APBRmetrics

Eurobasket 2011 is underway. Some NBA players are involved, even some superstars. Spain is the heavy favorite.
Because I also have ratings for euroleaugue and spanish ACB I thought it might be interesting to create some kind of rough power ranking based on the available player ratings. If there is no ranking for a specific player, because he did not play NBA/Euroleague/ACB, I take that as a bad sign. Those 3 leagues are the strongest in the world. Not having played in any of them probably means you're not that good.
Obviously, having a rating of, say, '-2' in the NBA is a lot better than having a '-2' in ACB. One would have to account for that.

Belgium:
MBenga (NBA, -2.6)
Beghin (EL, -0.2, ACB, +0.8)
Van Den Spiegel (EL, 0.0)

Bosnia:
Teletovic (EL, +0.2, ACB, +0.2)
Gordic (EL, -1.2)

Bulgaria:
Videnov (EL, +0.7)
Ivanov (ACB, -1.3)

Croatia:
Bogdanovic (EL, -3.0)
Tomic (EL, +1.6, ACB, +1.4)
Markota (EL, -0.9)
Popovich (EL, -0.1, ACB, -0.6)

Macedonia:
McCALEBB (EL, +0.9)
Ilievski (EL, -0.4)

Finland:
Salin (EL, -0.3)

France:
Parker (NBA, +0.3)
Batum (NBA, -1.3)
Noah (NBA, +1.1)
Gelabele (NBA, -0.2 (2008))
Diaw (NBA, -1.0)
Seraphin (NBA, -4.7)
Traore (EL, -0.2)
De Colo (EL, +0.5, ACB, +0.1)
F. Pietrus (EL, -0.2, ACB, +0.7)

Georgia:
Pachulia (NBA, -1.0)
Tskitishvili (ACB, -1.3)

Germany:
Nowitzki (NBA, +8.2)
Kaman (NBA, -1.8)
Jagla (EL, -0.8)

GB:
Deng (NBA, +5.2)
Clark (ACB, +0.2)
Freeland (EL, +0.0, ACB, +1.0)

Greece:
Bourousis (EL, +1.1)
Zisis (EL, -0.6)
Bramos (ACB, -1.6)
Calathes (EL, +1.5)
Fotsis (EL, +1.5)
Koufos (NBA, -3.2)
Kaimakoglou (EL, +0.1)
Papanikolaou (EL, -0.2)
Vasileiadis (ACB, +1.4)
Sloukas (EL, -0.1)
Mavroeidis (EL, +0.8)

That's it for now, I'll add more teams later. Greece is stacked with good Euroleague players, while France has a bunch of solid NBA players. Germany, of course, has the current NBA finals MVP.
From these teams I'd probably rank Greece over France (because they're playing by FIBA rules, not NBA), then Germany, then Croatia.

There's a very good chance that I missed a couple of players, so if you spot any errors, feel free to mention them. Later in the tournament I might do SRS or ridge regressed team ratings.

Player Ratings used
http://stats-for-the-nba.appspot.com/ranking_rec
http://stats-for-the-nba.appspot.com/euroleague-ranking
http://stats-for-the-nba.appspot.com/acb-ranking

Israel:
Eliyahu (EL, +1.2)
Halperin (EL, -0.7)
Blu (EL, +3.3)
Pnini (EL, -0.2)

Italy:
Gallinari (NBA, +0.8)
Belinelli (NBA, -1.3)
Bargnani (NBA, -3.3)
Mancinelli (EL, -0.8)

Latvia:
Blums (ACB, +0.9)
Freimanis (ACB, -0.1)

Of the teams listed so far, the only +2 or better players (in the different strength leagues) are 2 from the NBA and now one American born, European player.


On your full league RAPM lists there were only 9 players above +2 in the entire ACB and only 11 in the Euroleague for the time periods studied. Meanwhile they are over 50 above +2 in the NBA this past season. The NBA may have 2-3 times as many players as either of these leagues but it has 5 times the level of +2 or above RAPM players. I guess people could interpret that differently.

Some might say that there are few real star performers in Europe, even in a league of weaker talent, while the NBA has a bigger crop of them even with tougher overall competition. That a number of players who can't make the NBA go on to be among the RAPM leaders in the ACB and Euroleague probably says something about the overall quality of the leagues. 1/3 of the RAPM top 20 in the ACB are Americans who couldn't make or stick in the NBA while about 1/2 of the the RAPM top 20 in the Euroleague were as well despite the limits on American participation.

Some might think the greater number of strong RAPM scores in the NBA is due more to the star system than more productive talent. And maybe that the weak performing players in the NBA have worse overall impacts than the weak ones in Europe, maybe because of better systems and stricter coaching in Europe that get better buy-in from bench players than in the NBA. Even if you agree with that, it doesn’t necessarily make them worse players than the Europeans if the groups were put in the same context though as suggested above.

