Relax Gaming · measured RTP 100.1% · tested 2026-09-30
Marching Legions
Fifty 120x Marching Legions free spins purchases lost FUN 5,078 in the first demo run, 16 stuck before a second wave, yet 20,000 rounds over two runs measure RTP at 100.1% against 98.12%.
In numbers: the lab's RTP test on the official demo server measured 100.1% against 98.12% declared over 20,000 logged rounds in 2 runs; 25.6% of rounds paid; a natural bonus landed once per 3,333 rounds (6 in total) and a bonus round paid a median 47.3x the stake, mean 100.9x, trigger spin included; 100 bonus buys at 120x returned 0.83x of their price on average, 86% of them below the price; the biggest round paid 420.5x; the longest dry run was 33 spins.
Marching Legions demo
Same reels, real payouts
Cafe Casino runs the same Marching Legions with the same math and adds a welcome bonus for players from the United States:
300% up to $2,000 + 150 FS 18+ · T&Cs apply · the demo stays freeTest card
20,000 spins · FUN 2.00 bet · every request logged
| Spins | 20,000 · 2 runs |
| Bet | FUN 2.00 |
| RTP declared | 98.12% |
| RTP measured | 100.14% (approximate 95% CI 90.2–110.1) |
| Hit frequency | 26% · every 3.91 |
| Dead streak max / avg | 33 / 3.94 |
| Median hit | 1.20x |
| Max win (no bonus round) | 420.5x · run 2, round 712 |
| Free spins, natural | 6 · 0.3 bHz · every 3333.3 spins · mean bonus round payout 100.9x stake |
| Balance start → end | 5,000 → 5,058 |
- 012026-09-2410,000 rounds · 50 buysfiles & checksums ↓93.51%RTP5 files · SHA-256 ↓
- 022026-09-3010,000 rounds · 50 buysfiles & checksums ↓106.78%RTP5 files · SHA-256 ↓
each run keeps its own spin log, CSV and raw server journal with checksums; the figures on this card combine all runs
Bonus buy lab
100 buys × 120x · FUN 240 each
mean 0.83x · median 0.292× price (FUN 70.00) · 86 of 100 below price · best 9.91x · worst 0.00x
| Spins per bonus (avg / min–max) | 15.6 / 6–44 |
| Extended | 74% |
| Bonuses that paid nothing | 9% |
Marching Legions extends its free spins by waves: three horn symbols on reel five open the next wave with one more stack than the last, and only the first wave starts with one of them already counted. The retrigger rate on this page is the part of free spins rounds that reached a second wave: 34 of 50 tickets and two of three meter features in run one, 40 of 50 and three of three in run two.
- Spent
- FUN 24,000100 × FUN 240
- Returned
- FUN 19,92283.01% · approximate 95% CI 47–119%
- Net
- −FUN 4,078−17.0 buys
- Break-even or better
- 14 / 10014%
- Paid 2× price or more
- 9 / 1009%
- Price in base spins
- 120natural trigger every ~3,333 spins
One ticket equals 120 ordinary stakes. Natural bonuses occurred once per 3,333 ordinary rounds in this sample. Stakes multiplied by this interval measure gross wagering, not the net cost of obtaining a bonus. Purchase returns above include the complete purchased round and are divided by the ticket price.
Risk profile
how much bankroll the 2 runs in sequence needed
| 50 bets | bust at spin 157 |
| 100 bets | bust at spin 435 |
| 200 bets | bust at spin 2966 |
The runs are played back to back as one series, one bet size. A different session can bust earlier or never; treat this as a measured example, not a rule.
How it plays
19 rules screens from the game
Marching Legions plays 5x3 with 243 ways: symbols that match on neighbouring reels from the left pay regardless of row, and extra copies on a reel multiply the count. There is no Wild. The legionary is the top symbol and the only one that pays for a pair, FUN 2.00 for two on a FUN 2.00 stake and FUN 5.00 for five; the eagle standard pays FUN 6.00 for five, the laurel wreath 5.00, the gladius 4.00, and the three badge symbols 2.40, 2.00 and 2.00.
