = 55 points.

2. Points for winning trophies or reaching key rounds:
CL = 20 points for winning it, 10 for final, 5 for semis
EL = 10 for winning it, 5 for final
Domestic league* = 15 for winning it, 10 for 2nd place, 5 for 3rd
Domestic Cup* = 10 for winning it, 5 for 2nd place (this is high coz there are no points for winning games in this competition, unlike the above 3 competitions)

*Once we start looking at lower leagues than top 5, we start reducing these points
**Not accounting for supercups, those are glorified friendlys anyways...

3. Points from Domestic League:
- this is based on points the club wins in the last 3 seasons, with the same 3-2-1 weightage. If the league has less teams than 20, that is factored into the points so that its comparable.
- This will be multiplied with the leagues competitiveness factor, which is determined by the average points earned by the leagues clubs in Europe. Doing some calculations, I got the following competitiveness factor for the top 5 leagues:
1. EPL = 1.98
2. Bundesliga = 1.93
3. La Liga = 1.90
4. Serie A = 1.76
5. Ligue 1 = 1.73
(which sounds about right, based on last 3 years performances)


Does this make sense? Would love your suggestions on:

1. League competitiveness factor. It will become really compliucated when we start looking at leagues outside of europe - how do we factor that??
2. I plan to multiply the european competition points by 2 so that top european teams earn as many points through europe as they earn via domestic league, just to keep the balance. By this method, for eg, ManUtd earned 270 points via europe, 45 points for silverware, and 291 points from EPL... does that weightage stack up?

Any suggestions / critiques are welcome.

I may be overcomplicating here, but I think with seeding/ranking performance in tournaments, you can be more exact than just giving x points for reaching QF and y for SF, etc. Tournaments themselves have to be absolute - 1 team wins, 1 team qualifies, but when you're analyzing you could take into account:

-the manner of the defeat - e.g. losing 4-0 over 2 legs, though it has same effect, is a worse performance than going out on away goals or penalties.
-not only which stage a team goes out, but the ultimate fate of the team they go out to, i.e. losing to eventual winner ranks better than losing to a team who themselves go out in next round

So for tournaments you could actually rank teams from 1-32 (or however many teams), and dole the points out relative to this and it might be a little more accurate.

debaser wrote:I may be overcomplicating here, but I think with seeding/ranking performance in tournaments, you can be more exact than just giving x points for reaching QF and y for SF, etc. Tournaments themselves have to be absolute - 1 team wins, 1 team qualifies, but when you're analyzing you could take into account:

-the manner of the defeat - e.g. losing 4-0 over 2 legs, though it has same effect, is a worse performance than going out on away goals or penalties.
-not only which stage a team goes out, but the ultimate fate of the team they go out to, i.e. losing to eventual winner ranks better than losing to a team who themselves go out in next round

So for tournaments you could actually rank teams from 1-32 (or however many teams), and dole the points out relative to this and it might be a little more accurate.

That is a great idea, so I just tested it over a few teams - it doesn't make as much difference as you would think it would on the points, made no difference to eventual ranking. Thats coz it does not impact the domestic league (since everyone plays everyone), and has minimal impact on CL/EL coz you have to give more points to a team which goes a round further, than a team which lost to the eventual champion or runners up...

While I agree with you that it is a slight improvement, its not worth the effort that it requires... (I actually plan to publish these rankings on a website, and update them every month)...

any more suggestion? Looks like I overestimated the geekiness of this MB!!

lol blue I guess I underestimated yours =D


Some toughts:

bluenine wrote:
1. Points from European Competitions:
I plan to use UEFA Cofficients scoring system, now that they have corrected it to ensure you get more points for CL than EL. But with the following change:
- Only last 3 seasons points count. Current seasons points have a triple weightage, previous seasons have double weightage, and the season before only 1. For eg, if a club has scored 10 points this season, 5 last season, and 15 the season before that, then their points will be = 10*3 + 5*2 + 15*1 = 55 points.

