Fortune Rabbit vs Speed Winner: Two PG Soft Engines, One Log Each

Two engines, one disclosure habit

Fortune Rabbit and Speed Winner come from the same studio and share one fact before a single spin is played: neither build prints its return. PG Soft publishes the figure for each game on the studio's website and in a Game Information sheet that anyone can download, and a player at the reels sees neither. Everything else about the two machines is different, starting with the way a win is built. Fortune Rabbit is a three-reel count in which gold tiles land with a value printed on them. When five or more tiles fill the ten-position grid they pay the sum of their printed values, in the feature and in the base game alike. Speed Winner is a six-reel cascade in which the symbols that pay are cleared and new ones drop into the gaps. Every multiplier tile left on screen is then added into one figure that multiplies the whole spin. One machine adds and pays flat; the other adds and then multiplies.

The question this page asks isn't which of the two is better, since both keep the return off the screen and both earn the same mark for it. It is where each machine puts its money, how often it hands some over, and how far two logged sessions moved from what the sheets promised. Both sessions ran warm, which is the first thing to hold in mind and the last thing to draw conclusions from.

What the two sheets promise

Each sheet carries a certified return and a split of that return between the base game and the feature, and the two splits differ more than the two returns. Fortune Rabbit's 96.75% is built from 71.51 points of base game and 25.24 of feature, so by design a quarter of the money arrives through the fortune spins. Speed Winner's 96.72% is built from 78.50 points of base game and 18.22 of feature, less than a fifth through the free spins. The two frequency rows are printed without a definition, 0.77% for Fortune Rabbit and 0.44% for Speed Winner. Our reading of those rows is one trigger in 130 spins for the tile and one in 227 for the scatters. The caps sit twenty times apart, 5,000x against 100,000x, and Speed Winner's sheet adds a footnote of its own. Its billion-spin simulation produced a largest result of 8,345x, which is where the advertised figure meets arithmetic.

Both sheets carry a last update of 8 July 2025, and the frequency rows are quoted as printed. The one-in-N readings are ours, on the assumption that each row counts triggers per spin.

Two logs on one definition

The Fortune Rabbit session ran on 2 August 2026 at a flat 6.00 for 2,624 paid spins; Speed Winner's ran the next day at a flat 12.00 for 2,757. Both are counted the same way here, on paid spins only, so fortune spins and free spins sit outside the count. A round is measured as the balance move between its trigger and its settlement, and the return is money returned over money staked. Every multiple on this page is a multiple of the session's own stake, 6.00 on one machine and 12.00 on the other. Speed Winner's figures are the corrected ones of 16 September, after a booking error in its log was found and fixed, and its session report carries the details.

Sources: session reports FRBT-20260802-01 and SPDW-20260803-01, both published on this portal with their checkpoint tables. Speed Winner's below-stake share is measured on the 612 wins whose amount the wallet shows on its own. The 134 wins that settled together with a neighbour are left out of that one figure.

The clock: a random tile against four scatters

Fortune Rabbit's feature fires at random on any spin, and the log shows what random looks like at this size. Fifteen rounds arrived in 2,624 paid spins, one per 175, with the gaps between them running from 10 paid spins to 586. Speed Winner asks for four scatters, and its twelve rounds came one per 230, with gaps from 2 paid spins to 454. Five of the twelve were packed into one stretch of 167 spins late in the run, which is the cluster that carried the session. Both clocks ran slower than our reading of the sheets, and neither by a margin that twelve or fifteen rounds could settle.

The price of waiting is where the stake enters, because a silence is measured in bets rather than in minutes. Fortune Rabbit's longest silence cost 3,516.00 at 6.00; Speed Winner's cost 5,448.00 at 12.00. In stake units the two waits are nearly the same, 586 bets against 454, and the Fortune Rabbit wait is the longer one. I would expect a random trigger to produce exactly this, no rhythm, a few clusters and one silence that reads as a fault. Consistent with that, our three Fortune Rabbit logs disagreed on the rate by a factor of nearly four, between one trigger in 113 spins and one in 417.

