Introducing Challenge Data Farming
Predict future ETH prices, for 5000 OCEAN in prizes weekly. Now part of Ocean DF
Contents
1. Abstract
2. Introduction
3. DF, Updated
4. Challenge DF Details
5. FAQ
6. Conclusion
7. Appendix: Updates1. Abstract
Ocean Protocol’s monthly “Predict-ETH” data challenges have become weekly “Challenge DF” challenges, as part of Ocean Data Farming (DF). Challenge DF begins in DF Round 48 (DF48). Each week has 5,000 OCEAN total in prizes towards those who predict the price of ETH with the lowest error. It’s part of Active DF; Passive DF is unchanged.
Predict ETH price accurately to win! Weekly competitions and prize $. The first challenge is live now. Get started here.
UPDATE Nov 1, 2023: Ocean Predictoor reduces the relevance of Challenge DF. Therefore, as of DF Round 62 (starts Nov 9, 2023) the prizes to Challenge DF will be 1,000 OCEAN total. And, DF Round 65 (starts Nov 23) will be the final round for Challenge DF. The Appendix has details.
2. Introduction
2.1 Background: Ocean Data Challenges
Ocean Data Challenges are data science competitions run by the Ocean core team with $thousands in prize money. Past competitions include Dubai real estate prediction, OCEAN Twitter sentiment analysis, Catalunya air pollution, and more. There’s a steady pipeline of future challenges. Most competitions are judged on a combination of subjective merit (e.g. data analysis) and objective merit (e.g. error).
2.2 Background: Predict-ETH Data Challenges
There’s great value in being able to predict ETH well. It helps to make $ in buying / selling ETH, DeFi trading, yield farming or DeFi protocol development. Such predictions — if accurate — would be of high interest to buy, e.g. as a datafeed within Ocean ecosystem.
Accordingly, we’ve encouraged the Ocean community to get good at predicting ETH. Specifically, within Ocean Data Challenges, we’ve held monthly “Predict-ETH” Challenges, running from October 2022 to July 2023 (7 total). Participants had to predict the ETH price into the near-term future: 5 min, 10 min, …, 60 min ahead. Reward $ goes to those with lowest prediction error. Error is a 100% objective measure.
2.3 Background: Ocean Data Farming
Ocean Data Farming (DF) is a rewards program that incentivizes for growth of Data Consume Volume (DCV) [1] in the Ocean ecosystem. DF is like DeFi liquidity mining, but tuned to drive DCV. DF emits OCEAN weekly for (a) passive rewards and (b) active rewards.
We launched DF in June 2022, and arrived at DF Main phase in March 2023.
2.4 DF Evolution
We’d set expectations that we may tune DF to include data competitions [2][3]. Now, that’s happening.
We’ve moved the ETH prediction competitions from Data Challenges framework to Ocean DF framework for two reasons. First, the competitions were 100% objective, and so could be automated to lower administration overhead; DF already had relevant weekly ops to support such a competition. Second, it sets up DF for further evolution related to predictions, stay tuned;)
3. DF, Updated
This section describes how DF is updated, in terms of structure, budgets, and how people can earn OCEAN in DF.
3.1 DF Structure, Updated
Challenge DF is a new substream of Active DF [4].
Here’s the updated structure of DF:
- Passive DF. Lock OCEAN for veOCEAN, for passive rewards. Reward depends on veOCEAN amount, which in turn depends on amount of OCEAN locked and for how long.
- Active DF. Substreams:
– Volume DF. Here, people curate data by pointing their veOCEAN towards data assets. Reward depends on asset DCV, veOCEAN amount, and more. (This is what was previously simply called Active DF; now it’s one substream of Active DF.)
– Challenge DF. Here, people predict future ETH price, and rewards are a function of prediction accuracy. (This is the new substream of active DF.)
3.2 DF Budgets, Updated
The DF budget change is mild: 5,000 OCEAN weekly that was going towards Volume DF, is now going towards Challenge DF. Everything else stays the same. Let’s elaborate.
We’re currently in DF Main 1, with 150,000 OCEAN per week in rewards. On launch, this gets allocated as follows.
- Passive DF: 75,000 OCEAN (50%)
- Active DF: 75,000 OCEAN (50%), allocated as:
– Volume DF substream: 70,000 OCEAN
– Challenge DF substream: 5,000 OCEAN
There is no change to how DF Main ratchets up OCEAN rewards over time. This post has details.
3.3 Earning DF Rewards, Updated
There are three ways to earn and claim rewards: passive DF (like before), Active DF : Volume DF (like before), and Active DF : Challenge DF (new).
- Passive DF. To earn: lock OCEAN for veOCEAN, via the DF webapp’s veOCEAN page. To claim: go to the DF Webapp’s Rewards page; within the “Passive Rewards” panel, click the “claim” button. Ocean docs has details.
- Active DF
– Volume DF substream. To earn: allocate veOCEAN towards data assets, via the DF webapp’s Volume DF page. To claim: go to the DF Webapp’s Rewards page; within the “Active Rewards” panel, click the “claim” button (it claims across all Active DF substreams at once). Ocean docs has details.
– Challenge DF substream. To earn: make ETH predictions for specific times; the next section gives an overview of how. To claim: go to the DF Webapp’s Rewards page; within the “Active Rewards” panel, click the “claim” button (it claims across all Active DF substreams at once). Challenge DF README has detailed instructions.