A reel filled top to bottom with legionaries becomes a Marching Stack: it sticks, a re-spin follows, and before every re-spin each stack steps one reel to the left, so a stack that lands on reel five walks through four more positions before it leaves. Any further full stack landing during the re-spins falls in with the march, re-spins continue while a stack is on the screen, and a stack that leaves the reels is collected on the shield meter. When no stack is marching and at least two partial stacks have landed, the nudge feature may push them into a full stack and start the march that way.
The meter has five levels and fills at 330 collected stacks, which opens the free spins at the weighted average of the bets used during the collection. Free spins run in waves: the first wave starts with one Marching Stack on the far right reel and lasts until no stack remains, three horn symbols landing on reel five unlock the next wave, the first wave starting with one already counted, and each new wave begins with one more stack than the last, five from the fifth wave on. Buying the feature costs 120 times the stake, FUN 240 here, opens the first wave at once and leaves the meter untouched. The rules give 98.12% RTP for ordinary play and 97.13% for the purchase; the game data caps wins at 10,000x and puts the bet range at FUN 0.10 to FUN 1,000.
Respin
Sticky symbol
Free spins
Bonus buy
Bet range FUN 0.10 – 1000 per spin · payouts read from the game's paytable at bet FUN 2.00
Gallery
11 frames from the demo session
Verdict
Five stacks, three tickets
The second Marching Legions run lifted the series: 93.51% then 106.78%, 100.14% pooled over 20,000 spins at FUN 2 and 28.9 bets ahead against 98.12% in the rules, and the same mechanic paid both times: rounds with a marching stack, 666 and then 631 of 10,000, returned 67.0% and 67.9% of the money at a median of 4 and 4.5 bets, with 164 and then 132 of them paying nothing. What a march is worth depends on how many stacks move together: one averaged 6.4x and 5.62x, two 21.7x and 16.69x, three or more 95.9x and 113.2x, and four stacks landing together on round 712 of run two made 420.5x, the series record.
Through the shield meter the free spins arrived three times per run, once per 3,333 spins, and from 100 tickets at FUN 240: the tickets returned 0.577x and then 1.0833x of price, 0.8301x over the hundred, with 86 below their price, nine zeros and a FUN 70 median, and everything hung on the waves. Tickets that ended in the first wave averaged 0.092x and 0.023x, in the second 0.294x and 0.263x, in the third 1.218x and 0.748x, in the fourth 4.358x and 3.186x, and run two's three fifth-wave tickets, 50, 19 and 32, averaged 9.708x, each holding a FUN 1,215 spin from five stacks marching together, against a FUN 810 best spin in run one. The meter's own features paid FUN 78, 96 and 111 in run one and 81, 752 and 93 in run two, the 752 from three waves on round 6605.
Hits came on 25.65% and 25.44% of spins at a 1.2x median, the longest droughts were 33 and 30 spins, and a stack arrived while another was already walking in 124 and then 34 rounds. The best single spins outside the free spins were FUN 270 on round 6962 of run one and FUN 540 on round 4033 of run two; the balance closed 648.7 bets down and then 677.6 up, run two never falling more than 220.3 bets below its start, so the pooled low of 886.4 bets remains run one's.
- Three tickets reached a fifth wave in run two and averaged 9.708x their price
- Rounds with three or more stacks marching averaged 95.9x and 113.2x in the two runs
- Four full stacks landed at once on round 712 of run two, 420.5x in five re-spins
- Run two closed 677.6 bets ahead at 106.78%, above its start on 9,317 spins
- Below their FUN 240 price came 86 of 100 tickets, nine of them with nothing
- Twenty-six of 100 tickets ended in the first wave, at 0.092x and then 0.023x of price
- Nearly a quarter of the 1,297 stack rounds paid nothing, 164 and then 132
- The shield meter still needs about 3,300 spins per fill, three fills per run
Run 01·2026-09-24·93.51%
A full stack landed or was nudged into place in 666 rounds, one per 15 spins, and those rounds paid 67.0% of everything the series returned, though their median was 4 bets and 164 of them paid nothing. Grouped by the most stacks marching at once, a lone stack averaged 6.4 bets, two marching together 21.7 and three or more 95.9, and the two biggest rounds, 6962 at 175.8 bets and 6768 at 151.5, both had three stacks moving at once.
