OUR EXPERIMENT Simulation Reports Published · 10 min read · 2,303 words

Simulation Report: 10,000 Players Buy Bonuses Until the Money Runs Out

A 300-unit budget, 100-unit buys, up to thirty rounds. We counted who was still able to buy at the end and what was left.

In a 10,000-player simulation, 88.2% ran out of cash before they could afford another 100-unit bonus, leaving the median buyer with just 69 units of the original 300. That means almost nine out of ten players walk away with less than a quarter of their bankroll.

The report breaks down why the loss is inevitable: 79.6% of individual buys fall short of their cost despite a 96.5% theoretical return, and the average buyer stakes 902 units in 9.0 buys, which at a 3.5% house edge translates to an expected loss of 31.6 units. You’ll see the math behind each round, the contrast with low-stake play, and how the odds stack against any hopeful streak.

WHAT WE FOUND
  • 88.2% of simulated players could not afford another 100-unit bonus, with the median ending at 69 units.
  • 79.6% of individual bonus buys return less than their 100-unit price even though the game’s RTP is 96.5%.
  • The average buyer wagers 902 units over 9.0 buys, incurring an expected loss of 31.6 units at a 3.5% house edge.
  • Players spinning 1 unit from a 300-unit bankroll average 1,702 spins, yet 73.3% go broke within 3,000 spins.
  • In 100 bonus buys the chance of seeing at least one 1,000x hit is 64.9% and of a 2,500x hit is 21.7%.

The setup

The simulation gave each of the 10,000 simulated players a 300-unit bankroll and let them purchase a 100-unit bonus whenever they could afford it. A purchase costs 100 units up front, and the model runs until the player can no longer afford another buy, which automatically limits the number of rounds each player can take. All runs used the same game, the same RTP and the same 3.5% house edge that applies to every spin inside the bonus.

In this bonus-buy model 79.6% of individual bonus buys returned less than their 100-unit price, even though the long-run return of the buy is 96.5%. The 3.5% house edge translates into an expected loss of 31.6 units for the average bonus buyer, who puts 902 units of stake through the game in 9.0 buys. Because a single 100-unit buy puts the equivalent of 100 spins at risk, ten buys carry the same expected loss as 1,000 spins.

When the budget was fixed at 300 units, 88.2% of the simulated players could not afford another buy by the end of their run. The median final balance among those who stopped was 69 units, while the top 10% of buyers ended with 436 units or more. The average player therefore spent almost three times their starting bankroll on bonus rounds before running out of money.

For comparison, the same 300-unit budget applied to a spin-only strategy produced 1,702 spins on average per player. In that scenario 73.3% of players went broke within 3,000 spins, and the luckiest 10% finished with 853 units or more. Both ways used the identical game mechanics, RTP and house edge; the only difference was whether the bankroll was allocated to 100-unit bonus buys or to 1-unit spins.

Use the 300-unit starting bankroll and the 100-unit purchase price as the baseline parameters if you want to reproduce these results or test a different strategy.

How the buyers finished

Out of the 10,000 virtual players, 88.2% could not afford another 100-unit purchase once the session ended.

The middle of the distribution settled at a modest 69 units. That median figure tells you that half of all buyers finished the night with less than a quarter of their original 300-unit bankroll. It is a concrete illustration of how quickly the high-cost buys eat into a modest budget.

Only the lucky tail escaped the worst outcome. The top decile of players walked away with at least 436 units, a gain that more than offsets the average loss but is achieved by a tiny fraction of the field. Those ten percent are the statistical outliers that marketing stories love to showcase.

On average each participant pushed 902 units through the game, spread over 9.0 separate bonus purchases. In real terms, a player who starts with 300 units will, on average, see their bankroll shrink by roughly a tenth of its size after just nine buys.

Use the figure of nine buys as a practical ceiling; any additional purchase pushes the expected deficit beyond the budget’s safety margin.

The control group that just spun

The control group entered the same 300-unit bankroll but never touched the bonus-buy feature. Every round they placed a single-unit wager, letting the game run until the balance could no longer cover the next spin. By keeping the stake at the minimum, they exposed themselves to the pure volatility of the base game rather than the amplified risk of a 100-unit purchase.

Because each spin only drains one unit, the average player in this cohort survived many more rounds than the bonus-buyers. The simulation showed that the typical spin-only participant logged well over 1,000 spins before the budget dried up. That longevity comes at the cost of a steady, predictable bleed: the house edge of 3.5% chips away at the bankroll on every spin, turning a 300-unit start into an inevitable decline.

The bust rate reflects the relentless nature of the edge. A clear majority of the spin-only players were unable to keep playing past the 3,000-spin threshold. Those who survived that far tended to be the outliers, riding an unusually lucky streak that kept their balance afloat while the rest fell to zero. The distribution was heavily skewed: while most ended with nothing, the top decile still managed to finish with a sizable surplus, far above the median outcome.

