Whilt I agree that combining bets leverages the effect of the overround book to the punter's disadvantage, I think you've underestimated the size of the phenomenon.
For example, take two snooker matches: In match 1, between players A and B, both have an equal chance of winning. The same in Match 2 between players C and D. In a 'fair' book, you'd be offered evens for each. In reality, there'd be an overround book, meaning you'd probably get, say, 5/6 on any of the players winning their respective matches. This would give the bookie a total book for each of the matches of 109.09% (100 × (6⁄11 + 6⁄11)) i.e. 9.09% overround.
The combined outcome of those two matches could see either of the following pairs winning: AC, AD, BC, BD. Since all four are equally likely, the true odds of any one would be 1/4 (and 3/1 against), meaning a 'fair' price of 4/1. That would result in a £10 winning £40 (10 x (3/1 + 1)). Whereas, the actual return in the overround book (using the figures in the paragraph above), would be £33.61 i.e. 10 × (5/6 + 1) × (5/6 + 1). That represents odds of 2.361/1 i.e. 29.752% (100/3.3611), which when applied to all four outcomes gives the bookie a book of 1.1901 i.e. an overround of 19.01% i.e. his edge has more than doubled, compare to each of the singles!
On a double, the overround (0) can be calculated as 0 = B1 × B2 × 100 − 100, where B1 and B2 is the whole of each respective book e.g O = 1.0909 × 1.0909 × 100 − 100 = 19.01%. Obviously, the more in the accumulator, the greater the book gets overround, calculated for a bet of y selections as B1 × B2 × ... × By × 100 − 100.
To do as the OP hopes i.e. to turn £20 into £150 with 10/11 prices (at the longest), would take a minimum of 20 bets. If we assume a reasonable overround of 10% on each, the combined overound accross the whole is massive: something like 1.1 x 1.1 x 1.1 x 1.1 x 1.1 x 1.1 x 1.1 x 1.1 x 1.1 x 1.1 x 1.1 x 1.1 x 1.1 x 1.1 x 1.1 x 1.1 x 1.1 x 1.1 x 1.1 x 1.1 x 100 -100 = 640% i.e a total book of 740%. No wonder bookies love it!
