Got bored so made a formula for selection (get your calculators out!)

[Mansleg]

So long story short I wanted to find out who in my squad was suited to the positions/roles I wanted to use in a new tactic I had adopted (after losing 5 in a row on my trusted 4-3-3 ). I started out by listing up to 7 attributes that I thought were crucial to the position and then ranked them according to importance. Next I went through each potential candidate in my squad and noted down in a neat little table each attribute number.

Then I thought "how am I going to analyse this data and find out who is the best (and the worst) player for this position?" So I came up with this forumla for scoring each attribute: (20 - attribute score) x 1/attribute rank
The LOWER the score the better and the LOWER the total score, the higher a player is ranked for that position.

So for example, I am looking for an AMC and decide that the attributes I'm looking for are "off the ball", "positioning", "pace", "acceleration", "passing", "creativity" and "dribbling". Then I need to choose and order of importance for these attributes so I rank them as such: 1) Passing; 2) Creativity; 3) Dribbling; 4) Pace; 5) Acceleration; 6) Off the ball; 7) Positioning.

Then I need to mark up my player's score for each attribute. So we have player X who has a value of 12 for passing. So his score for passing is (20 - 12) x 1/1 = 8. Simple enough. Next his creativity is valued at 8 so his score for creativity is as follows: (20 - 8) x 1/2 = 6. Dribbling is 11 so: (20 - 11) x 1/3 = 3. Pace is 13: (20 - 13) x 1/4 = 1.75. Acc at 11: (20 - 11) x 1/5 = 1.8. Off at 13: (20 - 13) x 1/6 = 1.17. Pos at 9: (20 - 9) x 1/7 = 1.57. Thus, he recieves a total score of 23.29.

Do this for a few more players and I'll have some total scores which I can rank and then give to each player, which means I can see who is my 1st choice AMC, 2nd choice and 3rd choice etc.

I made a spreadsheet of this formula meaning all I have to enter is the attributes for each player and I receive a ranking!



Now obviously I'm not saying this idea is the be all and end all of judging your players (for all I know someone may have already done the same thing). I understand that this does not take into account selection on match day (ie. fitness, morale etc.) nor any other match day factors (weather, opposition etc.) nor does it use any other important information you should consider when deciding on a 1st choice striker/midfielder etc. such as their footedness, personality, preferred moves etc. It is simply a way of ranking players based on YOUR choice of attributes and YOUR choice of ranking those attributes.

Thoughts? Opinions? Anyone else made up a formula like this?

[Howie]

I like this, I'm impressed. I've always wanted to apply some logic where I can because at the end of the day it's a stats based game, however as you say there are some many factors as you've listed in your last paragraph that are difficult to calculate, as well as the hidden attributes too.

It's a guide but then playing everyone in the squad based on their form over the previous 5 games is probably more reliable.

[Mansleg]

Also would anyone be interested in me uploading the spreadsheet?

@Howie: Yes form is obviously a factor but I only came onto this because my whole team was absolutely dire and I literally started from scratch tactically (whole new formation and roles/positions) and rather than use trial and error or a coach's advice to find out who I should play where I wanted to do it on my own terms. Thanks for the positive feedback!

[kingjericho]

Good idea. I think your selection of attributes should be more extensive, though. How could you decide between AMCs and don't add Technique and Flair?

[Cougar2010]

Also would anyone be interested in me uploading the spreadsheet?

@Howie: Yes form is obviously a factor but I only came onto this because my whole team was absolutely dire and I literally started from scratch tactically (whole new formation and roles/positions) and rather than use trial and error or a coach's advice to find out who I should play where I wanted to do it on my own terms. Thanks for the positive feedback!

Its always good to see someone put some thought into the game

I must admit it is something I've tried to do in the past myself but I seem to have settled now for creating filters within the game to show the relevant attributes and then judge from that rather than using a separate program.

Once you've settled into the game its perhaps not something you would use all the time but just every now & again when you feel things aren't going right or to confirm your own judgment.

If you upload it I would be interested in taking a look.


One small comment I would have is how does it cope with unbalanced players? say 6 good attributes then lacking a key skill. Does it tend to underestimate or overestimate compared to a balanced player.

[robzilla]

Been working on a similar thing over here: http://community.sigames.com/showthr...ion-Calculator

[Mansleg]

Its always good to see someone put some thought into the game

I must admit it is something I've tried to do in the past myself but I seem to have settled now for creating filters within the game to show the relevant attributes and then judge from that rather than using a separate program.

Once you've settled into the game its perhaps not something you would use all the time but just every now & again when you feel things aren't going right or to confirm your own judgment.

If you upload it I would be interested in taking a look.


One small comment I would have is how does it cope with unbalanced players? say 6 good attributes then lacking a key skill. Does it tend to underestimate or overestimate compared to a balanced player.

Ok I'll look into uploading it soon.

And with regards to overestimation and underestimation the more I think about it the more I get confused! I'm no mathematician. But I guess a player is penalised more for their Rank 1 attribute twice as much as their Rank 2 attribute and 3 times as much as their Rank 3 attribute etc. Whereas their Rank 2 attribute produces a penalty thats 1 and a half times greater than the penalty on the Rank 3 attribute. So I am guessing it would underestimate a balanced player. Let me do an example.

A player with 19 in the Rank 1 and 4 in the remaining 6 categories has a score of 30.49 and a player with 4 in Rank 1 attribute and 19 in the remaining 6 categories has a score of 36.60!!! So the player with a 19 in THE key attribute and a 4 in the sub-key attributes is considered better than the player who aces all the categories except the most important.

Hey it isn't a perfect system as I only knocked it up in 10 minutes but I'm hoping I wouldn't ever encounter such extreme players

EDIT:
@robzilla: Nice! Think I'll take a look at this.

[Mansleg]

Here is the spreadsheet: http://www.2shared.com/file/WuksG9Yx...ayer_rank.html