You know what's coming. As is our wont, we forwarded this question to Bob, who explains what he said and what it means. Take it, Bob.
Only a slight misquote. What I said was that the overall return of the game isn't affected by whether its single line, Triple Play, Five Play, Ten Play, Fifty Play, or Hundred Play. That's different than whether or not there's any advantage to the multi-line games.
The variance of multi-line games is considerably less than that for the single-line version, although there are short-run differences as well as long-run differences. You'll see rules of thumb in some books as to how much less volatile these games are, but there are at least two distinct ways of calculating this.
You can figure it out for the same denomination, i.e., comparing single-line quarters (where you bet $1.25), Triple Play quarters (where you bet $3.75), up to Hundred Play quarters (where you bet $125).
You may also figure this out for the same total bet, i.e., comparing nickel Hundred Play, dime Fifty Play, 50ยข Ten Play, $1 Five Play, or $5 single line -- at all of which your bet is $25.
Both of these are legitimate ways to look at the problem, and both are discussed at length in my new book, Video Poker for the Intelligent Beginner. Until now I haven't seen any book that correctly differentiates between these two ways of calculation.
The raw data to do the calculations provided in the book comes from the Video Poker for Winners software. It's the only currently available software that provides bankroll calculations for Triple Play, Five Play, and Ten Play.
The amount you receive for dealt royal flushes doesn't affect the long-term return on a game. However, since you receive one of these only once every 650,000 initial hands (approximately), this has the effect of greatly increasing the number of hands it takes to reach the long term.
Whether a royal flush is dealt or not has no bearing on single-line games, of course, but significant bearing on multi-line games. This is the "long-run consideration" to variance that was mentioned earlier.