I don’t think the Lambda choices are a big issue but I am not sure. 3000 was used for the NBA and the Euroleague. Not sure if the different length seasons make any notable difference.

The main reason for less players above +2 in Euroleague/ACB is that there's less data for those leagues. Individual impact could still be smaller in Europe if they had the same amount of games (possessions, really) because of rules/court size/zone defense but "not as much data" is the main reason for that effect.

I'd have to give the nod for France over Greece...

Ukraine
Lishchu (EL, +0.1, ACB, +1.3)
Fesenko (NBA, +0.4)
Pecherov (EL, +0.1)

Russia:
Kirilenko (NBA, +1.5)
Fridzon (EL, -0.4)
Shved (EL, -0.3)
Monya (EL, +0.4)
Vorontsevich (EL, -0.5)
Mozgov (NBA, -3.6)
Bykov (EL, +0.1)
Khryapa (EL, +1.2)
Ponkrashov (EL, -1.2)

Poland:
Kelati (EL, +0.3)

Portugal:
-

Montenegro:
Dasic (EL, -1.5)
Dragicevic (ACB, +0.2)
Pekovic (NBA, -3.9)
Cook (EL, +0.4, ACB, +1.4)

Serbia:
Krstic (NBA, -1.2)
Savanovic (EL, 0.0)
Teodosic (EL, +0.1)
Keselj (EL, +0.4)
Macvan (EL, -0.2)
Tepic (EL, +0.4)
Perovic (EL, +0.3)
Rasic (EL, -0.2)

Slovenia:
Lorbek (EL, +2.4, ACB, +4.0)
G Dragic (NBA, -2.6)
Begic (EL, +0.5, ACB, -0.9)
Lakovic (EL, +1.5, ACB, -0.7)
Slokar (EL, -0.7, ACB, +0.7)
S Udrih (El, -0.4)
Smodis (EL, -1.2)
Ozbolt (EL, 0.0)
Jagodnik (EL, -0.2)
Muric (EL, +0.4)

Lithuania:
Home court advantage (HCA seems to be higher in FIBA than in the NBA. I suspect it's even higher in a tournament compared to a standard game, so +~8)
Kaukenas (EL, +2.1)
Kalnietis (EL, +0.1)
Songaila (NBA, +0.7)
Jasaitis (EL, -0.8)
Jasikevicius (EL, -0.5)
Pocius (EL, -0.3)
Delininkaitis (EL, +0.6)
Lavrinovic (EL, +2.4)
Jankunas (EL, +1.4)
Petravicius (EL, +0.4)
Valanciunas (EL, +0.2)
Javtokas (EL, +1.4)

Turkye:
Onan (EL, -0.1)
Preldzic (EL, +0.6)
Kanter (NCAA, N/A)
Ilyasova (NBA, +1.6)
Turkoglu (NBA, +2.5)
Asik (NBA, +1.6)
Savas (EL, -0.6)
Akyol (EL, 0.0)
Arslan (EL, -0.1)
Tunceri (EL, -0.5)

Spain:
P Gasol (NBA, +2.1)
M Gasol (NBA, +1.9)
Navarro (EL, +1.6, ACB, +4.1)
Fernandez (NBA, -0.6)
Reyes (EL, -0.3)
Ibaka (NBA, -1.3)
Rubio (EL, +2.5, ACB, +2.5)
Calderon (NBA, -0.7)
San Emeterio (EL, +0.8, ACB, +2.3)
Llull (EL, +1.4, ACB, +1.7)
Claver (EL, -0.3, ACB, +0.5)
Sada (EL, +0.5, +0.7)

From that it's relatively easy to pick a top 8: Greece, France, Spain, Turkey, Lithuania, Serbia, Slovenia and Russia. Only problem is that those 8 cannot meet in the quarterfinals because 5 of them are in group A+B, and only 3 in group
C+D.

Group A:
Lithuania Spain Turkey Poland Portugal GB
Group B:
France Serbia Germany Italy Latvia Israel
Group C:
Greece Croatia Bosnia Montenegro Macedonia Finland
Group D:
Russia Slovenia Georgia Bulgaria Ukraine Belgium

in the quarterfinals #1 of (A+B) plays against #4 of (C+D), #2 of (A+B) plays against #3 of (C+D) and so on.

