Qtum's PoS vs CASPER staking

source here https://www.reddit.com/r/Qtum/comments/788oa5/qtums_pos_vs_casper_and_the_nothingatstake_problem/

Qtum's PoS implementation and CASPER serve the same basic purpose and have the same overall goals, but their implementation and design is completely different. First, about CASPER:

• CASPER uses smart contracts in order to track the stakes and funds created.
• CASPER requires a minimum amount of coins to participate in staking
• CASPER requires making a "deposit" and locking your coins for a set amount of time in order to participate in staking (I believe this is something like 30 days)
• CASPER provides a partial solution to the nothing-at-stake problem by punishing stakers who misbehave (ie, they lose their deposit)
• CASPER's full nothing-at-stake solution is both the punishment feature, and users choosing off-chain centralized servers to determine which blockchain is legitimate in the case of multiple.

And now, compared to Qtum:

• Qtum uses basic Bitcoin Script opcodes and transactions to track stakes and funds created - We believe this is better just because it has been tested for far longer. It is slightly less flexible, but ultimately encourages decentralization and stability.
• Qtum does not require a minimum amount of coins to participate in staking. 1 coin can help secure the network the same way 1 million coins can, though 1 million coins will of course find more blocks, etc
• Qtum does not require locking your coins. It only requires that the coins mature for about 20 hours before they can be used for staking. Afterwards, the coins can be sent very quickly in case of network issues, private key compromise, etc. (though immediately after your coins have staked, you must wait 20 hours before they can be spent)
• Qtum plans to provide a different solution to the nothing-at-stake problem and will never punish stakers that misbehave other than kicking them off the network. We believe no software can be made perfect, and so the risk of a bug causing the staker to misbehave and lose millions of dollars is unacceptable to us
• Qtum's nothing-at-stake solution will be implemented in 2018 and will be an optional consensus feature (ie, no fork etc required). Basically, we will put some data on the Bitcoin blockchain that can be later verified through SPV. This makes it so that it is impossible for an attacker to lie about when a particular block was created, which is the biggest vector used to construct a nothing-at-stake attack setup.

Beyond that, we are confident that the nothing at stake problem is pretty overblown, at least with the way our consensus system works. It's incredibly expensive to construct the setup for the attack, and when it is all setup, the only nodes you can really convince to follow it are ones synchronizing to the network the first time... and it's incredibly easy to detect by checking any centralized off-chain resource such as block explorers. It is also very easy for us to add checkpoints etc to the code at a more frequent pace and do faster releases in order to ensure that nothing-at-stake chains never have any significant amount of power.

• many comments

Casper: what is the minimum amount to stake
https://www.reddit.com/r/ethereum/comments/67edhh/casper_what_is_the_minimum_amount_to_stake/
https://ethereum.stackexchange.com/questions/11010/will-staking-ether-work-with-any-amount