I ran a simulation to see how many of each denom would be required if a 2.5X BB bet after the flop was called. Here are the results.
This wasnt supposed to be scientific, but just a spreadsheet to give me some idea of WTF is going on.
What was really interesting to me were the ratios. You can see below, the number of chips needed in their ratio format was 12/12/5/19, which seems inline with what most folks here recommend being a 12-12-5-6 starting stack.
However, the $1,000 chips seem seriously lacking in the starting stack breakdown. I've never actually tested this in theory, but maybe some of the other guys can chip in (excuse the pun)?
Sure, maybe there are 1/2 the amount of players around the table once you get to the 400/800 blind level, so doubling the amount of starting $1k chips per person (so going from a 5 to 10 ratio in the original breakdown), and you also need to account for the color-ups (but this only adds 5 X $1k chips, you use $5000 chips for the rest), but that 10-11 ratio still seems way short of the 19 ratio I came up with. My gut says I need more $1k chips.
One way I was thinking about solving this issue was to make the rebuys 10 X $1000. But I'm not sure how many of these I should do vs 2 X $5k.
Note: 33 ratio $5k chips arent required because from what you guys have told me, the game normally ends are 2000/4000, so the $5 chip ratio would actually be 5 (so 12-12-5-19-5).
Thought I would mention this Re: the amount of T1,000 chips you may need.