## INTERESTING: “Deal or No Deal” strategy

Did you ever see “Deal or No Deal”?

As an IT Architecture / BPR kinda guy, I think you have to have a strategy to get through the game. (Hey, I play penny slots too. No genius here.)

The only way to get a MILLION is to have two MILLION cases at the end.

Strategy is critical to escape with a prize.

I think that can apply some elementary statistics or probability theory.

You can predict the “banker’s offer” approximately and that leads you to a process to play.

It seems to me that you can compare the offer based on picking a low or high case. Then you make the decision from that.

Suppose, for example, at any point in the game, the distribution of cases can tell you what to do relative to the offer.

The offer is “fair” if EV = ∑ (p sub i * V sub i) is equal to the Banker’s Offer.

When receiving the Banker’s Offer, one can calculate the EV (ALL) of the current set of cases, the EV (HIGH) dropping the highest case, and the EV (LOW) dropping the lowest case. In a truly random game, you can calculate the potential loss EV (LOW) – EV (ALL) . And the potential gain EV (HIGH) – EV (ALL).

The offer can then be subject to the same test.

IF there is only ONE high value, the OFFER is usually cut in half.

By comparing the potential gain versus the potential loss, that determines if the contestant should take the offer or not.

You can’t pick cases, (although 26 seems to have a MILLION more than probability would dictate), but you can have a money management strategy.

# # # # #