Contrast this with the bonus-buy group, whose median balance collapsed to under 100 units after just a handful of purchases. The spin-only strategy spreads risk across many low-stakes events, so even though the overall loss expectation is the same - 3.5% of the total amount wagered - the bankroll lasts considerably longer. In practice, a player who sticks to one-unit spins can expect to see the same percentage loss as ten bonus buys, but over 1,000 individual outcomes rather than ten amplified ones.

If you decide to play with a 300-unit budget, remember that each spin costs exactly one unit of stake. Track that unit carefully; once the balance reaches zero, no further play is possible.

Round by round: mostly below the price

Each bonus-buy round is a single gamble, and the bulk of those rounds fall short of the purchase price. As noted earlier, the majority of buys lose, so the session’s cash drain is driven by tiny, frequent deficits rather than occasional windfalls.

The first tier of “clip-worthy” rounds pays 100x or more. The model assigns a 20.268% chance per buy, which translates to an average wait of one in five buys. In a block of 100 buys the chance of seeing at least one 100x hit is 100.0%. When such a hit occurs the expected outlay to land it at a $1,000 buy price is $4,934.

The next tier jumps to 500x or more. Here the chance per buy drops to 3.719%, giving an average wait of one in 27 buys. The probability of seeing a 500x hit within 100 buys is 97.7%. A hit at this level costs, on average, $26,887 in purchases at $1,000 each.

The high-volatility segment of the model produces a 1,000x or more payout once in 96 buys. The per-buy chance is 1.041%, so the average wait is one in 96 buys. The chance of encountering at least one 1,000x hit in a 100-buy session is 64.9%. A streamer buying at $1,000 per bonus would on average spend $96,085 for every 1,000x hit shown in a highlight.

At the extreme end, a 2,500x or more payout appears once in 410 buys. The per-buy chance is 0.244%, yielding an average wait of one in 410 buys. The chance of seeing at least one 2,500x hit in 100 buys is 21.7%. The expected spend to land such a highlight at $1,000 per buy is $409,626.

Short series are luck, not information

When you line up just a few bonus buys, the result is dictated by pure variance rather than any emerging pattern. Our model ran 20,000 simulated players through exactly ten bonus purchases on a game that truly returns 96.5%. In that batch, 31.7% of the series finished with a net profit. The same volume of simulations for 50 purchases produced a profit in 36.1% of the runs.

The spread of outcomes makes the luck factor obvious. For the ten-buy tests, nine out of ten simulations produced an observed return between 29% and 255%; the median observed return sat at 71%. In the 50-buy batch, nine out of ten showed returns between 52% and 172%, with a median of 88%. The unlucky five percent of ten-buy series never exceeded a 29% return, while the lucky five percent reached at least 255%. With 50 buys, the bottom five percent capped at 52% and the top five percent broke through 172%.

Those extreme tails shrink as the series lengthens. A test of 200 bonus buys ended in profit in 38.4% of 20,000 simulations, and the nine-out-of-ten range narrowed to 70%-132% with a median of 94%. Extending to 1,000 buys, profit occurred in 31.6% of 4,000 simulations, while the observed return stayed between 83% and 111% for nine out of ten runs and the median settled at 96%. The reduction in spread shows that the volatility of short bursts is not a signal of skill or information; it is statistical noise.

If you decide to chase a short streak, treat any win as an outlier. A ten-buy run gives you only a 31.7% chance of ending ahead, and a 50-buy run improves that chance to just 36.1%. Therefore, when budgeting a bonus-buy session, calculate expected loss based on the long-run 96.5% return and do not let a lucky short run dictate the size of your next buy.

Smaller buys, longer life

When the same dollar amount eats a smaller slice of the bankroll, the session stretches dramatically. A streamer making $2,000 buys from a $100,000 balance risks 2% of the balance per buy; a viewer copying the excitement with $20 buys from $200 risks 10% per buy, five times the relative stake. That five-fold difference in exposure translates into a stark contrast in how long a bankroll survives under repeated bonus buying.

In the 20,000-player simulation where each buy cost $20 against a $200 bankroll, the pressure was immediate. A viewer with a $200 bankroll making $20 bonus buys was unable to afford another buy within 25 buys in 38.7% of 20,000 simulated cases and within 100 buys in 76.8%; the median viewer lasted 33 buys. The model capped at 100 buys, so most players ran out well before reaching the theoretical maximum.

Switching to a $4 buy size, which is only 2% of the same $200 bankroll, the picture changes dramatically.

The contrast is stark: at 10% per buy, more than three-quarters of players run out before hitting 100 buys, while at 2% per buy, less than one percent do. The median lifespan jumps from 33 to 100 buys, a threefold increase in opportunities to encounter a high-paying bonus round. Stretching the bankroll therefore raises the odds of catching a game-changing win.

If you want your bankroll to survive a full hundred-buy session, keep each purchase at or below 2% of the total balance - for a $200 bankroll that means $4 buys, not $20.

What the report means for a real evening

When you walk to a slot with a 300-unit bankroll and buy the 100-unit bonus, the math is unforgiving. In our bonus-buy model 79.6% of individual bonus buys returned less than their 100-unit price, even though the long-run return of the buy is 96.5%.

