Continuing the discussion from Scrypt to Cluster data:
As noted for the system messy, I’ve had the value change on me from
Cracking the system: §280
Beating the system creator’s time: §35
Cracking the system: §264
Beating the system creator’s time: §33 (2nd solution)
I’m pretty sure I’m not wrong, because I can (I already have) take a screenshot of my current script and it is:
I started with 26,500 and completed the 6 systems noted in my linked forum post.
Plus it doesn’t seem to be rewarding me less than average ping while I only used 3 packets, and I’m pretty sure (90%+) that’s the minimum.
These are intended behaviors.
Reward values depend on other solutions to the same system, so they will vary over time (but eventually stabilize as the system is solved more).
The “fewer than average pings” reward is a “less than,” not a “less than or equal to.” You only get it if you beat the average, which means it may not be available for all systems.
For more OOC information on how rewards work and what our reasoning is, see our blog post on the topic.
So… systems are basically a race to see who can beat the creator first for the most scrypt.
Or perhaps @percent20 should create bots that artificially inflate a systems value.
We totally understand any concerns about manipulation of this rewards system and any consequences that are damaging to players’ experiences. We take those concerns seriously.
The reward system will probably be tweaked, balanced, and adjusted as we see how it plays out in practice. It’s also designed for high-volume play; there will be some odd artifacts of smallness until we have more players.
We want the amount of bonus scryp you get for extra tasks to be eclipsed by the reward you get for completing a new system. We may need to adjust this proportion to cause that effect.
We can’t make a perfect guess at the difficulty of a system in an automated fashion. Whether or not solving a system early is better depends on the efficiency of other early solutions, including the creators. Rewards can therefore increase or decrease over time.
Our goal for this whole reward system is for folks to make and play more of each others’ systems. We’re not particularly trying to encourage hyper-optimization of rewards, although we’re reluctant to tell you how you should play.
However, deliberately manipulating system rewards in ways other than through the design of a system isn’t cool. Using a bot to artificially inflate or deflate a system’s value is something we would notice in our stats and take steps to stop.
I’d just like to continue this by adding that over the first two weeks of recording scryp, I’ve noticed that system rewards sometimes (for around 25% of systems) decrease 10-25% over the first few times they are solved. Seems like players that log-in often will get an additional 2-5% scryp over players that login once or twice a week.
Of course, there are still more systems I expect will get hit with a downgrade cough
cough, but these are just preliminary