Fifty tickets at FUN 240 returned 0.577x their price, FUN 6,922 back from FUN 12,000 against the 97.13% given for the buy. The demo balance ran dry at ticket 41, and the series continued in a fresh session. The 16 tickets that ended after the first wave averaged 0.092x, 22 that ended in the second 0.294x, ten that ended in the third 1.218x and the two that reached a fourth 4.358x.
Ticket 32 was one of those two: 33 spins across four waves, and on the 24th spin four stacks marching together paid FUN 810, 405 bets, followed by FUN 324 on the next, FUN 1,408 in all. Ticket 43 made FUN 1,338 in three waves, with a FUN 810 spin of its own, and ticket 8 FUN 684 in four; nothing else passed FUN 480, and the other 49 tickets averaged 0.469x without ticket 32.
A quarter of spins paid, 25.65%, and the meter filled three times, at rounds 3170, 6430 and 9248, for rounds worth FUN 78, 96 and 111 with their trigger spins, about 1.5% of the money.
Run 02·2026-09-30·106.78%
Three tickets reached a fifth wave and each of them had the same spin in it: five stacks marching together paid FUN 1,215, 607.5x the stake, on ticket 50, ticket 19 and ticket 32, which finished at FUN 2,379, FUN 2,331 and FUN 2,280, 9.91x, 9.71x and 9.5x of the FUN 240 price. Those three carried the sample: fifty tickets returned FUN 13,000, 1.0833x of price, yet 42 came back below it, five paid nothing, tickets 12, 15, 28, 36 and 40, and the median was FUN 91.50; by the wave a ticket reached, the ten that ended in the first averaged 0.023x, 21 in the second 0.263x, 13 in the third 0.748x, three in the fourth 3.186x and the three in the fifth 9.708x, with 17.4 spins per ticket on average and 44 at most.
A full stack marched in 631 rounds, one per 15.8 spins, and those rounds paid FUN 14,510.80, 67.9% of everything the series returned, though 132 of them were empty and the median was 4.5x; a lone stack averaged 5.62x, two marching together 16.69x and three or more 113.2x, and a stack that landed on reel five was worth 7.96x on average against 5.27x for one on reel one. Four full stacks hit reels two to five together on round 712, which paid 420.5x in five re-spins, more than any round of either run, round 1234 did the same for 340.5x, and round 4033 held a FUN 540 spin, 270x, the best single spin outside the free spins; in 34 rounds a second stack fell in beside one already on the move.
The shield meter filled three times, at rounds 3175, 6605 and 9838, and the middle one ran three waves for 376x, FUN 752, with a FUN 270 spin, 4.3% of the return from three features; the other two paid 40.5x and 46.5x. A quarter of spins paid, 25.44%, at a 1.2x median, and the longest drought was 30 spins from round 5288; spin 9 put the balance ahead and it stayed above its start on 9,317 of 10,000 spins, bottomed only 220.3 bets down on spin 695 and closed 677.6 bets ahead, 106.78% of the stake, with 2,000-spin blocks of 124.3%, 95.0%, 96.2%, 111.3% and 107.1%.
Test record
2 runs · 20,000 rounds · 100 buys · open data · closed rig
This game has been through 2 separate test runs on the official demo server, each a fresh session of 10,000 rounds at a fixed bet. Every figure on this page combines the runs into one sample, as if they were one long session; the record below keeps each run apart, with its own numbers and its own downloadable log.