If your pocket is smaller, the share of the bankroll you risk per buy matters dramatically. By contrast, a viewer with the same $200 bankroll making $4 bonus buys (2% of the bankroll per buy) was unable to afford another buy within 25 buys in 0.0% of 20,000 simulated cases and within 100 buys in 0.7%; the median viewer lasted 100 buys. The same game, same RTP, same starting budget therefore shows a stark gap between a 10% stake and a 2% stake.

A 2% stake keeps the early bust rate at zero and pushes the failure rate past the 100-buy mark to under one percent. In plain terms, if you want to stay in the game long enough to see any meaningful variance, keep each bonus buy at two percent or less of your total bankroll.

A practical rule of thumb follows directly from those numbers.

Apply the same scaling to larger balances. By matching the 2% rule you keep the odds of an early bust below one percent and give yourself a realistic chance to ride out the variance that the game’s 96.5% RTP promises.

THE VERDICT

Bonus buying on a 300-unit bankroll is a losing proposition; 88.2% of simulated players end broke and the median ends with just 69 units. Stop buying the 100-unit bonus and stick to low-stake play if you want to keep any of your money.

The data behind this piece

Every figure in the text comes from the runs below. They are computer simulations and exact calculations, not real-money play, and they can be reproduced: the model parameters are stated and the random seed is fixed.

Buying the bonus vs spinning into it on the same budget

How we ran it: 10,000 simulated players start with 300 units. Group A buys the bonus round for 100 units per buy (up to 30 buys, stopping when the balance cannot cover another buy). Group B spins at 1 unit (up to 3,000 spins). Both use the high-volatility model calibrated to 96.5% RTP: any-win frequency 25.4%, bonus round on average once per 200 spins, 36% of the return comes from the bonus round, top payout 4,949x the stake; the bonus-buy return is set to the same 96.5%. Random seed: 2027

Same game, same RTP, same starting budget
Way of playingCould not continueMedian final balanceTop 10% final balanceAverage stake put through the game
Bonus buys at 100 units88.2%69436902
Spins at 1 unit73.3%broke8531,702

How misleading short test sessions are

How we ran it: Repeated test sessions of 10, 50, 200 and 1,000 bonus buys (and 200-spin demo sessions) on the high-volatility model calibrated to 96.5% RTP: any-win frequency 25.4%, bonus round on average once per 200 spins, 36% of the return comes from the bonus round, top payout 4,949x the stake. For each session length we record the observed return and whether the session ended in profit. Random seed: 2027

Observed return of a 96.5% game by test length
Bonus buys in the testTests that ended in profitUnlucky 5% saw at mostMedian observed returnLucky 5% saw at least
1031.7%29%71%255%
5036.1%52%88%172%
20038.4%70%94%132%
1,00031.6%83%96%111%

How many dead bonus buys sit behind one viral clip

How we ran it: 4,000,000 simulated bonus buys at 100 times the stake on the high-volatility model calibrated to 96.5% RTP: any-win frequency 25.4%, bonus round on average once per 200 spins, 36% of the return comes from the bonus round, top payout 4,949x the stake. We count how often a single buy pays at least 100, 500, 1,000 and 2,500 times the stake, and price the buys needed to see one at $1,000 per buy. Random seed: 2027

Frequency of clip-worthy bonus rounds
Bonus paysChance per buyAverage waitChance to see one in 100 buysMoney spent per hit at $1,000 a buy
100x or more20.268%1 in 5100.0%$4,934
500x or more3.719%1 in 2797.7%$26,887
1,000x or more1.041%1 in 9664.9%$96,085
2,500x or more0.244%1 in 41021.7%$409,626

Copying streamer bonus-buy sizing on a normal bankroll

How we ran it: 20,000 simulated viewers make up to 100 consecutive bonus buys on the high-volatility model calibrated to 96.5% RTP: any-win frequency 25.4%, bonus round on average once per 200 spins, 36% of the return comes from the bonus round, top payout 4,949x the stake, stopping when the balance cannot cover the next buy. Two sizing rules are compared. Random seed: 2027

How long a $200 bankroll lasts
Sizing ruleMedian buys before running dry (max 100)Out of money within 25 buysOut of money within 100 buys
$200 bankroll, $20 buys (10% per buy)3338.7%76.8%
$200 bankroll, $4 buys (2% per buy)1000.0%0.7%

Questions players ask

Can I make a profit by buying bonuses with a small bankroll?

No. 79.6% of bonus buys lose money even though the long-run return is 96.5%. With a 300-unit bankroll the average buyer loses about 31.6 units over 9 buys, and 88.2% run out of cash before they can afford another purchase.

Is playing the base game at 1 unit any better?

Yes, but the risk remains high. With a 300-unit bankroll players made an average of 1,702 spins; 73.3% went broke within 3,000 spins, while the top 10% finished with at least 853 units, showing far more variability than the bonus-buy approach.

What are the chances of hitting a huge multiplier like 1,000x?

In a session of 100 bonus buys the probability of seeing at least one 1,000x hit is 64.9%, and the chance of a 2,500x hit is 21.7%. Those rare events drive the hype, not the expected outcome.

Check it yourself:Run the simulators yourself