Login

Username:

Password:

Join the community

Don't have a forum account yet? We need you to join the discussion! It only takes a couple of minutes to sign up.

Using Stats for Player Analysis

   
Hi guys

I am in the process of creating formulas to monitor player effectiveness in each position. I have a striker and a Winger formula but I need some help.

Does anyone know where I can find:

Tackles made

Passes
Pass Completion %

Or any other good stats. Im currently using ActimStats.

My findings so far.

Players get a score. Higher is better.

Premiership 2006/2007
Wingers (using Goals, Assists, Crosses, Shots, Penalties Scored and Missed, Offsides, Fouls, Yellow and Red Cards). It is then averaged into per 90 minutes.

Cristiano Ronaldo 1.116
Ryan Giggs 0.866
Arjen Robben 0.787
James Milner 0.765
Stuart Downing 0.734
Jermaine Pennant 0.679
Morten Gamst Pederson 0.667
Aaron Lennon 0.585

Some interesting Results eg. Lennon last so far.

What do you guys think.
   
Goalkeepers: (using clean sheets, saves, average goal conceded, goals, assists and discipline)

Petr Cech 0.867
Jose Manuel Reina 0.771
Tim Howard 0.675
Thomas Sorensen 0.620
Carlo Cudicini 0.574
David James 0.571
Marcus Hahnemann 0.570
Edwin Van der Sar 0.566
Jussi Jaaskelainen 0.481
Jens Lehmann 0.458
Mark Schwarzer 0.439
Paul Robinson 0.356
Scott Carson 0.256
Shay Given 0.140
Henrique Hilario -0.256

What do you think?
   
It's a decent enough idea. I recommend using Excel, or some other similar program with spreadsheet capability, to keep track of the stats and to make calculating scores easier. I'm curious, what's your formula?

One more point about formulas and scoring. I think each player should get a mark out of 100 (scored on a Bell curve) for each individual statistic, and their overall score should be an average of all of their marks.
   
Sounds like a good idea

I do a formula for each position.
positive contributions give the player points eg. Strikers get 1.5 points per non-penalty goal. Points get taken away for fouls, cards etc.

Each contribution is weighted per position and importance. eg. goals > fouls etc.

Afterwards the players total is divided by the number of minutes he has played and the multiplied by 90 to get a contribution per full game.

The only major flaws I have found are comparing between positions and with keepers their average goals conceded per game is tallied. Saves are also noted but keepers who play less time might concede a high average but don't have time to make saves. This is what harmed Shay Given. I am considering changing it so total goals conceded replaces average per game, If that makes sense.

Here are my current formulas:

Strikers

90{(Non –Penalty Goals x 1.5) + (Assists) + (Penalties Scored x 0.2) - (Penalties Missed x 0.8) + Shots/100 - (Offsides x 0.05) - (Fouls x 0.05) – (Yellow Cards x 0.1) – (Red Cards x 0.5)}
______________
Minutes Played

Wingers

90{Non-Penalty Goals + Assists + (Crosses/10) + Shots/100 + (Penalties Scored x 0.2) - (Penalties Missed x 0.8) - (Offsides x 0.05) - (Fouls x 0.05) – (Yellow Cards x 0.1) – (Red Cards x 0.5)}
______________
Minutes Played

Goalkeepers

90{(Clean Sheets) + (Saves/10) + Non-Penalty Goals + Assists + (Penalties Scored x 0.2) - (Penalties Missed x 0.8) – (Average Goals Against x 10) - (Fouls x 0.1) – (Yellow Cards x 0.1) – (Red Cards x 0.5)}
_______________
Minutes Played

I need to know passes and tackles etc. to do CMs and Fullbacks etc.
Image

:cry: RIP Vincenzo Paparelli. Killed by an AS Roma flare in 1979. :cry:
:cry: RIP Gabriele Sandri. Killed by a Police Officer in 2007. :cry:
FORZA LAZIO

   
Sorry about the double post.

Another problem is that you cannot measure mental attributes and leadership attributes, or physical eg. speed

So don't use it entirely to decide who is better than who. Use it as a guide.
Image

:cry: RIP Vincenzo Paparelli. Killed by an AS Roma flare in 1979. :cry:
:cry: RIP Gabriele Sandri. Killed by a Police Officer in 2007. :cry:
FORZA LAZIO

   
It just seems like a very arbitrary formula (though any of this kind would be to some degree) because you're taking it upon yourself to assign weights to different areas of the game while you're also deciding which areas of the game bear any importance at all. Also, with this system you can't compare the contribution of players in different positions. Here's a bit more about my suggestion - a rough outline basically, I'll use a fictional example (all these stats will be completely made up).

Let's say that the effectiveness of an attacker is based on only two things (limiting at two only for the purpose of this outline, in reality it should definitely be more). These two things are: goals per game, and shots on target/total shots.

The 3 top players in each statistic, let's assume they're the same 3 for both to keep things simple:

Goals per game:
1. D. Kuyt 0.5 (100)
2. A. Shevchenko 0.4 (80)
3. W. Rooney 0.1 (20)

Shots on target/total shots:
1. A. Shevchenko 0.8 (100)
2. D. Kuyt 0.4 (50)
3. W. Rooney 0.3 (37.5)

Using those stats, I've added the scores they would receive in brackets. The formula to get the scores for any statistic is: player stat/best stat * 100%.

Therefore the best strikers in the EPL are:
1. A. Shevchenko (90)
2. D. Kuyt (75)
3. W. Rooney (28.75)

So by using this system you can compare how well Shevchenko did in his position and in his league to how well someone like Dida did in his position and league.

What do you think, better or worse than your idea?
   
They both have strengths and weaknesses, although I think your idea could go places :D . The only problem I see is that if you add one person to It you may have to redo everything but that is not too bad. I still think we should weigh the significance. So eg. For a striker, using your method, goals are out of 100 and discipline out of, lets say 20.

We could even try doing both tests to get an average of best player. :D
   
Sorry about the Double Post

I guess there will always be debate in terms of how important something is so I guess we should find a happy medium. I am a bit biased towards short passing and possession but I'll try and be impartial.

I definitely think that we need to weigh thing whichever way we do it because some things are obviously more important than others.