- 20,000rounds, all runs
- 26%rounds that paid
- 421xbiggest round
- 1 / 3,333natural bonus rate
- 0.83xmean of 100 buys
- 33longest dry streak
01run Run 012026-09-24 measured return93.51%-4.6 pts vs declared
- 95% band
- 83.6–103.4%
- rounds that paid
- 26%
- biggest round
- 176x · round 6,962
- longest dry streak
- 33
- natural bonuses
- 3 · one per 3,333 rounds
- bonus buys
- mean 0.58x · median 0.25x · 88% below price
- balance start → end
- 5,000 → 3,703
- deepest drawdown
- 1,034x
02run Run 022026-09-30 lab note measured return106.78%+8.7 pts vs declared
- 95% band
- 89.4–124.1%
- rounds that paid
- 25%
- biggest round
- 421x · round 712
- longest dry streak
- 30
- natural bonuses
- 3 · one per 3,333 rounds
- bonus buys
- mean 1.08x · median 0.381x · 84% below price
- balance start → end
- 5,000 → 6,355
- deepest drawdown
- 466x
Marching Legions' second run, 30 September, returned 106.78% after 93.51% on 24 September, 100.14% pooled, the rules saying 98.12% for a series that now spans 20,000 spins at FUN 2. Marching stacks came 631 times against 666 and paid 67.9% of the money against 67.0%, the largest round rose from 175.8x to 420.5x on round 712, and the FUN 240 tickets returned 1.0833x of price after 0.577x, three of them past 9x from a fifth wave.
2 test runs combined into one sample (2026-09-24, 2026-09-30): 20000 ordinary paid rounds, including naturally triggered features, at a fixed FUN 2.0 bet; 100 bonus buys at 240.0 each (120.0x) on the official Relax Gaming demo server (jurisdiction MT, RTP profile 98.12%); every request and response logged.
The test rig plays the official demo build over the network and records every server round: bet, win, intermediate steps (cascades, respins), free spins, collect. Screenshots and video were captured in a separate browser session and are not part of the statistical samples. The text was read against the log by an editor before publication. How we test → One test, step by step →
FAQ
6 questions · answered from the log
Is Marching Legions at its 98.12% after two runs?
Yes, slightly above it: 100.14% pooled across 20,000 spins of FUN 2, 93.51% in the first and 106.78% in the second, 28.9 bets ahead in all. Marching stacks supplied 67.0% and 67.9% of each run's money, the meter's bonus rounds 1.5% and 4.3%.
Is a fifth wave the whole difference?
Yes, among the tickets, mostly. A fifth wave was worth about 9.7x the price in run two: tickets 50, 19 and 32 reached it, ran 44, 35 and 42 spins and paid FUN 2,379, FUN 2,331 and FUN 2,280, each with one FUN 1,215 spin from five stacks marching together. Run one never got past a fourth wave, where two tickets averaged 4.358x; across both runs 74 of 100 tickets reached a second wave, 31 a third, eight a fourth and three a fifth.
Are the stacks any more frequent in run two?
No, slightly less: 631 rounds with a march against 666, one per 15.8 spins against 15, paying 67.9% of the money against 67.0%. A lone stack averaged 5.62x, two together 16.69x and three or more 113.2x, and a stack starting on reel five, 103 of them, averaged 7.96x against 5.27x for one starting on reel one.
Which base round paid most?
The 420.5x of round 712 in run two: four full stacks landed at once on reels two to five and marched for five re-spins. Round 1234 repeated the four-stack start for 340.5x and round 4033 held the largest single spin outside free spins, FUN 540, 270x; run one's best was 175.8x on round 6962.
Did the meter fill again without buying?
Yes, three times, as in run one, at rounds 3175, 6605 and 9838, roughly every 3,300 spins. Round 6605 ran three waves for 376x, FUN 752 with a FUN 270 spin, the other two paid 40.5x and 46.5x, and the six meter features across both runs returned FUN 1,211 between them.
What are the biggest spins of the series?
FUN 1,215, 607.5x, three times in run two, once on each of tickets 50, 19 and 32 with five stacks marching; run one's best was FUN 810 on tickets 32 and 43. Outside the bought features the record is FUN 540 on round 4033 of run two, and the 10,000x cap listed in the game data was never approached.
Comments and reviews
Leave a comment or your own test of this slot
- spins20,000
- bet2.00
- result+29x
- RTP measured100.1%
- bonuses6
- max win, no bonus420.5x


