The shape of a round: a sum against a multiplier

Here the two machines part company, and the parting isn't visible in the median. Fortune Rabbit's fifteen rounds had a median of 32.5 stakes, and Speed Winner's twelve had a median of 32.2. Anyone pricing a round from the middle of the table would price them the same, and they'd be wrong about one of them. The means say otherwise: 39 stakes against 107, and the gap between the two averages is the whole comparison. Fortune Rabbit's rounds are sums of printed tiles, so they ran from 2.5x to 127x and stayed in a band. Three of the fifteen paid less than 100.00, two paid more than 500.00, and everything else sat in between. Speed Winner's rounds are cascades under a summed multiplier, so eleven of them paid between 4.2x and 102.3x and the twelfth paid 841.15x. The round ran for eighteen free spins and closed with a single payout of 9,281.40, a 299.40 line total under a summed x31.

One round therefore decided one session, and no round decided the other. Speed Winner's best round was 62% of the session's net result, whereas Fortune Rabbit's best round was 17% of its net. Fortune Rabbit's best base-game hit, 732.00 from a five-tile count, outpaid fourteen of its fifteen rounds. The feature shares follow the same line, and they run against the sheets. Fortune Rabbit's rounds delivered 17.4% of everything returned, against a sheet that budgets 26.1% for them. Speed Winner's rounds delivered 31.3% against a sheet that budgets 18.8%, so two sessions inverted two sheets. The honest reading is that a dozen rounds say a great deal about the shape of a round and almost nothing about its average.

Between the features

Fortune Rabbit's base game paid once in four and a half spins, and half of those payments returned less than the stake. Its longest no-win run was 37 spins, and its deepest drawdown was 752.40, or 125 bets, before the tiles caught up. Speed Winner paid on just over one spin in four, with 44% of its readable wins below the stake. Its longest run without a payment was 22 spins, and its deepest drawdown was 2,241.00, or 187 bets. The busier surface belongs to the cascade, and so does the deeper hole. A game that pays more often can still walk further downhill between the payments that matter.

The best base-game hits tell the same story from the other end. Fortune Rabbit's tiles pay in the base game as they do in the feature, and its best base-game spin, 122 stakes, was almost the size of its best round. Speed Winner's best base-game spin, 70 stakes, was an x10 sum over an 84.00 cascade, and it was a twelfth of the best round. On one machine the feature is a longer version of the base game, and on the other it's a different game on the same reels.

What two warm afternoons can and cannot say

Both sessions closed ahead, 127.65% and 149.43%, and the second figure is the highest this portal has published. Neither is a property of its game, and the sheets say 96.75% and 96.72%. The two returns are draws from distributions nobody outside the studio has seen in full, and a third afternoon on either machine could as easily run cold. What the logs can say is structural, and it survives the luck. Fortune Rabbit spreads its money, with a base game that pays small change often and rounds that pay thirty-odd bets more often than not. Its cap has a single route, ten top tiles on one screen, and no log came near it. Speed Winner concentrates its money, with a base game that pays a little more often and slides a little deeper. Its rounds share Fortune Rabbit's median and add a tail, and the tail carried the whole afternoon.

The practical difference is the budget, and it runs in opposite directions on the two machines. A Fortune Rabbit stake sized for the wait between features misreads the machine, because the tiles pay between them. A Speed Winner stake sized for the median round misreads that one, because the median isn't where the money lives. On disclosure the two are level, and the mark against each build is the same: the return sits on a website, not on the screen. Speed Winner also sells its feature at 50x the stake, above its median round and below its mean, which prices a wide distribution rather than a shortcut. Fortune Rabbit sells nothing, because a random trigger doesn't need a button. Both builds play free on our Demo Shelf, and a few hundred demo spins on each will show the two clocks better than any average; 18+ applies there too.

Sessions FRBT-20260802-01 and SPDW-20260803-01 were logged on the provider's public demos on 2 and 3 August 2026. The two reviews, Fortune Rabbit and Speed Winner, hold the full session tables and the frame captions. Both engines sit beside their studio-mate on the PG Soft hub, and the cascade family is charted on our cascade slots guide.