- Heavy growth in the beginning of the 2nd week.
- More steady vs. previous epochs (no sybil boost)
- .
Analysis of Epoch 5 metrics and user behavior
| epoch | budget | budget_allocated | donations | leftover (%) | generosity (%) |
|---|---|---|---|---|---|
| 1 | 101.469 | 75.913 | 5.527 | 25.2 | 7.3 |
| 2 | 138.233 | 137.941 | 3.925 | 0.2 | 2.8 |
| 3 | 241.301 | 239.465 | 25.484 | 0.8 | 10.6 |
| 4 | 214.744 | 203.187 | 4.788 | 5.4 | 2.4 |
| 5 | 220.261 | 205.297 | 17.630 | 6.8 | 8.6 |
Note: patron mode not included.
| epoch | potential_users | allocators | donors | ga_beneficiaries | turnout (%) |
|---|---|---|---|---|---|
| 1 | 515 | 362 | 347 | 15 | 70.3 |
| 2 | 586 | 366 | 332 | 14 | 62.5 |
| 3 | 603 | 320 | 285 | 15 | 53.1 |
| 4 | 593 | 303 | 279 | 16 | 51.1 |
| 5 | 898 | 439 | 422 | 232 | 48.9 |
Note: patron mode not included.
Lower turnout % vs Epoch 4 is slightly deceiving. If the 60+ sybil accounts made their allocation, we would have ended up with higher turnout this Epoch.
Users rarely change their allocation decision (more stability of results due to QF, less heavy-weight testing by team members). Growth vs Epoch 4 mostly due to the sybil addresses pulling out (60+ addresses with no allocation changes).
As usual donations (in absolute terms) are dominated by the whale category, which is by far the most important contributor to the overall generosity rate. Allocation to budget ratio is relatively stable in other categories, but we can see some growth in the M category.
Non-participating users are over represented in the XS category (which is a good thing - syblis?). Note that the vast majority of users fall into to full donor category (>98% of budget allocated to projects) and that the mixed category (2%-98% of budget allocated to projects) are quite an important part of the middle class users.
The falling average number of projects supported can be attributed to the introduction of GLM marketing campaigns. “Old” users keep the average at a level similar to Epoch 4, while users who received GLM giveaways in Epoch 5 tend to concentrate their votes more (less than 4 projects on average).
If we applied linear distribution, only 7 projects would receive funding, with ca. 90% of the MF being captured by 5 projects supposed by the single whale. Note that Protocol Guild did not receive the highest amount of donations and yet managed to be at the top of the leaderboard when we applied QF.
Interesting stuff: 1) lower variation in the number of donors, 2) similar total donation amounts lead to entirely different shared in Matching Funding (due to QF).