Egraam
As loaded with PIs as it gets so not really "basic" anymore, but still some good old-fashioned 4-4-2

Best you can probably do is go to the player search tab and there you can see players that are natural in chosen positions, but that means user would need to export files separately for each position
owing said: It should definitely be possible to create a script that calculates a staff members level based on the attributes. The code itself shouldn't be much dissimilar to the one already existing, as you would mainly have to change the algorithm that calculates and represents the level of the staff rating you want to show. This tool is an existing tool that lets you calculate the star rating of a specific coach, or multiple coaches using an exported html-file, much like in the script.

Here are the coach formulas if anyone was interesting in writing their own tool (I have my own bit of code but don't want to bother with making any fancy stuff)

def_tac = 6 * coach['Def'] + 3 * coach['TCo'] + 2 * ddm
def_tech = 6 * coach['Def'] + 3 * coach['Tec'] + 2 * ddm
att_tac = 6 * coach['Att'] + 3 * coach['TCo'] + 2 * ddm
att_tech = 6 * coach['Att'] + 3 * coach['Tec'] + 2 * ddm
pos_tac = 6 * coach['Men'] + 3 * coach['TCo'] + 2 * ddm
pos_tech = 6 * coach['Men'] + 3 * coach['Tec'] + 2 * ddm
fitness = 9 * coach['Fit'] + 2 * ddm
gk_shot = 9 * coach['GkS'] + 2 * ddm
gk_handling = 6 * coach['GkH'] + 3 * coach['GkD'] + 2 * ddm

Where ddm equals to Discipline + Motivation + Determination, Tco is Tactical attribute, Tec Techical and so on (these are the column names in game)

You will then get rating on scale from 0-300 which corresponds to training star values:
270-300 = 5*
240-269 = 4.5*
210-239 = 4*
180-209 = 3.5*
150-179 = 3*
120-149 = 2.5*
90-119 = 2*
60-89 = 1.5*
30-59 = 1*
15-29 = 0.5*
Well-balanced formation, wasn't explored much yet

Mostly TI changes compared to the previous version but I've also been quite happy about play of Inverted Wing-Backs with attack duty
Here is what I've been using for now to get through the beta, some decent results so far.

The whole point of extensive tactic testing is to find the best tactic possible. People here are trying to "break" the engine and the game, and set pieces are part of that. None of the top tactics play realistically, because that is not their point, they are supposed to win, utilising every tool possible. Besides, after a while it becomes obvious what type of set pieces are the best and almost every top tactic uses them, so they aren't skewing anything.

If you are looking for something realistic than I would advise you to look for tactical recreations or build something yourself and not go for the top tactics, because realism and winning, for most part, don't go hand in hand.
Valencia predicted 9th. Set Pieces from Bombyte. Getting good results against top teams especially.





I've found 3 defender setup with the central one on cover + higher line and drop off more to get pretty good results. 3 DMs should guarantee defensive safety, but the tactic still scores plenty.
TIs are mostly from the current top tactic





Hi,
I've been playing FM more casually for the past 2 years due to severe burnout from the game, but have recently started testing and creating some tactics again.
It's a 4-1-4-1, should provide good defensive stability while still being effective when it comes to scoring.
Set pieces are from the current top tactic (41212 Re-Cookie VI).





Waterfall V2
Voila V1
Voila V3.5 AF
Motion
Hacker V2
Threedom 343
Threedom V2
Demo Tapes V2
Uploaded V2, don't know if it's really much better than V1 but the Poacher goalscoring numbers in an elite team are so absurd (62 league goals), I thought someone might enjoy it. Underdog performance is ok (Watford), but nothing super impressive (GD close to 0, or negative in most of the tests I've done)

I'll be working on 3 at the back as well in the future, I even had some decent results with wide-center backs, I just need to explore the game a bit more first and do more testing before I get something out. But it will be coming :thup: