RTG also have this artificial rule that a free spins feature will be halted early once 40,000x line bet is reached - this is certainly not random, it takes away the top slice of truly awesome payouts, yet leaves in place the really crappy bonus rounds that pay zero.
Thanks for the welcome all!
Re the capping at 40000x line bet:
It's actually very common in the land-based industry to cap wins derived from a particular paid game (this is essential in many markets due to government regulation. In Australian States, for instance, capping is set at $10,000 or $5,000 (depending on State) irrespective of the amount bet (which can be up to $10)).
This does not apply to some jackpot wins.
However, in all instances (other than extremely low caps such as 50p/35 pound machines in the UK), the effect of such a cap on the mathematics is utterly negligible.
Capping out in the extremely rare instances of wins of that magnitude makes way less than 0.1 of 1% difference..
The Microgaming slot "5 Reel Drive", one of their early ones, is using short reel strips, but they are weighted. Some calculations and simulations were done that showed that playing different numbers of paylines made the payback different. The variation was between 88% and 96% if I can remember, with 96% being from playing all 9 lines, and playing 5 lines being the worst. Further, individual paylines paid out more than 100%, but the software does not allow individual lines to be selected.
Re combination layout making a difference to overall RTP:
With all of the games we produce for both markets (net and land-based) a full cycle test is run at min lines, multiple lines and max lines to ensure that there is actually nil difference (or within 0.1%) between total theoreticaly single-line combinational maths and the actual result.
This is also a requirement with most government testing authorities.
It's possible that a game like the as you say in the MG example, there is a very, very short game cycle (total number of possible combinations), and a reel strip layout that makes some difference, but I'd be surprised if it was that high...can't rule it out though!
At any rate I can't think of a single game that I've designed (and we're clocking 300 in the current company and over 800 in total, albeit not all deployed) that would have a significant variance based on line distribution and reel strip layout.
Since RTG slots have some kind of weighting to match the operator settings, it is possible that strategy could result in better value from the game.
Re reel strip weighting:
It's extremely rare to have reel strips that are exactly the same as one another.
Weighting (having different numbers of particular symbols on different reels) is necessary to balance game RTP, hit rates, feature hit rates and so forth.
The main example I can think of in relation to non-weighted strips were some of the old stepper (mechanical slots), but even these evolved to have the same number of total symbols on a strip, but different weighting of them (eg: all 5 reels might have 30 total symbols, but there is difference between the distribution of higher paying symbols and lower paying).
There's nothing sinaster in doing this, however I tend to find that unless required due to a bizarre jurisdictional requirement, it's best to have varied length of each reel strip in order to achieve a more playable game.
Yes, it is the ability of RTG operators to weight the slot games to give payouts between 93% and 97%.
The important thing to note here is that I have never come across a net or internet system that would allow individual operators to have control over RTP on the fly (mind you, I've only ever dealt with reputable systems so I'm sure there's some software teams out there that directly run casinos and probably have dodgy functionals like this).
In reputable cases, however, it would be crazy to have this form of control.
Hiking back to Australian land-based as another example, operators can only change their average RTP settings once a month.
As with both internet and land-based systems that I know of, operators can only select between available RTP settings (eg: between 95% and 97.5%).
There are exceptions, but again they're jurisdictional in nature.
UK AWP machines, for instance, are non-random in certain instances.
They have a requirement to have a "reactive" component to their random calculations.
Before any given spin these machines test to see whether RTP over a given cycle or timed period (such as previous 30 days) is currently running high or running low by a certain degree (for instance, higher or lower than 3% of expected average RTP).
If it's running low, it awards a series of top prizes (in these cases 35 pounds - though soon to shift to 50 shhh) in order to bring RTP back into line...so a player may hit 4 or 5 top prizes literally in succession.
This is why players develop strategies in these markets to throw away (through voiding reel nudges that are part of these games) small wins in order to decrease overall RTP over a period.
Conversely, if it's running high, the machine enters a cycle of "sorry buddy, no way in hell are you getting any form of prize" until RTP has fallen back into line.
Haven't come across this function in the net providers I've dealt with though.
...yet leaves in place the really crappy bonus rounds that pay zero.
I hate it when that happens!
The effect of randomness I suppose...but you'd have to be pretty unlucky to get nothing from a free game series...and personally I'd never design a game that had an instance win bonus round that couldn't yield a win!
Da Dog