Hold and Win: What the Base Game Pays and What the Bonus Pays

Every Hold and Win slot is two machines sharing one balance: a base game that pays small change while you wait, and a bonus board that pays in lumps. This page splits five of our own logged sessions along that seam, 13,694 paid spins and 128 triggered rounds, and adds a sixth machine with no bonus as a control. The question is plain, how much of what came back came from each half, and the answer changes how a session on these games should be budgeted.

Two machines on one balance

Our Hold and Win guide explains the engine and sets five pooled clocks side by side. This page asks a narrower question of the five sessions logged on this portal, one per carrier: 777 Coins, Royal Joker, Thunder Coins, Gold Express and Little Farm. Each log is cut into two piles of money, and every credit in it goes to one pile or the other. The first pile holds everything the base game paid, line wins and the instant collects that Gold Express and Little Farm run between features. The second pile holds everything paid inside a triggered round, Hold and Win boards and, on the two dual-feature carriers, Free Spins as well.

A sixth log sits beside the five as a control, and it was chosen for what it lacks. Burning Wins is a plain five-line fruit machine from Playson, the studio behind Royal Joker and Thunder Coins, and it has no feature of any kind. Whatever it returns, it returns through the reels, so it shows what a base game looks like when nothing else is coming. None of the six sessions is pooled with anyone else's here, because the split needs a spin-by-spin ledger and only our own logs carry one.

Six logs on one definition

Every session ran on the provider's demo at a fixed stake of 1.00, with natural triggers only and no feature bought. A spin is a paid spin, and respins and free spins aren't counted as spins, because nothing is staked on them. A round's payout is everything credited between its trigger and its settlement. The shares below are money over money: what each half paid, divided by everything staked in that session.

Session777 CoinsRoyal JokerThunder CoinsGold ExpressLittle FarmBurning Wins
Paid spins2,5002,7762,9602,7252,7332,889
Returned in all2,173.003,077.002,368.003,405.252,698.002,572.00
Session return86.92%110.84%80.00%124.96%98.72%89.03%
Paid by the base game1,074.001,181.001,507.001,299.101,459.752,572.00
Base game, share of the stakes42.96%42.54%50.91%47.67%53.41%89.03%
Paid inside rounds1,099.001,896.00861.002,106.151,238.25no rounds
Rounds, share of the stakes43.96%68.30%29.09%77.29%45.31%no rounds
Rounds, share of everything returned50.6%61.6%36.4%61.9%45.9%no rounds

Sources: session reports 777C-20260721-01, RJHW-20260729-01, TCHW-20260729-01, GEXP-20260726-01, LFRM-20260727-01 and BWCL-20260805-01, each published on this portal with its checkpoint table and its round register. The round totals in this table are the totals of those registers, and the base-game figure is everything else the session returned.

The base game pays about half the stake back

The first finding is how little the five base games differ once the rounds are taken out. They returned between 42.54% and 53.41% of the money staked, and the intervals around those figures overlap from end to end. A base-game return measured on 2,500 to 3,000 spins carries a margin of seven to nine points either way, so five logs can't rank five base games. What they can say together is that the reels alone hand back about half the stake. Between rounds the balance therefore falls by about half a bet per spin. The erosion came to 57 bets per hundred spins on 777 Coins and on Royal Joker, 49 on Thunder Coins, 52 on Gold Express and 47 on Little Farm.

The five carriers reach that half of the stake by two quite different roads. The three single-feature carriers play a 3x3 board with five lines, where the smallest win is the stake itself and nothing pays less. Their base games paid on one spin in ten to eleven and a half, with a median payment of 4 bets. The longest silences in the three logs ran 62, 63 and 78 spins. The two dual-feature carriers play twenty and twenty-five lines on a 5x4 board, and their base games paid on one spin in five. Nearly one payment in two came in under the stake, 44% on Gold Express and 45% on Little Farm, and the median payment was about one bet. Twice as many payments bought back the same half of the stake, and the longest silences shortened to 40 and 26 spins.

