Yep so if we assume voting power is a good proxy and a determining factor for validator voting power then a lot of the trends we inferred from the previous graphs like Orion getting higher rewards because of higher voting power and Hashed losing a lot of rewards due to the calls for decentralization (this graph was also an excuse to use the new table before finishing off this dashboard!).

So the biggest change I see from week-0 to week-7 is the rewards for hashed have gone down significantly and the rewards for Orion have gone up a lot. The rest of the validators in the top 15 look relatively bunched up with a more even distribution (probably redistribution from the higher peaks in the early weeks) but no real change otherwise.

So I think through our graphs we've seen some patterns emerge from the different plots we've made let's summarise them to conclude our analysis:

• Terra's top 15 validators haven't seen that much churn in the last 2 months. Some validators have moved up and some down but the movements haven't been significant.
• The re-shuffling amongst the top 5 was due to Do's call for decentralisation incentivized via larger airdrops for delegators who delegate to smaller validators.
• Orion made a large jump due to their sought out product and users wanting extra tokens by staking to them.
• Hashed was the biggest loser over this period with its reward yield going down heavily over the last couple of weeks.
• There are some large gaps between validator voting power (like between 4 and 5 or 6 and 7) making it hard for the system to be truly decentralised. It seems like many delegators prefer to delegate to validators with high-quality service to prevent slashing, and also prefer to not be agile with validator choice to maximise for validators.
• Some delegators do care about decentralisation and/or maximising airdrop yields, their proactivity has led to rewards for validators from 5 to 15 becoming more even over the last 2 months.

Lastly, let's see the voting power graph over the last two months for various validators to see if our inferences align with a change in voting power (which partially guides the rankings for validators):

Okay, so the first graph shows us that the top 5 validators haven't seen much churn in the last two months apart from some reshuffling. With 4 validators being at rank 2 providing the most variety at any rank in the top 5 when the reward is averaged across weekly data. It seems like there is more competition for 6 to 11 with 4-5 validators holding those positions very commonly.

The next graph shows us the change in ranks for various validators in the top 15 (a rank of 0 indicates a position outside the top 15 in a given week). We want to look for lines that slope downwards: like Orion which went from 12th in Week-0 to 2nd in Week-8. The biggest losers seem to b e Hashed which went from 2nd to 5th and Mr.K which went from 6th to 13th.

The last weekly graph we'll look at is rewards over different weekly time-periods with colors representing the validator:

The graphs above show the top 15 or so validators in two different formats: the first being the top 15 grouped by their month (i.e. using the group by as x-axis) and the other grouped by ranks (the x-axis is the rank, y-axis the reward and the colors represent the validator in the graph).

The second graph gives us the ability to look at the movements of each bar (denominated by their color) in terms of their rankings. It also helps us understand the diversity around each rank to see how much churn there is for a given rank.

The only validates to stay on the same rank (represented by a singular bar corresponding to the x axis) are the ones on 5 and 6 are Staked and DSRV. It also seems like on either side of 5 and 6 there has been a ton of musical chairs over the last two months amongst those on each side. This is because there is a significant gap between (4 and 5) and (6 and 7) in terms of rewards (making it harder for any validator to jump from one side of 5/6 to the other).

Next, let's look at the validator rankings distributed over the last two months in chunks of weeks:

There seems to have been some shakeup in the top 10 validators on Terra. Mosaic and Certus One moved out of the top 10 (positions 8 and 10 respectively to positions 11 and 12 respectively). Orion Money and Chorus One took their positions in the top 10. There was a larger shakeup within the top 5 with hashed moving to 2nd from 1st, B-Harvest from 2nd to 3rd and Staking Fund moving from 3rd to 1st. Orion Moved from 13th to 4th as well due to them rewarding delegators with extra airdrops of their tokens. Within 6 to 10 Dokia Capital went from 9th to 7th, Mr.K from 7th to 9th and Terran One being the largest loser with a loss of 4 positions (from 4th to 8th).

This was all probably largely spurred by a call from Do Kwon to decentralise the Terra validator set. We know that he said the delegators to the top 5 validators would not get $NEB airdrops. The top 5 is where we see most of the flux in rewards, but the overall rewards distribution over the last two months sadly doesn't seem different enough to signal meaningfully higher decentralisation.

Let's also visualise all the information in the two paragraphs above into a plot so that we can view it easily:

Let's start off by clearly defining that we'll be looking at the top 15 Terra validators (since we want to focus on all contenders for the top 10) for the last two months on a monthly basis to see if the top 10 has changed in the last month. Moreover, we drill down into the rankings and look at the top 5 validators along with doing more drilled-down weekly analysis on validator rankings.

(NB:

• A validator's ranking is based on the amount of rewards they received i.e. the top validator on any given day receives the most rewards in UST terms

• We look at rewards in LUNA terms using the end_block tag in the transitions table to begin with , the denomination of rewards doesn't really matter since our goal is to compare the validators instead of show their absolute rewards on some sort of scale.

)

Let's start off by listing out the top 10 validators and their rewards for the month before last and last month: