Theo Walcott vs. Gareth Bale

There is some discussions going on whether Theo Walcott or Gareth Bale is the better player. @benjaminpugsley put an interesting blog post online where he neatly compares various statistics of both players. His conclusion is
Bale exhibits dominance over Walcott in every category.
So, according to a bottom-up view, Bale is clearly the better player. That said, as always bottom-up numbers are difficult to interpret. No complete model of football exists and therefore it is not obvious for each statistic if actually a high or a low number is good. Sometimes the answer is "it depends". In Benjamin's analysis, for example, a high ratio of a player's goals over the team's total goals is seen as positive as one interpretation is that he is a main contributor to the team's success. However, an alternative explanation is that the player cannibalizes goals that team mates could score, too. He maybe only seeks many shots and thus takes opportunities from the team mates. A third interpretation is that the passes he receives are so good that any player could score. So while he does a good job in scoring, he could be easily replaced with a team mate without the number of scored goals changing dramatically.

I have no idea which of the three explanations is the correct one. It could even be a mixture of them or some more other reasoning can be found. This is a general problem of the bottom-up approach. It delivers valuable insight and we can learn a lot, but it is hard to base a player comparison on it because we are sure to miss something.

So let's see what Goalimpact says to those two players. Goalimpact simply measures the player’s team’s goal difference when the player is on the field. It doesn’t look at how he effects his team’s goal difference. The value is corrected for other influences on the team's goal difference to increase accuracy
  • the players alongside him
  • the strength of the opponent
  • whether he has home advantage
  • red cards shown to either side
  • luck
The value is transformed into a human readable scale where 100 is the average over all players and the best player in the world has 190. In terms of Goalimpact, Walcott is superior to Bale by a rather big margin. Walcott has a Goalimpact of 141.5 and Bale only of 117.8. Both are well above the average of 100 and would be valuable contributors to every Premier League team. If we compare the development of their Goalimpacts and those of their team mates over time, we get the following chart.

As we can see, both are above their team averages. This indicates that the teams, on average, play less successful without them than they do when these players are on the field. However, the distance to the team average is much larger in Walcott's case. Bale's team seems to compensate his absence rather well as the distance is only 3.9 points. They play only marginally worse without him. For some time in 2011 the performance with and without him was even equal. In stark contrast, the Goalimpact of Walcott is 13.6 points higher than that of his team mates. If he is not playing, the team will have a hard time to compensate it and, on average, they are not managing it to full extend. The average goal difference without Walcott worse than with him.

Conclusion

Top-down sees Walcott as better than Bale. The disadvantage of bottom-up is that we can't really tell which player is superior and the disadvantage of top-down is, we have no clue why Walcott is better than Bale. Maybe he is working more for defense. Maybe Bale is cannibalizing goals from his team mates. Maybe something completely different.

FC Arsenal vs. FC Bayern München: Preview

Tuesday will be the first leg of the round of sixteen matches between FC Arsenal and FC Bayern München. The first match will be in London.

As I couldn't find any predictions for the line-ups yet, I post here the line-ups leading to the highest total score according to the GoalImpact as shown in brackets. I'll refine the line-ups if I get more information. The GoalImpact measures the impact a player has on the expected goal difference of its team. The average over all 100,000 players in the database is 100. To get a feeling of the scale, I suggest you check the league averages.

FC Bayern München


                                                                   Neuer (154,0)

               Lahm (180,2)           Dante (125,2)            Van Buyten (136,2)     Alaba (117,5)

                                         Schweinsteiger (177,6)           Martinez (102,8)

                                    Müller (162,4)             Kroos (143,3)         Ribery (146,6)

                                                                   Mandzukic (124,1)

Substitutes: Robben (156,3), Contento (119,5), Gustavo (139,5), Gomez (147,7)
See here for the remaining players.


FC Arsenal
                                                                 

                    Podolski (116,2)                   Cazorla (125,4)                       Walcott (142,3)

                                     Ramsey (117,7)   Wilshere (122,7)   Arteta (122,2)

                    Vermaelen (133,7)    Mertesacker (146,1)   Koscielny (133,6)     Sagna (139,7)

                                                                 Szczesny (119,2)


Subs: Squillaci (134,6), Jenkinson (107,6), Coquelin (108,0), Rosicky (130,6), Oxlade-Chamberlain (121,5), Gervinho (120,3), Arshavin (135,8), Diaby (123,7), Giroud (110,5)

Injured: Gibbs (125,3)

No clearance for Champigons League: Monreal (109,7)

Summary

In general, the team of Bayern München seems to be superior. However, Arsenal has a strong right wing that will create pressure on Bayern's weakest point, Contento. Bayern, on the other hand, has a world class right wing with Müller and Lahm that seems to be unstoppable for the inner defense player Vermaelen. So my prognosis is that the wings will be key and that Bayern has an advantage here.

Update: Changed lineup to expectation as published on kicker.de

Update: Exchanged Giroud (110,5) for Ramsey Tod reflect actual Lineup Art beginning ob game.