Taking already played games into account my ranking would be (group in brackets):
Spain(A)
Lithuania(A)
Turkey(A)
Greece(C)
France(B)
Russia(D)
Slovenia(D)
Serbia(B)
..
Ukraine(D)

which would lead to

A+B
Spain
Lithuania
Turkey
France

C+D
Greece
Russia
Slovenia
Ukraine/Croatia/Montenegro

Spain will crush Ukraine or any other team at #4 from (C+D). Lithuania over Slovenia, Turkey over Russia, Greece over France
which should lead to: Spain over Turkey, Lithuania over Greece
Spain over Lithuania

10 guys over +2 on RAPM in their respective home leagues for 24 teams and few than one-third of the teams have any. Lithuania has 2, Spain 3, no one else more than 1.

J.E. wrote: Greece over France

haha, we'll see

Good stuff, this sadly is all we really have to watch for a long time.

JE, what do you think is a decent translation of RAPM from Euroleague to the NBA?

Bobbofitos wrote:

J.E. wrote: Greece over France

haha, we'll see

Good stuff, this sadly is all we really have to watch for a long time.

JE, what do you think is a decent translation of RAPM from Euroleague to the NBA?

Yesterday Greece lost to Macedonia by 14, which probably means your prediction was better than mine :'(.

When comparing Euroleague ratings to NBA ratings we need to keep in mind that the Euroleague ratings might be off quite a bit because of limited data. Top Euroleague players probably don't get the ratings they deserve

Player Euroleague-Rating NBA-Rating
Garbajosa +2.3 -> -1.9
Kleiza +1.8 > -2.8
Navarro +1.6 -> -4.2
Childress +1.2 -> -3.9
Jaric +1 -> -0.1
Splitter +0.4 -> -3.4
Asik +0.4 -> +1.2
Jacobsen +0.1 > -0.5
Nachbar 0 > -2.7
Pekovic -0.4 > -3.5
B. Brown -1.1 > -3.2
Jawai -1.6 > -0.5

Average: -2.1

I guess that roughly means someone with a rating >+2 in Euroleague should be a solid NBA player/starter. Players >0 in Euroleague should be an OK 6/7/8th man. Unless that player's skillset is a particularly bad fit for NBA rules and gameplay, which might have been the case with Navarro. Athleticism is probably a big factor.

Also of note: I don't think we have pace data for Eurobasket but Turkey is currently ranked 1st in defense with Ilyasova and Asik being the minute leaders for that team. Both have good RAPM defensive numbers in the NBA as well. Asik is also #11 in DREB und #2 in blocks so far. Ironically Bargnani is #1 in blocks

If "top Euroleague players probably don't get the ratings they deserve"- in the Euroleague- then the interleague performance rating difference may be bigger than 2 points for them.

The gap for players below the top in the Euroleague might be even bigger. In theory the inter-league performance difference might be expected to have one average value but there could be some threshold of player quality or NBA suitability below which the bottom falls out more rapidly than the average if the context shifted from the Euroleague level and range to the NBA environment.

I am thinking of point blowouts in non-conference college mismatches after a certain quality gap is exceeded. And about the difference in what can happen when a player and another player just a step or two worse than the first player (when they are compared directly in a lower level of competition) join a good pickup game. There is some level of casual player who can fit decently in a good game without too much negative impact. Below that level the negative impact may grow quickly. The actual performance difference can be more than expected by the difference between the 2 lower level players measured against each other and players in their range.


Only 1 in 8 of the listed European League players who moved from Europe to the NBA without playing in the US first had a positive NBA RAPM estimate (Asik). The proportion of positive estimates in all NBA is probably only around 1/3rd (it may vary from year to year or between single season and multi-season but 1/3rd is what I remember finding before).

Most of the euro-imports probably played mostly against reserves and had low minutes initially or long-term. Navarro played pretty big minutes right away and had a fair amount of starts in his one and only NBA season that scored a very low NBA RAPM rating. The difference in performance in that more intense than average transition context was bigger than the average inter-league difference.


If the interleague performance rating difference is 2 points that implies the average NBA team (NY, Phoenix) would beat the average Euroleague team (Valencia, Zagiris) by 10 points per game. Using the data here http://en.wikipedia.org/wiki/Euroleague ... _NBA_games
the average margin was 13.7 (or an average of 2.74 per position) and the NBA has an overall record of 36–7. Now the teams might not be exactly average for their leagues and there is a home court edge to the NBA and friendly games might not be the best judge of true point differential but it is in the same ballpark as the suggested RAPM league differential but somewhat larger (compatible with my earlier comments than the gap could be larger). Without likely garbage time drag on differential it could be even larger still. The 3 games of league champion vs league champion also had an average margin of almost 15 points.

