Starlight Princess 1000 — Game Details

A borrowed engine at an honest price

I read Starlight Princess 1000 as a fairly priced multiplier ladder standing on a paytable too weak to hold it up. Every option the player controls returns within a hundredth of a point of 96.50%. Standard spins, Ante Bet and the 100x buy are one bet at three speeds, and none of them hides a surcharge. What all three buy is exposure to the top of a sixteen-value multiplier ladder, because the symbols underneath pay too little to matter on their own. Honest pricing on a thin engine is the whole case for this game, and I find it is also the whole case against it.

None of the engine is new. Starlight Princess, the game this one follows, already paid symbols anywhere on the screen, tumbled winners away and summed multiplier symbols when a sequence ended. The mechanisms that define the sequel, pay-anywhere tumbles, multipliers up to 1000x and a free spins total multiplier, are the same three that define Gates of Olympus 1000, while Gates of Olympus and Sweet Bonanza run close variants of the formula. Beyond a higher multiplier ceiling, I find nothing structural that the 1000 adds. A game with no mechanism of its own is judged on what its maths costs and what it can reach, so that is where I take the rest of this review.

Eight of the nine symbols are filler

I count a paytable that tops out at 50x, for 12 or more of the best symbol, and the 15,000x cap is 300 times that. Three groups of 12 need 36 positions, so a 30-position screen holds two at most, and the richest regular-symbol screen pays 50x plus 25x from the second high symbol, 75x in all. Reaching the cap from the paytable alone would take 200 of those screens back to back. That is not a route to the cap. The multipliers are.

Below the top symbol the drop is steep. Its lowest tier, 10x for 8 or 9, equals the second high symbol's middle tier and the best low symbol's top tier. The other eight regular symbols pay 2.5x or less for 8 or 9, four of them pay less than the stake, and a fifth returns exactly the stake. Most of the paytable's value waits behind the 12-symbol threshold, where the fourth high symbol jumps from 2x to 12x and the best low symbol from 1.5x to 10x. Twelve matching symbols is two fifths of the screen.

The single best pay on the table does not belong to a regular symbol at all. Six or more scatters pay 100x, double the best symbol's top tier, and those same scatters start free spins. Even the paytable's peak points at the bonus round.

The cap cuts the top rung short

The multiplier ladder runs through sixteen values from 2x to 1000x, and eleven of them are 25x or less. Adding the sixteen together, I get 2,010, of which 250x, 500x and 1000x account for 1,750, or 87%. The reach of the ladder lives in three rungs.

Those three rungs collide with the cap quickly. A single 1000x on any sequence win of 15x or more reaches 15,000x, and 15x is only the top tier of the third high symbol. Put the same 1000x on the best symbol's 25x middle tier and the nominal 25,000x is cut to 15,000x, discarding 40% of it; on the 50x top tier the cap discards 70%. Because base-game multipliers in one sequence are added, a 500x and a 1000x together make 1,500x, which caps the best symbol's cheapest 10x win. The top rung pays its full face value only on wins of 15x or less, down to the 250x it makes of the cheapest 0.25x win.

The lower rungs meet a paytable that gives them little to work with. A 20x multiplier on 8 or 9 of any of the eight lesser symbols returns between 5x and 50x, and a multiplier in a sequence without a win returns nothing. Base-game multipliers need a win in the same sequence to pay at all, and most of the wins available to meet them are small.

The cap itself is not decoration. At 15,000x with a chance of 1 in 2,749,771 per spin, capped rounds supply about one unit in every 177 the game returns at 96.50%. For a single outcome I call that a real share of the maths, and it is why the ceiling reads as part of the design rather than a number for the lobby.

Free spins pay for the order of events

The round opens with 15 spins and adds 5 each time 3 or more scatters land. The running total grows only when a multiplier lands on a winning spin, and the total multiplies only wins on which a new multiplier lands. One event both builds the total and cashes it. A win without a fresh multiplier pays the flat paytable, at most 75x per screen of the tumble sequence from regular symbols, whatever the total has reached.

That rule makes sequence matter more than content. Take two events: a 2x multiplier on a 50x win and a 100x multiplier on a 2x win. With the 2x first, the round pays 100x and then 102x on the 2x win, 304x in all. With the 100x first, it pays 200x and then 102x on the 50x win, 5,300x in all. Same symbols, same multipliers, more than seventeen times the payout. Nothing visible on the reels says which kind of round is under way, and I count that as the round's main weakness, because a large total can sit idle while the spins run out.

The same rule decides how the cap binds in the bonus. Once the total, fresh multiplier included, reaches 300x, 12 of the best symbol is enough on its own to end the round at 15,000x, and whatever spins remain are forfeited.

Ante Bet and the buy change speed, never value

Ante Bet raises the stake by a quarter and doubles the natural chance of free spins. Per unit staked I make that 2 divided by 1.25, or 60% more triggers. Paytable figures stay tied to the base bet, so every base-game pay is worth four fifths as much against the stake actually placed. With the return fixed at 96.50%, the value the paytable loses against the bigger stake has to come back through the extra rounds. I read Ante Bet as the free spins round bought in instalments.

Buy Free Spins costs 100x and returns 96.49%. I make the expected cost of a purchase 3.51x the bet, against 3.50x for 100x of standard spins, a difference of a hundredth of a bet. Its best case is 15,000x, 150 times its price, which is a fair ceiling for the money. To return its own cost the round has to find 100x across its 15 starting spins, and from the paytable alone that takes anything from two top-tier groups of the best symbol to 400 of the cheapest 8-symbol wins. The buy is fairly priced for a round whose break-even rests almost entirely on the ordering problem above. The two options cannot be combined, because the buy is disabled while Ante Bet is on.

The price difference that does matter sits outside both. Casinos can run the game at 95%, where the house keeps 5 units per 100 staked instead of 3.5, which makes the same game about 43% more expensive to play. Choosing between buying and spinning moves the return by 0.01 points, while the casino's setting moves it by 1.5. The rules screen shows which setting is running.

Nobody should come here for the base game

I see this suiting a player who wants a fairly priced shot at a large multiple and accepts that the spins between multiplier hits return little. For that player the buy is sound, since it costs a hundredth of a point in return, and Ante Bet is the slower equivalent. Anyone who wants the paytable to hold a balance steady between features should look elsewhere, because eight of the nine regular symbols cannot. Anyone who already knows Gates of Olympus 1000 gains no mechanism by switching, and anyone at a casino running the 95% setting pays 43% more for an identical game.

Honest on price, borrowed in design and thin between multipliers, it is a game I place at the top of competent and short of good.