When the bonus is better than the main game - and vice versa

1) The essence of the choice

In slots, the lion's share of RTP often sits in bonuses, but the path to them is different in price and risk. We choose where EV/hour and risk for the bank are better for your goal (WR, "snapshot" of a skid, tournament, non-sticky cash phase), and not what "seems hot."

2) Mini-math solutions

Designations:

• ABP is the average bonus payout in xBet.
• BF - middle backs to bonus.
• RTP\_ base/RTP\_ buy - RTP of the base game/when buying a bonus.
• α - Ante/Double Chance markup to the rate (often + 25%).

Formulas:

1 paid spin bonus EV:

• $$
• EV_{b/spin} \approx \frac{ABP}{BF}
• $$
Ante effect (A/B measurement):
• $$
• \Delta EV_{b/spin} \approx ABP\!\left(frac{1}{BF_{Ante}}-frac{1}{BF_{base}}ight) - \alpha
• $$

If $\Delta EV\le 0 $ - Ante only accelerates flow.
Price per hour:
• $$
• \ text {Bet }imesext {spin/h }imes (1-RTP)
• $$
• Buying a bonus: it is more profitable than the base only if RTP\_ buy $\ge $ RTP\_ base, or value time compression (less "empty" spins), taking the increase in variance.

3) When the bonus is better than the base

1) RTP\_ buy $\ge $ RTP\_ base and median bonus not too low
- typical for slots with sticky Wild, increasing multipliers, repeated retriggers. Buying "cuts" empty time.

2) High ABP and acceptable BF
- if $ ABP/BF $ already gives a decent EV on the spin, and the slot in the database "starts up," it is reasonable to spin up to a bonus (or take Buy with a good RTP\_ buy).

3) Multiplier tournaments/stream content
- target: quickly catch × N. Buying speeds up attempts (within the bank and rules).

4) Non-sticky cache phase (rare, spot)
- if the purpose is "snapshot" and the bank allows, sometimes reasonably 1-2 purchases in high-vol title (Unit-Risk ≤1% bank/purchase), immediately with Profit Lock.

5) Persistent meters are almost full
- in games with portable fees, proximity to the warranty increases the local $1/BF $. Here it is logical to "squeeze" to the bonus.

4) When base is better than bonus

1) Wagering wager (WR)
- frequent events, low variance are needed. Base with HF ≥30 -35%, BF <120-150, cascades/ways, without Buy and more often without Ante.

2) RTP\_ buy $\le $ RTP\_ base or very low bonus median
- the purchase eats the bank and gives rare "peaks"; the base with the same RTP is cheaper by the hour.

3) Bank Restricted/Cap Risk
- Buy requires series (5-10) and Unit-Risk 0.5-1% bank/purchase. There is no such reserve - play the base.

4) Bonus T & Cs
- max bet/ban Buy/Ante. Any purchase under the bonus = forfeiture risk.

5) Goal - "flat grind "/missions/rakeback
- the base gives turnover stably and predictably.

5) Volatility and the bank: what's changing

Buying a bonus dramatically increases variance. We need a larger bank and seriality:
• Unit-Risk: 0.5-1.0% bank to buy.
• Batch: ≥ 60-100 × of purchase price.
• The base in "flat" slots requires 200-500 bank rates; in high-vol - 500-1000 +.

6) Clear selection algorithm (2 minutes)

1. Open "i": RTP\_ base and, if any, RTP\_ buy; Ante terms (does RTP change).
2. In the demo for 50-100 spins, rate HF, BF, ABP (estimated), modifiers, drawdown.
3. Count $ EV _ {b/spin} = ABP/BF $.
4. For Ante, do A/B (without/with Ante at 100 spins) and count $\Delta EV $.
5. Compare RTP\_ buy vs RTP\_ base and the median bonus (10-20 purchases in the demo).
6. Link to the goal: WR → base, snapshot/tournament → bonus (if RTP\_ buy ok), limited bank → base.
7. Set the limits: SL/SW/Trailing/timer; Unit-Risk for shopping.