Ok, we're onto Round 2. Anyone who is NBA starved (as I am), ESPN360 has done a great job w/ the actual feeds as well as the announcers. (They've been very, very proficient, and you don't get any of the JVG "I'm the center of attention" interfering with the game) Replays are available for another ~week or so, I highly recommend Spain v Turkey (as well as a bunch of other games, but particularly that one, as most of the players are NBAers).

Here are the groups:

1. France 2/0 4
2. Serbia 1/1 3
3. Spain 1/1 3
4. Turkey 1/1 3
5. Lithuania 1/1 3
6. Germany 0/2 2

1. F.Y.R. of Macedonia 2/0 4
2. Russia 2/0 4
3. Slovenia 1/1 3
4. Greece 1/1 3
5. Georgia 0/2 2
6. Finland 0/2 2

Results from your first group carry over (the 2 teams that you've played) which is why France has a leg up (as they beat Serbia AND Germany, Finland lost to Macedonia and Greece, etc.). It is my understanding that 8 of the remaining 12 teams make it to the next phase, where there is a convoluted tournament progression from that point onward.

Group 1 (it's technically group E) was the much tougher group, as all the remaining teams are technically strong. Germany is a dog to progress given they're in the 0-2 hole and none of 4 teams above them look like shoe-ins to slip up. France, as the ~2nd strongest team (in my estimation behind Spain) and with a game up on everyone seem like a sure thing. Lithuania also benefits from the massive HCA, as JE has pointed out. My guess is the group finishes France-Spain-Lithuania-Turkey, with Serbia and Germany slipping up. If you have free time, most of these games promise to be exciting.

Group 2 (Group F) is the softer group. Russia and Greece are good but I'd list the remaining 4 teams behind all of Group E. Anyway, with 2 teams starting 0-2, (and neither being that good) it's pretty easy to just say the group finishes in this particular order. Although I think Greece are stronger than Macedonia, they actually lost to them by 14 in round 1. I wouldn't be that surprised if either/or/both of Greece/Slovenia bypass them. That said I'm pegging the group to finish Russia-Macedonia-Greece-Slovenia.

SRS and ridge regressed team ratings Not perfect, because nobody from A+B has played anyone from C+D yet.
Macedonia (FYROM) really surprised me, and many other people I guess. McCalebb is kicking ass. He was born in New Orleans.
Group stage concludes on Monday.
Taking ridge regressed team ratings the forecast would be

Group E:
Spain
France
Lithuania
Turkey

Group F:
Russia
Macedonia
Greece
Finland

Spain over Finland
France over Greece *COUGH*
Lithuania over Macedonia
Russia over Turkey

Spain over Lithuania
Russia over France (difference in strength is <0.1 for both games though)

Spain over Russia


Kirilenko, Asik, Parker, Batum and others seem to be in very good form

Newest power rankings (using ridge regression with a lambda of 5)
France's rating was hurt by resting 2 of their starters in the last game.

J.E. wrote:Newest power rankings (using ridge regression with a lambda of 5)

Code: Select all

Spain	9.96
Lithuania	7.48
Russia	7.2
Greece	4.26
Turkey	4.18
France	4.02
Slovenia	2.52
Germany	2.28
FYROM	1.92
Finland	1.08
Georgia	1.02
Montenegro	1
Ukraine	0.28
Croatia	-0.98
Serbia	-1.24
Poland	-1.5
Bosnia	-2.16
Italy	-2.4
Latvia	-2.66
GreatBritain	-3.96
Bulgaria	-4.04
Bosnia-Herzegov.	-4.98
Belgium	-7.12
Israel	-7.32
Portugal	-8.82

France's rating was hurt by resting 2 of their starters in the last game.

This is blind to the fact all games have been played IN Lithuania, right? What if you added a basic home court modifier, how much does that affect Lithuania's power ranking? (I suppose we could argue it's a flat X points, so... simply take that away!)

It seems bizarre to slate Turkey ahead of France. How has Macedonia done so well??? (They beat Russia!) We're onto the quarter finals though, with these teams left:

Spain Slovenia
14.09.2011 18:00

F.Y.R. of Macedonia Lithuania
14.09.2011 21:00

France Greece
15.09.2011 18:00

Russia Serbia
15.09.2011 21:00

So, I was wrong about Serbia falling in favor of Turkey, and Germany at least made it interesting as far as perhaps qualifying. Macedonia beating Russia was a mild upset, but I still don't have high hopes for them.

Spain should crush Slovenia; I have to imagine Macedonia's good luck story ends, mainly because they're in Lithuania; Russia got the easiest pairing and should continue to advance. The game everyone is poised for, I expect, should be France v Greece. (We should have a friendly bet, JE; my money is on France advancing =) )

Correction, Macedonia lost. A dude I know just reminded me by showing me this:



Some reason I thought they won. Oh well!