3b. Now I might be gone. It's a 1/8 chance of £240,000, a 1/8 chance of £3,750, and 3/8 chances of £60k and £15k. EV of that is (£240k + £180k + £45k + £3,750)/8 = ~£57k. That's in the territory I want on blue/£250k, and the uncertainty effect doesn't take too much off when there's a 7/8 chance of a five-figure amount at least. I'd just about take that, I think...

4. £22k or a 1/4 chance of £250k. Is £250k four times as useful as £22k? Let's work backwards. If £55k or so is what I consider half as good as £250k, then for me to prefer the shot at £250k over the certain £22k, I'd have to be uncertain at £22k on a blue/£55k offer, and that I would not be. No Deal to the gamble; if this came up with the Alice offer, I would consider it (and would also consider it actually less bad than what actually happened in Alice's game, as at least it creates a bonus agonising decision - the only hard decision Alice faced was the swap).

5. By further induction, to ascertain that £250k is eight times as useful as £10k, I'd need to consider No Dealing £10k if it was the offer against the amount I would take in 4. This amount isn't even double the £10k in the first place, so by induction I am not taking this gamble on either.

My tipping points based on this logic: £16,500, £6,500. The latter seems a touch low, but only just, and the logic is clear; if £a is my tipping point on blue/£250k, £b is my tipping point on blue/£a, and £c is my tipping point on blue/£b, and £x is half as useful as £y when £x is my tipping point on blue/£y, £b confers 25% of the utility of £250k, and £c confers 12.5% of the utility of £250k. Hence, £b is my tipping point on gamble 4, and £c is my tipping point on gamble 5.