On with the statistical analysis. I looked at all my games within the last 3 years and tried to see what I could work out from them. We are talking over 1000 games but I stripped out any friendlies or games where either side was playing any formation other than 352, 451, 442 and 541. These are probably the only formations that are regarded as being competitive in todays environment.
Additionally I only looked at home games. Away games tend to score less and concede more goals so a difference in playing styles in away games would bias the data. note also these figures are based on my teams so half the games I will be home, half away, and also reflect the various home advantages.
Offside Traps
One thing I checked for was offside traps. How much difference does it make, what about with different referees.
TPs v OST |
NO |
YES |
S>=7 |
S=6 |
S=5 |
|
|
|
|
|
|
Chances |
275 |
119 |
59 |
22 |
30 |
Goals |
82 |
35 |
16 |
7 |
8 |
Conv |
30% |
29% |
27% |
32% |
27% |
|
|
|
|
|
|
Games |
62 |
32 |
19 |
6 |
6 |
Cha/Game |
4.4 |
3.7 |
3.1 |
3.7 |
5.0 |
|
|
|
|
|
|
Goals/Game |
1.32 |
1.09 |
0.84 |
1.17 |
1.33 |
The above chart shows how effective through plays are in different situations and how many games are in the sample. For this data I only used games where the defence and midfield of the 2 teams was broadly similar strengths.
Playing TPs when the opponent doesn't use offside traps ends up with an average of 1.32 goals a game. Looking at the games as a whole where the opponent played OST then there are only 1.09 goals/game.
This is not the whole story though. and I have broken down these chances according to the skill of the referee
I often play OST on S6 and S5 refs as a margin call. S6 refs look ok, but with a small sample of data. With S5 refs it really is negligable.
I would like to stress the sample sizes here are tiny, especially on S6/5 refs.
Using a playmaker
I also looked at the benefits of using a playmaker. Same sample data as above except I ONLY checked games where the away side pid a win bonus so I could filter out inactive teams.
When the home side didn't use a PM they scored an average of 0.8 goals/game. Using a playmaker increased this to 1.34 - a significant jump.
I then what happened depending on opponent marking options.
If the away side manages to mark the playmaker they reduce the goals massively to a goal a game. Marking anyone other than the PM helps a little weighing in at 1.3 goals/game conceded. Marking nobody at all is slightly weaker at 1.4 goals/game.
The moral of this story - always pick a playmaker, you probably still do better if he's marked than if you don't have one.
PM Marking |
None |
YES |
Mark PM |
Mark Any1 |
None |
|
|
|
|
|
|
Chances |
257 |
1258 |
215 |
1129 |
171 |
Goals |
66 |
390 |
73 |
339 |
53 |
Conv |
26% |
31% |
34% |
30% |
31% |
|
|
|
|
|
|
Games |
84 |
292 |
75 |
267 |
39 |
Cha/Game |
3.1 |
4.3 |
2.9 |
4.2 |
4.4 |
|
|
|
|
|
|
Goals/Game |
0.79 |
1.34 |
0.97 |
1.27 |
1.36 |
An interesting point is when I ran the data with matches where the home side was significantly stronger. Here's the same chart as above:
PM Marking |
None |
YES |
Mark PM |
Mark Any1 |
None |
|
|
|
|
|
|
Chances |
347 |
1088 |
128 |
1068 |
239 |
Goals |
115 |
403 |
43 |
390 |
86 |
Conv |
33% |
37% |
34% |
37% |
36% |
|
|
|
|
|
|
Games |
70 |
195 |
35 |
193 |
39 |
Cha/Game |
5.0 |
5.6 |
3.7 |
5.5 |
6.1 |
|
|
|
|
|
|
Goals/Game |
1.64 |
2.07 |
1.23 |
2.02 |
2.21 |
Now the home side gets more chances and scores more goals. Having a PM adds less benefit now than if the sides are close. and if you manage to mark the PM then you gain a MASSIVE benefit. It seems if you are facing significantly weaker opposition then using a PM is potentially more of a risk than it is worth.