To answer Jeremy—David Hammett was hired on to the Greed crew to figure out how best to have reasonable odds for the top shelf questions.
Sort of. Greed was the only show I worked with that I saw from development to wrap party. My initial contribution was helping them determine the relative difficulty of questions with multiple answers; whereas Millionaire featured a 1 in 4 chance of being right just by guessing, the 4 out of 6 question on Greed was 1 in 15. (Being good at math was an asset for getting involved with Dick Clark Productions.) Later, I was at each taping in front of a large dry-erase board with a camera trained on it to project to Chuck Woolery what each player's stake in the game was; I'd update the totals as the game progressed. (The monitor he saw was located between the middle two contestants; you can sometimes see a small portion of it during wide shots.)
My later work with shows concentrated more on average payouts, although in some instances I'd have to point out that averages didn't tell the whole story... wide swings in the payouts, especially for a short series, might also be important. Occasionally my observations would lead to changes during the development of a show. For example, Russian Roulette eventually understood the first questions needed to be $150 instead of $100 because the scores might need to be divided three ways at some point. On another show (that did not get picked up), I had to stop them from introducing a game element that might have them dividing a player's total by zero. And so it goes.
