Re: A Statistical Comparison of Tevez, Dzeko, and Aguero
UPDATE - Last season's stats and overall stats for the last two seasons (100 games)
Good news for those who think Aguero/Tevez make the best strike partnership; I've finally found something you can use!
So, having done the same statistical analysis of every competitive game last season (54 in total), I have data from last year to compare to this year, and I've also been able to combine the two seasons to get averages that are even more statistically sound.
Unfortunately, it's not really possible to compare the strike partnerships last season because of Tevez' little golfing holiday. He played so few minutes with Aguero or Dzeko, that the statistics are unreliable (however, I did roll those minutes into this year's data and will show those further down).
Here are the individual numbers from last season;
Total playing time (games @ 94 min/game);
Aguero 38.3 (incl. 3.5 as sub), Tevez 7.8 (1.6), Dzeko 25.9 (3.4), Balotelli 20.3 (2.8)
Goals per game (parenthesis shows goals as a starter and as a sub);
Aguero 0.78 (0.72/1.41), Tevez 0.51 (0.64/0.00), Dzeko 0.69 (0.58/1.47), Balotelli 0.84 (0.80/1.08)
Note: the data are better when they start games than when they come on as a sub, because there's a much bigger sample size, and even Tevez' starting data are a bit unreliable because he played so few games.
Team goals per game as a starter (for/against);
Aguero 2.22/0.75, Tevez 2.41/0.48, Dzeko 2.17/1.20, Balotelli 1.88/0.86
Putting this in terms of equivalent goals for the whole season (54 games);
Team goals for the season as a starter (for/against);
Aguero 100/34, Tevez 109/22, Dzeko 98/54, Balotelli 85/39
Note: the Tevez' data are not nearly as relevant as the Aguero, Dzeko, and Balotelli data due to the very small number of games he played.
Here's the data for the strike partnerships last season;
Total playing time (games);
Tevez/Dzeko 1.0, Aguero/Dzeko 16.3, Tevez/Aguero 2.6
Only the Aguero/Dzeko data are statistically meaningful for last season, but the data for the other pairings will be added to this season's data later on. The team goals for/against are as follows;
Team goals per game (for/against);
Aguero/Dzeko 2.45/0.92 (this year it's 2.21/1.24)
Overall Data (100 competitive matches, last season and this season)
Total playing time (games @ 94 min/game);
Aguero 64.3 (incl. 6.2 as sub), Tevez 40.3 (3.8), Dzeko 51.1 (8.5)
Goals per game (parenthesis shows goals as a starter and as a sub);
Aguero 0.72 (0.67/1.13), Tevez 0.50 (0.52/0.26), Dzeko 0.63 (0.49/1.30)
Even after combining the data to cover 100 games, the sub data is still a very small sample size for Tevez, and not that much better for the others, so I'll keep focusing on the starting data.
Team goals per game as a starter (for/against);
Aguero 2.03/0.88, Tevez 2.00/0.88, Dzeko 2.04/1.17
Putting this in terms of equivalent goals for both seasons (100 games);
Team goals for two seasons as a starter (for/against);
Aguero 203/88, Tevez 200/88, Dzeko 204/117
Strike Parterships;
Total playing time (games);
Tevez/Dzeko 16.2, Aguero/Dzeko 27.6, Tevez/Aguero 16.2
The team goals for/against are as follows;
Team goals per game (for/against);
Tevez/Dzeko 2.54/1.05, Aguero/Dzeko 2.35/1.05, Tevez/Aguero 2.29/0.68
Putting this in terms of equivalent goals for two full seasons (100 games);
Team goals for two seasons (for/against);
Tevez/Dzeko 254/105, Aguero/Dzeko 235/105, Tevez/Aguero 229/68
This shows that the team scores more goals when Tevez and Dzeko are on the field together, than both non-Dzeko combinations. The Dzeko partnerships also concede a lot more. The goal differential is as follows (for 100 games);
Team goal differential for two seasons;
Tevez/Dzeko 148, Aguero/Dzeko 130, Tevez/Aguero 161
So what does this tell us over two full seasons?
1) Aguero has outscored the others as a starter (Dzeko and Tevez are about the same).
2) The team has scored almost exactly the same number of goals per game, regardless of which striker starts, though they've conceded more when Dzeko starts.
3) We've scored the most goals when Tevez/Dzeko are on the pitch, but we've conceded the least (by quite a bit) when Aguero/Tevez play
4) If we judge the best partnership by the biggest differential between team goals scored and conceded, then over the last 2 seasons, the Aguero/Tevez partnership comes out slightly on top.
5) All that running around and "putting in a shift" that Aguero and Tevez do doesn't seem to help us score any more goals (we actually score less), but it does appear to prevent the other team from scoring. When we play Dzeko, we're likely to score more, but also concede more.