Here is are the numbers for the distribution:
Contestant | Dev % | Testing % | Total BTC | @ $800 per BTC |
Total | 100% | 100% | 150 | $120000 |
As the mastercoin protocol and mastercoin foundation is all about transparency and consensus, I think also the judging should follow that approach.
Please consider writing the algorithm that you use (words are enough, but for sure code is preferred), and all participants could put the evaluations openly online - then the calculation will be totally transparent.
I'll give it a shot. For this contest (and the last one), I followed the following steps:
- Normalize all feedback to fix those which don't add to 100%
- Attempt to determine how careful each person's feedback was (i.e. the same % for everyone is not very careful, but different percentages for each person with detailed comments on how you chose those percentages is very careful)
- Weight each person's feedback by how careful they were
- Get a percentage for each person based on the combined weighted feedback
- Perform a second pass through weighting step, changing each person's grade for themselves from zero to the result from the first pass
That last step fixes a problem where if you do a fantastic job grading other people you would have gotten slightly penalized for it, since your feedback doesn't give anything to yourself but is being counted heavily toward the final score.
I'm sure everybody would have their own way of doing this, but I hope this is close to fair!
I hope to do these payments tomorrow in BTC. I have a lot of figuring to do before I can do the MSC-payments across all bounties to date, but that will come next.
dacoinminster - thank you for the hard calculation work, and thank everyone else for the mutual efforts behind those bounties.
I still think that automated calculation (code) could really improve the process - it would be more clear how things are done, and it will save the bounties maintainer hard work for each of the next distributed bounties.
As for transparency - I would rather seeing how each person evaluates me, and show how I evaluate them. This only makes things more efficient and direct. It would be another step away from centralization.
In the future we could even use mastercoin protocol to automate the evaluation & calculation. Every participant could send coins to other participants (these can be some evaluation coins), and then the monthly evaluation code could gather these numbers and give a clear cut updated result.
Regarding the MSC-payments - please avoid sending many class A payments directly from exodus. It would be better to send all monthly MSC payments to some external address, and from that address to pay using multiple class B transactions to all. Let's take care of the utxo space.