I'm not a fan of splitting the first round into two parts. It means that, really, the randomness of it all it lost - you get the easiest five questions first, then the harder five. There's not a whole lot of difficulty variation there, at least, not close to having all ten questions clumped together.
Of course, you pointed out the problem with having a safe haven in the first round. With the random difficulty and values, you could have someone with as much as $31,000 (post-halving) after five questions, or as little as $300. The three-question safe haven cuts the maximum to $25,000, but also the minimum to $50.
I took a look at the past ten contestants who have not lost or walked by Q5 at the WWTBAM Bored, and on average, contestants have a bank of $24,210 after question 5, and $16,510 after question 3. After halving, that's $12,105 and $8,255. You could easily tack on a $5,000 safe haven after questions 3 or 5, but here's the problem with that: the banks vary
wildly, especially at question three. There, the halved banks ranged from $1,550 to $16,000, with four being under $5,000, two from $5,000-$9,999, and four at $10,000 or over. At question five, there is much less variation: seven were from $14,000-$18,500. Still, the other three were from $2,050 to $6,250.
I'm going to outline some various scenarios with different mid-round safe havens now, pointing out things that are good and things that are bad.
Scenario 1: $5,000 safe haven at question five.
It sounds okay. At five questions, however, your halved bank may be as little as $800. In fact, if you have the minimum possible bank for each question, you *could* (and I stress could) go EIGHT questions without taking a single risk - this would require jumping high amounts towards the beginning and getting lower amounts from there on. In that case of eight questions, which would require getting as little money as possible, you would have $5,800 (halved) after that, only risking $800 for the next question. Now, it's very unlikely this would happen, but less extreme versions would pop up more often, making for bad TV when for half the show the contestant takes a free guess at every single question. So that one's out for me.
By the way, a fixed-money scenario at question three would do the same thing. And you could always lower the safe haven value, but then you're not playing for much more than the guaranteed $1,000, and the idea becomes pointless.
Scenario 2: A safe haven worth half your bank at question three.
Jump two questions, get anything less than $3,000 on the remaining question in that block of three, and it doesn't matter. So that's not going to work.
Scenario 3: Same thing as scenario two, but at question five.
You could still have as little as $800 if you jump two of the first five questions. Other laughably low-valued scenarios include $1,300, $1,750, and $1,800, plus many, many more.
Now, how do we fix these? Simple: Bring up the value of the lower amounts, in order to make sure that you can't still have peanuts by the time of the proposed safe haven. I think this tree would work fine (these amounts are, of course, not halved):
$1,000, $1,500, $2,000, $2,500, $3,000, $4,000, $5,000, $10,000, $15,000, $25,000
That totals up to $69,000, only a $400 increase over the previous tree. At five questions, you'd have to have at
least $2,250 with half your bank, so a safe haven could work there with that tree.
(I'll try and take up less space in this topic in the future.
)
) I guess the next on will be Poland then...
- in Season 9, versus 1.75 and 1.43 in seasons 8 and 7, respectively. The reason for so many contestants, when the clock should have theoretically put more in the hot seat? They're playing less questions. This season has seen the lowest question count out of any season of syndicated Millionaire, at 2,206. By comparison, the seasons with the Clock format had an average of around 2,706.5, and seasons 1-6 had about 3,304.
)