... and we will be fining fairly. >:D

Initial Replication == 1 allows us to shows our best and worst sides, but option # 1 is preferable due to:
1. Relationships to other participants;
2. There is a bear's paw for every cunning fox;
3. My script is much more might.

43, ...

So someone was cheating Sidock credits to win some silly competition?

So someone was cheating Sidock credits to win some silly competition?

Looks that someone tried to convert competition to silly thing.
A ridiculous story comes to mind in such cases:

Bear, wolf, fox and hare are playing cards.
A bear says: "Don't cheat! Whoever cheats will get a kick by brazen auburn muzzle!"

Good luck!

So someone was cheating Sidock credits to win some silly competition?

Not just sidock and not just this competition.

So someone was cheating Sidock credits to win some silly competition?

Not just sidock and not just this competition.

What other project and what other competition?

I read there is adaptive replication to deal with these cases. https://github.com/BOINC/boinc/wiki/Adaptive-Replication

I don't know if any of the number there can be tuned though, like how many consecutive tasks a host needs to be trusted. If the validator can also periodically flag results from trusted hosts to double check, it might just be enough to automatically weed out cheaters. At least their cheating is going to be harder.

I read there is adaptive replication to deal with these cases. https://github.com/BOINC/boinc/wiki/Adaptive-Replication

It is a good idea. Values from template can be overrided for some fraction of workunits during tasks generation.