This means you're calculating yourself the club points, without adding 33% of league coefficient to their points totals, right?
Because otherwise that would be an unfair advantage for bigger leagues teams

Speaking of Uefa points system, it's basically 2 points Win, 1 point Draw, 0 points loss, plus 3 points for accessing the group stages and 1 point for every round advance, isn't it?
Assuming I got it right, I'd say that progressing further should be rewarded more.
For example, teams A & B meet in a tie, each wins one leg, team A goes thru then lose both legs in the following tie.
That means 3 points for team A and 2 points for team B added to their season CL total.
I'd say this reflects poorly on the fact that team A went one stage further, which is usually considered quite a relevant accomplishment.
Unless team A went thru with a penalty shootout (which means that in the end it perfomed evenly with team B), I think going thru should reward more points, probably even progressively increasing (say N points for 1st KO stage, 2N for quarter-finals and so on).

I'm also not sure that group stage matches should net as much points as KO stages matches.
A win with a hopeless 4th place team in the last group match is not exactly like a win in QF first leg.
Anyway this is probably just going too complex.

Another idea from the top of my mind is awarding a little bonus for away wins in the KO stages.

Finally, in a relatively short competition like CL, where you face just a small subset of the partecipating teams, I think there should be some weighting in of the opposition a team have to face.

I think you could calculate a group coefficient, similar to the leagues coefficient for adjusting league results, based on the previous 2 seasons points rankings of the partecipating teams.
This could go even further, for example factoring in the current teams ranking in the KO ties.

bluenine wrote:
2. Points for winning trophies or reaching key rounds:
CL = 20 points for winning it, 10 for final, 5 for semis
EL = 10 for winning it, 5 for final
Domestic league* = 15 for winning it, 10 for 2nd place, 5 for 3rd
Domestic Cup* = 10 for winning it, 5 for 2nd place (this is high coz there are no points for winning games in this competition, unlike the above 3 competitions)

I think in domestic leagues you should look at european competitions slots to give bonus.
Getting 4th in a league with 4 CL spots is an achievement, a target clubs aim and prepare to fight for.
Getting 4th in a league with 3 spots is not.

Similarly, getting 2nd or 3rd in the EPL is basically the same thing achievement wise, you just fullfil the target of automatically qualifing for CL without preliminary rounds, and you didn't win the league.
There's already the league points factor separating those teams in the ranking, so I don't think that being te first of the losers without other practical benefits should be rewarded much more.


An idea: shoudl you give little bonuses for repeat success and multiple success in a single season?
Winning back to back CL or performing a treble is very prestigious, so probably it should award a little more than the obvious outcome of the plain more points obtained.

bluenine wrote:
3. Points from Domestic League:
- this is based on points the club wins in the last 3 seasons, with the same 3-2-1 weightage. If the league has less teams than 20, that is factored into the points so that its comparable.
- This will be multiplied with the leagues competitiveness factor, which is determined by the average points earned by the leagues clubs in Europe. Doing some calculations, I got the following competitiveness factor for the top 5 leagues:
1. EPL = 1.98
2. Bundesliga = 1.93
3. La Liga = 1.90
4. Serie A = 1.76
5. Ligue 1 = 1.73
(which sounds about right, based on last 3 years performances)

Well I think there's not much to add to that.

bluenine wrote:
Does this make sense? Would love your suggestions on:

1. League competitiveness factor. It will become really compliucated when we start looking at leagues outside of europe - how do we factor that??

Well probably you just shouldn't factor that.
Different confederations are separated monads and there's no easy way to cross-evaluate them apart from a biased multiplier factor.
So IMHO you should just presents separated confederation tables, it's up to the reader to figure out how being first in Asia compare to being first in Europe.

One way, out of the top of my head, to compare leagues in different confederation could be looking at the players in every league, counting their recent caps for National Teams, and sum them up factoring in FIFA NT rankings.
This could give you a rough esteem of the talent value, and hence competitiveness, of a league, but I guess this should be excruciatingly time consuming to compute, and anyway it's still somewhat theoretical.

bluenine wrote:
2. I plan to multiply the european competition points by 2 so that top european teams earn as many points through europe as they earn via domestic league, just to keep the balance. By this method, for eg, ManUtd earned 270 points via europe, 45 points for silverware, and 291 points from EPL... does that weightage stack up?

I think it's fair at first sight.
As a rule of thumbs, IMHO winning one of the big 4 leagues should award more points on average than winning EL, and less than winning CL

Are you accounting for the various Super Cups (European/Domestic)? What about the Intercontinental Cup?