Burning Wins shows what the other design looks like on the same kind of board. With no feature to fund, its reels returned 89.03% of the stakes in our log, from a payment on one spin in fourteen and a half. The smallest win on its paytable is seven bets, the median payment was 7 and the mean was nearly 13. Set beside its two studio-mates, the control shows where some forty points of return went. Royal Joker and Thunder Coins pay smaller sums a little more often on the same board and the same five lines, and the difference is parked in the bonus. The fruit machine's balance fell 11 bets per hundred spins, while the two Hold and Win boards fell 57 and 49.

What the rounds paid

Here the five logs agree on one number and on little else. The median round paid 45, 46 and 49 bets on the three single-feature carriers, and 38 and 38.75 on the two that split their clock with Free Spins. A median is the round a player meets on most evenings, and on these boards it is worth forty-odd bets whatever the box promises. The means run from 41 to 90 bets, and the whole spread sits at the top of each table. Three rounds carried 47% of the round money on 777 Coins and 57% on Royal Joker, where one 664x board was over a third of everything the rounds paid.

Rounds777 CoinsRoyal JokerThunder CoinsGold ExpressLittle Farm
Triggered rounds15211646, of them 45 played out in the log30
One round per167 paid spins1321855991
Range that count supports, at 95%one per 101 to 29886 to 214114 to 32444 to 8164 to 135
Median round45x46x49x38x38.75x
Mean round73.27x90.29x53.81x46.80x41.27x
Best round191x664x102x169x98x
Rounds under 20x13193
Rounds of 100x or more3415none
Top three rounds, share of the round money47.2%57.2%31.1%21.4%21.6%
Longest wait between triggers634 paid spins594707360464

Round statistics cover the rounds that the log followed to their settlement, which leaves out one Free Spins round on Gold Express. The ranges are exact Poisson intervals on the number of triggers, the rule our methodology page sets for any count of rare events.

Gold Express and Little Farm trade the size of a round for the traffic of rounds. Their rounds arrived every 59 and 91 paid spins, between one and a half and three times as often as on the single-feature boards. Their top three rounds carried only a fifth of the round money, against 31% to 57% on the other three. Inside those shared clocks the two features aren't equals, at least on one of the two machines. On Gold Express the 28 Hold and Win boards had a median of 42.5 bets and the 17 finished Free Spins rounds a median of 15.9. The boards paid 54.50% of the stakes there and the Free Spins 22.79%. On Little Farm the two features ran level, fifteen rounds each, with medians of 40 and 35.1 bets.

The clocks deserve their ranges, because fifteen or twenty rounds is a small count. Fifteen rounds in 2,500 spins leave the 777 Coins clock somewhere from 101 to 298 spins per round, and sixteen rounds leave Thunder Coins at 114 to 324. Those ranges overlap, so the two logs can't say which board is the quicker. Gold Express is the one case a range can support: 46 triggers put its clock at 44 to 81 spins per round, clear of all three single-feature ranges.

The bonus decides the session

Put the two halves back together and the five sessions sort themselves by the second one. Ranked by session return they run Gold Express, Royal Joker, Little Farm, 777 Coins and Thunder Coins, from 124.96% down to 80.00%. Ranked by what the rounds paid they run in exactly the same order, from 77.29% of the stakes down to 29.09%. Ranked by the base game they run Little Farm, Thunder Coins, Gold Express, 777 Coins and Royal Joker, an order with no bearing on the result. Thunder Coins had the second-best base game of the five and the worst session; Royal Joker had the weakest base game and the second-best session.

The arithmetic of the wait shows why the second half decides and the first one does not. On 777 Coins an average gap of 167 spins cost 95 bets of base-game erosion, and the median round refunded 45. Royal Joker's average gap cost 76 bets against a median round of 46, and Thunder Coins' cost 91 against 49. On those three boards the round a player usually meets pays back about half of the wait that bought it. The account is squared, when it is squared, by the few rounds at the top of the table. The dual-feature carriers run closer to even, 31 bets of wait against a 38-bet median on Gold Express and 42 against 38.75 on Little Farm. I would expect exactly this from a format that keeps its value in a long tail. The ordinary round is a partial refund, and the rest sits in the few large boards and in the jackpot tiers above them.

