This is such a complicated topic. I've run a bunch of experiments on it using the match engine outside the game, and I've also decompiled the match engine into its own functions and pieces. Below is my best attempt to summarize everything (yes, with the help of AI. I went back and forth with it to get the clearest summary I could, so its not slop, and the result is clearer than what I would have written alone). Let me know if anything is confusing and we can chat about it. Hopefully it's useful for anyone building training plans, or anyone who just wants to know how this works.
TL;DR:
Each player has eight tactical-area values (0-100), shown as the pies on the Tactical Familiarity screen. Only two really matter for results:
chance creation and
pressing intensity.
Out-of-possession shape matters about half as much; the other five barely matter. The headline numbers (overall effectiveness, buy-in, next match info) have almost no effect on results. "Per season" means 38 matches.
What to do
- Get chance creation and pressing intensity above 50 for every regular starter (a pie of 6+ segments). Below 41 instead of just above 50 costs about 4-5 points a season per area. Pushing on towards 9-10 segments adds only 1-2 more.
- Then out-of-possession shape. Leave the other five to match play.
- Check players, not the team pie. One starter at 30 loses while a teammate at 90 gains, and the team pie hides it. New signings and players in new positions are usually the low ones; training is the quickest fix for them.
- Bands, not points. A change matters only when it crosses a band (16, 34, 41, 51, 67, 81, 91). 70 to 78 changes nothing; 48 to 52 does.
- Simple tactics help everyone. Every 10 points of complexity costs 2.5 points in every area for every player.
- Don't tweak settings right before a match. Each notch away from what the team knows costs about 3 points in that area until they've played it.
Quick glossary
- Area value: a player's familiarity with one tactical area, 0-100. A pie has 10 segments of 10 points.
- Dice: every few seconds the engine rolls a random number for each mental and technical attribute. A player with 15 Anticipation plays, in each moment, like whatever he just rolled. Familiarity makes some rolls bigger or smaller.
- Band: values are grouped (below 16, 16-33, 34-40, 41-50, 51-66, 67-80, 81-90, 91-100). Within a band the effect is the same.
- Match exposure: how used the team is to the exact settings you play, built up by playing them.
- Study: one of my simulation experiments, running the match engine thousands of times with settings I control.
1. What the screen shows
"Checked" = my formula matches the game's own numbers; "code" = read from the decompiled engine.
| Panel item | What it is | How sure |
|---|
| Tactical Areas pies | Area values, 0-100. In Player View a player's pie is his average of all eight areas | Checked (PSV match, 11 players) |
| Player Tactical Ability | From his mental attributes (2.1) | Checked, exact |
| Tactical Complexity | Three complexity numbers (2.4) | Checked (order), code |
| Capability, Role Suitability | Position rating and role fit; only feeds overall | Code |
| Player Buy-In | Coaching, relationships, head-coach tendencies | Checked (order) |
| Next Match Info | Opposition report and player knowledge | Checked (order) |
| Overall Effectiveness | 0.5 × average of the eight areas + 0.25 × buy-in + 0.15 × next match + 0.1 × player ability | Checked, exact |
The pies are the
result, not the training input: they already include tactical ability, exposure, complexity and training.
2. What goes into an area value
In plain words: up to
25 points from the player's mental attributes, up to
30 from playing these settings, up to
25 for a simple tactic, and up to
60 from training. Anything above 100 is cut off.
area value = round(0.25 × tactical ability + 0.3 × exposure + 0.25 × (100 − complexity)) + round(0.6 × training), kept between 0 and 100 (checked exactly on 20,000 records)
2.1 Tactical ability. The middle value of Anticipation, Decisions, Composure, Concentration and Versatility, converted to 0-100: (middle − 1) × 100/19. A middle value of 14 gives 68, i.e. 17 points in every area.
2.2 Match exposure. The team remembers how familiar it is with every setting it has used in each area. Exposure = familiarity with the nearest known setting, minus 10 per notch away from it. Fully used to a setting (99) and you move it one notch: about 89, so the area loses about 3 points. In- and out-of-possession shape come from the two formations. In the PSV match, established starters had 99 but newer players 0-26, costing them up to 30 points in every area.
