Skip to content

Training Formula V2.0

Source: Vladar [1996140], thread t=16182535. Karma 15,082. Collaborative research with Kivou, Pyrit, Proxima (Journal of Torn Sciences).


The Formula

dS = (S × ROUND(1 + 0.07 × ROUND(LN(1 + H/250), 4), 4) + 8 × H^1.05 + (1-(H/99999)^2) × A + B + RAND(C)) × (1/200000) × G × E × (1+PERK%) × ...

Where: - dS = stat gain per train - S = your current stat (capped at 50,000,000 in formula) - H = starting Happy - G = gym dots (e.g. 7.3 for George's) - E = energy per train - PERK% = each training perk as decimal (5% = 0.05, multiply not add)


Stat-Specific Values (Key!)

Stat B (offset) C (randomness) A
Speed +2,000 ±1,350 1,600
Dexterity +1,500 ±1,000 1,800
Strength +1,700 ±700 1,600
Defense -600 ±1,500 2,100

Defense is objectively the worst stat to train. It has a -600 base offset (vs Speed +2,000). Speed gives the best raw gains.


Base Formula at Low Happy

When Happy = 0: dS = S / 200,000

So at 200K stats, 0 happy: gain = 1 per energy. At 50M stats (cap), 0 happy: gain = 250 per energy (before gym dots and perks).


Happy Loss Per Train

dH = ROUND((1/10) × Energy × random(4 to 6), 0)

Roughly 40-60% of energy spent is lost in Happy after each train. So training 10E costs you 4-6 Happy.

This is why timing boosts right after a 15-minute reset matters.


Perks Stack Multiplicatively

All perks multiply, not add:

+2% property × +10% faction × +1% education × +30% book
= 1.02 × 1.10 × 1.01 × 1.30 = 1.468 = +46.8%

Not 2+10+1+30 = 43%.


Key Practical Takeaways

  1. Speed trains give best gains. Defense is worst. Prioritise speed when building your stats.
  2. Training gains are randomised. Defense has the highest variance (±1,500), making it unpredictable.
  3. The Happy effect is via power function at low stats: The 8 × H^1.05 term completely dominates when stats are low, meaning Happy is exponentially important early.
  4. Gym dots multiply everything: A 7.3 gym gives 1.83× the gains of a 4.0 gym per energy spent.
  5. Training once for all energy = same as multiple trains (Chedburn confirmed, small rounding aside).

  • Training Gains Formula (Vladar)
  • Jumping vs Energy Training (Omanpx)
  • Training Strategy Guide (Anomalousman)
  • War Education Guide (Apo)