How to Solve Profit Instability in Game Machines That Fluctuates Every Week
An arcade owner in Jakarta showed me his weekly revenue data for a single fish table machine over six months. Week 1: 3,200 dollars. Week 2: 4,100 dollars. Week 3: 2,700 dollars. Week 4: 4,800 dollars. Then back down to 3,100 dollars in Week 5. The six-month average was 3,700 dollars per week. But the actual weekly results swung between 2,400 and 5,100 dollars — a range of over 100 percent from lowest to highest. When I asked what he thought was causing the swings, he said customers. Some weeks more people came and played. But his door counter data showed consistent weekly traffic, varying by less than 5 percent. The customer count was stable. The revenue was not. Weekly profit instability has causes that have nothing to do with how many people walk through the door.
Separate Revenue Instability From Player Traffic Fluctuation
The first step is to determine whether the instability is caused by changing player volume or changing per-player economics. Calculate two metrics for each week: total player sessions (how many distinct play sessions occurred) and average revenue per session (total machine revenue divided by total sessions). Plot both on the same time axis.
If player sessions fluctuate in the same pattern as total revenue, the instability is being driven by traffic. You need to understand why some weeks attract more players. The answer is usually external: holidays, local events, competitor promotions, or seasonal patterns. Marketing and promotions can smooth this out. If player sessions are stable while total revenue fluctuates, the instability is per-session economics. Each player is generating highly variable revenue. This is a machine or operational problem, not a market problem.
The operator in Jakarta had sessions varying by only 3.2 percent week to week while revenue varied by over 40 percent. His problem was clearly per-session economics, which meant the instability was inside the machine or in the operational factors specific to each week. Traffic had nothing to do with it.
Check Whether Payout Timing Coincides With Revenue Dips
For machines where revenue is unstable despite stable traffic, the most common cause is inconsistent payout timing. The machine payout percentage is configured to a specific rate — say 85 percent — but the timing of when those payouts occur matters enormously for weekly revenue. If the machine happens to pay out a disproportionate share of its 15 percent hold during a single week, that week will show low revenue regardless of how many players visited.
Analyze the payout events by week. If a particular week shows a concentration of large payouts that account for the revenue dip, the instability is being caused by natural statistical clustering — the random payout sequence happened to cluster larges wins in that particular week. This is normal and will average out over longer periods. The fix is not to change the machine. It is to track revenue on a monthly rather than weekly basis, which smooths out the clustering effect. Monthly revenue for a properly configured machine should be stable within a few percentage points. Weekly revenue will always fluctuate within a wider band due to the randomness of when large payouts happen to occur.
Look for External Interference Patterns by Shift and Day of Week
If payout clustering does not explain the instability — if the swings persist even in monthly averages — the cause is likely external interference that operates on a schedule. I have seen cases where a machine consistently underperforms on Tuesdays and Thursdays but performs normally on other days. The cause was a nearby business that operated a high-power RF device — in one case a large induction heater — on those specific days. The device was not trying to attack the gaming machines. It was just doing its normal business function. But the RF energy it emitted was enough to disrupt machine operation and bias outcomes.
To test for this, segment your revenue data by day of week rather than by week. If certain days consistently underperform across months, the cause is environmental and tied to that day. Investigate what nearby businesses or equipment operate on a schedule that correlates with the underperforming days. You may not need to shut down the interfering source. Simply moving the affected machines to a different location in the venue, further from the interference source, may be enough to stabilize revenue.
Examine Whether Machine Configuration Changes Between Weeks
Some operators or technicians adjust machine settings — payout percentage, bonus frequency, credit value — in response to what they perceive as the machine being too tight or too loose. These adjustments create exactly the instability they are trying to manage. The machine is set at 85 percent one week. After a bad week for players, the operator bumps it to 88 percent to keep customers happy. The following week, revenue drops because the machine is now paying out three percent more. The operator sees the lower revenue and reduces the payout to 83 percent. Revenue spikes. The operator thinks the adjustment is working. In reality, the configuration changes are the driver of the instability, not the solution to it.
Track every configuration change made to every machine. Log the date, the parameter changed, the old value, the new value, and the reason for the change. Review this log alongside your weekly revenue data. If revenue instability correlates with configuration changes, the instability is self-inflicted. Set your machine parameters to the manufacturer recommended values, leave them alone for at least one full month, and only adjust after reviewing 30 days of stable data that shows a genuine need for adjustment. Most machines, left alone at their recommended settings, will produce stable revenue over monthly periods without intervention.
Install Independent Monitoring for a Baseline Period
If the causes above do not explain the instability, you need independent measurement to establish what the machine is actually doing versus what it is reporting. Install external counters on the coin acceptor and bill validator outputs. Install a bus data logger on the external communication port. Run these for four weeks. Compare the independent measurements against the machine internal reports.
If the independent measurements show stable patterns while the machine reports show instability, the machine reporting system is the problem — data is being manipulated or logged incorrectly. If the independent measurements show the same instability as the machine reports, the instability is real and is being caused by something affecting the physical machine operation. The controlled swap test described in earlier articles applies here: exchange the physical location of the unstable machine with a stable machine. If the instability follows the machine, the cause is internal. If it stays with the location, the cause is environmental.
Look for external equipment changes near your venue. Profit instability that appears suddenly and affects multiple machines simultaneously often has an environmental trigger. A new piece of equipment installed in a neighboring business, a building renovation that changed the electrical wiring layout, a new cell tower activated nearby, or even a change in your own building equipment such as a new air conditioning compressor can introduce electrical noise or RF interference that destabilizes machine operation. When instability appears abruptly, walk the perimeter of your venue and the surrounding area looking for anything that has changed since the machines were last stable. New antennas, new construction, new signage — anything that could emit electromagnetic energy. The source of the instability is often visible if you look for it.
Frequently Asked Questions
How much weekly fluctuation is normal? For a single machine, weekly revenue fluctuation of plus or minus 15 percent is normal due to the randomness of payout timing and minor variations in session count. Fluctuation exceeding 25 percent on a recurring basis, especially when standard deviation of weekly revenue exceeds 20 percent around the monthly average, indicates a cause beyond normal variance. Multiple machines in the same venue should show independent fluctuation patterns. If multiple machines show the same pattern — all up one week, all down the next — the fluctuation is being driven by a venue-wide factor (traffic, interference, or scheduling) rather than individual machine variance.
Should I change my pricing to stabilize revenue? Pricing changes do not stabilize per-session revenue instability caused by the factors described in this article. If the instability is caused by payout clustering, pricing changes will not affect it. If it is caused by external interference, pricing changes will not block the interference. If it is caused by configuration changes, changing the price as well adds another variable to an already unstable system. Stabilize the underlying operation first. Only consider pricing adjustments after you have achieved stable revenue at the current pricing level.
How long should I wait after implementing fixes before expecting stable results? Allow a minimum of four weeks after implementing any fix before evaluating whether revenue has stabilized. Weekly results in the first two weeks may still show the residual effects of previous instability. By week four, the trend should be visible. If instability persists after four weeks, the fix did not address the root cause, and further investigation is needed.