2.3 Training. A training number per area (12-68 in the PSV line-up). Sessions raise it; untrained areas fade. Every 10 training points add 6 to the area. How sessions turn into this number is outside the match engine and not decoded yet.
2.4 Complexity. 60% team part + 30% instructions-and-roles part (instructions switched on, plus the player's role and duty) + 10% formation part (0 in all my cases). The PSV tactic's team part was 78 (a Chelsea tactic I tested: 4). Complexity 78 adds only 5.5 points instead of up to 25: about 20 points lost in every area.
2.5 Two PSV starters, same area:
| Starter 1 | Starter 2 |
|---|
| Tactical ability | 68 → 17 | 53 → 13.25 |
| Exposure | 99 → 29.7 | 0 → 0 |
| Complexity | 70 → 7.5 | 77 → 5.75 |
| Base (rounded) | 54 | 19 |
| Training | 39 → 23 | 20 → 12 |
| Area value | 77 | 31 |
Both match the game's records. Most of the gap is exposure: Starter 2 was new to the tactic.
3. What the values do in a match
Each die is rolled between 0 and about the attribute + 2, re-rolled for the team about every 24 seconds of match time. Play reads the die, not the attribute. After each roll, area values stretch or shrink two dice:
| Area value | <16 | 16-33 | 34-40 | 41-50 | 51-66 | 67-80 | 81-90 | 91-100 |
|---|
| Die × | 0.50 | 0.75 | 0.85 | 1.00 | 1.15 | 1.30 | 1.45 | 1.60 |
The change is capped at 5 either way, roughly 2-3 attribute points on the 1-20 scale. At ×1.30 about half of all rolls already hit the cap, so going above about 70 adds little, while dropping below 41 starts taking points away.
| Area | Dice it scales |
|---|
| Chance creation | Anticipation, Off the Ball |
| Pressing intensity | Anticipation, Tackling |
| Out-of-possession shape | Concentration, Marking |
| Passing style | Decisions, Composure |
| Pressing trap | Decisions, Positioning |
| Build-up | Technique, Vision |
| In-possession shape | First Touch, team cohesion |
| Line of engagement | Work Rate, Aggression |
The headline numbers scale other dice (overall: a morale die on dribbles and placed shots; buy-in: Sportsmanship, Teamwork, Professionalism; next match: Pressure). In 80,000 simulated matches none had a reliable effect on results.
4. What each area is worth
In 80,000 simulated matches I set one area to a chosen value for
every player at once, with the other areas random, and measured points against a baseline of 45 (where dice are unchanged). Points per season, accurate to about ±1.
| Area value | 8 | 25 | 37 | 45 | 58 | 73 | 85 | 95 |
|---|
| Chance creation | −3.6 | −3.2 | −2.5 | 0 | +2.1 | +2.4 | +2.9 | +3.2 |
| Pressing intensity | −3.3 | −3.0 | −2.7 | 0 | +1.8 | +2.8 | +3.1 | +3.5 |
| Out-of-possession shape | −2.3 | −1.7 | −1.2 | 0 | +0.2 | +0.8 | +1.3 | +0.9 |
| Build-up | −0.5 | +0.1 | −0.3 | 0 | +0.6 | +1.2 | +1.0 | +1.2 |
| Passing style | −0.8 | −0.5 | −0.3 | 0 | +0.3 | +0.4 | +0.5 | +0.7 |
| In-possession shape | −0.6 | −0.2 | −0.2 | 0 | +0.3 | +0.3 | +0.7 | +0.5 |
| Pressing trap | −0.3 | −0.1 | −0.3 | 0 | +0.2 | +0.3 | +0.3 | +0.1 |
| Line of engagement | −1.1 | +0.2 | −0.2 | 0 | −0.4 | −0.6 | −0.8 | −0.5 |
Why this order? An area matters as much as the attributes behind its dice. Anticipation is the only one with a large effect on its own, and it sits under two areas. Concentration comes next; the rest are worth little.