What one log per game can carry

Each figure on this page describes one session, and the two halves aren't equally firm. The base-game shares rest on thousands of spins and still carry seven to nine points of margin either way. Burning Wins carries fourteen points of it, because its wins are fewer and larger. The round shares are far looser than that, and Royal Joker's 68.30% is the clearest case. On the evidence of 21 rounds it could sit anywhere between 14% and 122% of the stakes, since one 664x board moves the whole figure. The split of a session is therefore a description of that session, and the next log of the same game will split differently.

Two things survive that caution, and both of them are of practical use to a player. The base games of all five carriers returned about half the stake, in five logs from two studios, and no interval reaches far beyond three fifths. None of the five builds prints its return, and the figures in circulation, 95.62% to 95.75%, come from third-party databases that we can't confirm against the builds. If those figures hold, the rounds have to supply the rest over the long run. On these base games the rest is somewhere between a third and three fifths of everything staked. The second survivor is the median round, which held between 38 and 49 bets on five different boards. No other figure the rounds produced was as steady, and it is the one I would budget a session around.

How to read a Hold and Win session

The practical reading follows directly from the two piles of money described above. The base game won't carry a session on any of these five boards, since it hands back half the stake and keeps the rest against the next round. A bankroll should therefore be sized for the wait and not for the average. The longest waits in these logs ran 634, 594 and 707 paid spins on the single-feature boards. At half a bet per spin, each of those waits took 300 bets or more before the board opened. On Gold Express and Little Farm the longest waits were 360 and 464 spins, shorter in spins and no cheaper per spin.

The second reading concerns the median round and what a player may ask of it. Forty-odd bets is what a Hold and Win round usually pays on all five boards. A session that needs more than that from its next round is asking for the top of the table. On 777 Coins three rounds of fifteen carried nearly half of the round money, and on Royal Joker a single 664x board out of twenty-one rounds put the session ahead. Thunder Coins never found such a round in sixteen tries, and its session closed at 80.00% with the second-best base game on the page.

The five reviews hold the full tables, the frames and the pooled clocks: 777 Coins, Royal Joker, Thunder Coins, Gold Express and Little Farm. The control has a review of its own, Burning Wins, written on the same protocol. All six builds play free on our Demo Shelf, where a few hundred spins will show the wait better than any table; 18+ applies there too.

Quick answers

How much of a Hold and Win slot's return comes from the bonus?

In our five logged sessions the rounds paid between 36.4% and 61.9% of everything returned, and the base game paid the rest. Measured against the stakes, the base games returned 42.54% to 53.41% and the rounds 29.09% to 77.29%. The base share is the steadier of the two, about half the stake on every board, while the round share moves with a handful of large boards.

Does a busier base game mean a better session?

It did not in these five logs, and two of them show the opposite. Thunder Coins had the second-best base game of the five and the worst session, and Royal Joker had the weakest base game and the second-best session. The five sessions ranked by return in exactly the order of what their rounds paid. The base game sets the pace of the losses, and the bonus sets the result.

Why is Burning Wins in a Hold and Win comparison?

It is the control for the two Playson boards in the comparison. Burning Wins comes from the same studio as Royal Joker and Thunder Coins, plays the same 3x3 board with five lines, and has no feature at all. Its reels returned 89.03% of the stakes in our log, against 42.54% and 50.91% for its two studio-mates. The gap shows how much of a Hold and Win game's return is moved out of the base game to fund the board.

Are these splits the games' real numbers?

They are the numbers of one logged session per game and nothing more. A base-game share measured on 2,500 to 3,000 spins carries seven to nine points of margin either way. A round share that rests on 15 to 46 rounds can be tens of points out. None of the five builds prints its return or a split of it, so there is no studio figure to set these beside.

Erik Lindholm
Written by Erik Lindholm

Former studio-side game mathematician: paytables and RTP models, now checked from the player's side. Full profile and method →