7) Numerical examples

A) Ante can be pointless

ABP ≈ ×80, BF\_base = 160 → $EV_{b/spin}=0{,}5×$.
Included Ante + 25% steak, BF\_ Ante = 110 → additive:
• $$
• \Delta EV \approx 80\left(frac{1}{110}-frac{1}{160}ight)-0{,}25 \approx 80(0{,}0091-0{,}00625)-0{,}25 \approx 0{,}228-0{,}25=-0{,}022
• $$

→ worse, this is just an accelerator of pace.

B) Buying more profitable than base (RTP)

RTP\_ base = 95.5%, RTP\_ buy = 96.4%, purchase price = × 100, median ~× 38.
In the long series, Buy gives + 0.9% RTP and saves time to bonuses → more rationally if the bank pulls the series and the goal is a "snapshot."

C) WR scenario

RTP = 96%, target WR 25 × per base 100 cu → Cost\_ {WR} ≈ $0 {,} 04 × 25 × 100 = $100.
Purchases do not help close the WR cheaper, but make the variance higher than the → base.

8) Tactics and limits (by mode)

If playing base (flat grind/WR):
• Rate 0.8-1.2% bank/spin; rate 300-450 sp/h.
• Without Ante (until $\Delta EV $ is proven); without Buy.
• Limits: SL − 30%, SW + 40-60%, Trailing 50-70%, timer 30-60 min.

If you "crush" the bonus (purchases/hunting):
• Unit-Risk 0.5-1% bank/purchase; batches of 5-10; SL batches − 25-35%, SW + 60-100%.
• After a major hit: Profit Lock 60-80%, Trailing 50-60%, pause.
• Do not raise the purchase price "out of the red" (no dogons).

If you twist the base to the bonus:
• The permitted "ladder" is only from profit, ≤1 step and in the mouth guard (high-vol ≤1% bank/spin).
• If the actual BF deteriorates (50% worse than your log) - pause/shift.

9) Bonus or base checklist

RTP\_buy $\ge$ RTP\_base? Yes → bonus is acceptable; no → base.
Is the median bonus (for 10-20 demo purchases) acceptable?
ABP/BF in the database gives a decent $ EV _ {b/spin} $? Yes → you can "squeeze" the base.
Purpose: WR → base; tournament/snapshot → bonus.
Bank: enough for a series of purchases? No → base.
T & Cs: no ban/cap max bet/Buy/Ante?
Limits: SL/SW/Trailing/timer/Unit-Risk recorded?

10) Typical errors

Buy bonus at RTP\_ buy below the base "for the sake of speed."
Enable Ante without A/B measurement $\Delta EV $.
Play WR with purchases/Ante - grow variance without benefit.
Chasing "almost a bonus" is a UX deception, not an increase in probability.
Catch up purchase price/rate from drawdown.
Violate max bet/ban Buy/Ante under the bonus - the risk of canceling the win.

11) Short profiles (copy)

Profile "WR-even (base)"

RTP ≥96%; HF ≥30–35%; BF <120–150.
Bet 0.8-1.2% bank; 300-450 sp/h; SL −30%, SW +40–60%, Trailing 50–70%.

Profile "High-vol "snapshot" (bonus/purchases) "

RTP\_buy ≥ RTP\_base; median is not "bottom."
Unit-Risk 0,5–1%; series 5-10; PL 60–80%; Trailing 50–60%; without dogons.

Profile "Let's live up to the bonus base"

Good $ ABP/BF $, modifiers frequent.
Bet 0,5–1,2%; A/B Ante; one short press from profit maximum.

Conclusion

The bonus is "better" when RTP\_ buy is no worse than the base, the median is adequate, the goal is a quick "shot "/tournament, and the bank pulls the series. The base is "better" when stability is needed (WR/grind), RTP\_ buy is disadvantageous, the bank is limited or T & Cs cut Buy/Ante. The decision is made by numbers: RTP, $ ABP/BF $, $\Delta EV $ from Ante, "hour price," required bank and limits. Then you play not "on sensations," but according to a model where risk and time are subordinate to your goal.