FWIW I'm with dooglus and mechs on this. As regards disclosure, I in the top ten investors at the moment.
I invested in just-dice for a long term stable investment - based on e.g. 5000 BTC wagered a day, 50 profit a day. For me, I would rather have a load of smaller bets slowly meeting the 1% expected profit rather than a load of variance. In my ideal investor world, the max bet would be tiny and thus the variance would be tiny and we would rely on bet volume, however the max bet is certainly a huge attraction to some gamblers, so we want it high enough to attract people to the site, but low enough to still maintain a constant stable profit over time.
High max bet means that investing is more like gambling - if I wanted to be watching nakowa bet 300 or 400 a pop, I would just be betting 10 BTC on 50/50 myself, but I don't want to do that - I recommend that those people that do prefer that 'investment strategy' go ahead and play instead of invest.
For those saying to simply 'reduce your investment if you don't like the risk' - that isn't the answer either - if I've invested a certain amount of BTC on just-dice then I am expecting to make a certain amount per day on average and that amount is directly proportional to the amount I've invested - reducing my investment reduces my profit but keeps my risk exactly the same.
So for me the question is, how many of the bets are actually > 80BTC - I would guess that the percentage is pretty small, it's probably a very small number of gamblers who exceed this limit, and without those people (pick a set of days over the past month or so when nakowa wasn't online), we're looking at maybe around 5000 BTC a day average bet - this still gives around a 35% annual return. I'm personally very happy with this. If I wanted to go high variance I would gamble!
So, the balance has to be struck here in making the max bet small enough that we don't scare gamblers away. For me, 0.25% is pretty good given the amount of investment, so I back mechs and dooglus in the recent change.
Another Option
On another option that hasn't been explored much, perhaps we should have an option where investors can choose to be based on amount wagered instead of amount won - at say a substantial reduced amount - example would be at 40% house 'risk cut'.
This is best explained with an example e.g. for an investor owning 1% of the house invested with current house edge of 1% - 5000 BTC gambled a day that's 50 profit or 0.5 BTC a day with pure winnings (as JD is now). With pure-wagered the penalty would be, say, 40% cut so it would be 5000 BTC a day, 50 expected profit, 1% own of the house but 40% cut so 0.6 / 100 * 50 = only 0.3 but that value is guaranteed.
To make it more elaborate, make it so there is a sliding scale of 0 to 100 on how much risk you want to take. If you chose to go 'winnings based' (this is how JD is now) then you take the risk of win and loss as people win and lose. If you choose to go 'wagered based' then you take the risk of not profiting as much when people win, but a lower but more stable income. To make the math work, the excess profit from those people who are on 'wagered based' when the house is up goes into a pot, and that pot is used to pay these people out when the house loses. I will have to think about the math a bit more to make certain it all adds up.
Keen observers will notice that 'wagered based' is very much like the model for letsdice.com - but with a lower (1%) house edge.
Will