Expect further evolution in Active DF: additional substreams, tuning substreams, and budget adjustments among substreams. What remains constant is passive DF, and total OCEAN rewards emission schedule.
4. Challenge DF Details
4.1 Punchline
You can participate in Challenge DF by making ETH predictions for specific times. The Challenge DF README has detailed instructions [5]. We share key aspects here.
4.2 Key dates
Each Challenge DF competition is weekly, in line with the rest of DF.
- Submission deadline: Every Wednesday at 23:59 UTC when the Data Farming round finishes.
- Prediction at times: Every Thursday at 00:05 UTC, 00:10, …, 1:00 (12 predictions total).
Challenge DF competitions start now. Therefore the first submission deadline is Wednesday Aug 2, 2023 at 23:59 UTC, as part Data Farming Round 48 (DF48).
4.3 Criteria to Win
The winner = whoever has lowest prediction error (normalized mean-squared error, or NMSE).
To be eligible, competitors must produce the outcomes that the README guides. This includes:
- Created an Ocean data NFT
- Filled it with (encrypted) predictions
- Transferred it to the Ocean judges before the deadline
- And a bit more; the README has full info
4.4 Challenge DF Prizes
Weekly Prize Pool: 5,000 OCEAN
- 1st place: 2,500 OCEAN
- 2nd place: 1,500 OCEAN
- 3rd place: 1,000 OCEAN
4.5 Winners Reward Distribution
We will identify winners by the blockchain address they use in the competition (on Mumbai).
Rewards get distributed as part of DF rewards: we allocate OCEAN to winners’ accounts on Ethereum mainnet as part of Active Rewards contract.
If you are a competitor and want to know if you won, go to the DF Webapp “Challenge DF” section and see the listing of winners for each previous DF Round.
To claim your reward, go to DF Webapp “Rewards” page. Then, within the “Active Rewards” panel, click the “claim” button. This will claim rewards across all Active DF substreams at once.
5. FAQ
- Q: Data Farming is supposed to be about driving DCV. How does Challenge DF help?
- A: ETH price predictions — if accurate — make for price feeds that people will readily pay for to trade against and more. Challenge DF catalyzes creating such feeds, and has more streamlined ops than the previous Predict-ETH Data Challenges.
- Q: Are Challenge DF rewards a function of stake, e.g. veOCEAN held?
- A: No. Challenge DF rewards are only a function of accuracy. (For now.)
- Q: Predict-ETH had a blog post announcing each challenge, and another post announcing winners. Will you have the same here?
- A: Partially. There is already a weekly DF blog post that wraps up the previous DF round and gives parameters for the next DF round. To keep ops streamlined, the blog post will not list winners. If you had submitted to Challenge DF and want to know whether you won, go to the DF Webapp “Challenge DF” section and see the listing of winners for each previous DF Round. Your winnings are claimable as part of regular Active DF rewards.
- Q: Going from Predict-ETH Data Challenges to Challenge DF, does the prize pool stay the same?
- A: Approximately, yes. The changes are (a) weekly not monthly, and (b) denominated in OCEAN not USD. Let’s elaborate. Each monthly Predict-ETH competition had total 5000 USD worth of OCEAN worth of prizes. Dividing that by 4 (approx. 4 weeks in a month) gives us 1250 USD per week. Since DF is automated, it’s cleaner to calculate everything in OCEAN not USD. With current OCEAN price about 0.35 USD, then 1250 USD is 3571 OCEAN. We round up to 5000 OCEAN.
- Q: What’s the difference between “Predict-ETH” and “predict ETH”?
- A: “Predict-ETH” refers to the specific set of Ocean Data Challenges to predict the price of ETH. “predict ETH” is simply a verb combined with the “what”, around someone predicting the price of ETH; it’s not referring to any given product.
6. Conclusion
Challenge DF is a new substream of Active DF, which rewards 5,000 OCEAN total in prizes towards those who predict the price of ETH with the lowest error. Expect Active DF to evolve further, towards driving DCV in Ocean ecosystem. Passive DF and total OCEAN rewards emissions are unchanged and will remain so.
Predict ETH price accurately to win! Weekly competitions and prize $. Get started here.
7. Appendix: Updates to Challenge DF
(This section was added on Nov 1, 2023)
Challenge DF was always conceived as an onboarding mechanism to becoming a predictoor, ie someone who runs a bot predicting ETH and more within the Ocean Predictoor ecosystem. Since Predictoor was introduced, predictoor onboarding has greatly improved. It’s no longer worth doing Challenge DF first.
Therefore, since the launch of Challenge DF (DF48), there are two important updates.
- Reduced payouts. As of DF62, Challenge DF payouts are reduced from 5000 OCEAN / week to 1000 OCEAN / week. DF2 counting starts on Nov 9, 2023 and concludes on Nov 16, 2023. (Predictoor DF is introduced in DF62.)
- Wind-down. DF65 is the final round for Challenge DF. After that, the 1000 OCEAN / week will be redistributed equally to Volume DF and Predictoor DF, therefore each having 37500 OCEAN / week.
For further details, see the Ocean Data Farming Series post, which includes “Predictoor DF” posts and other DF posts, all in chronological order.
Notes
[1] Data Consume Volume (DCV) is the USD$-denominated amount spent to purchase data assets and consume them, for a given time period (e.g. one week)
[2] Expectation setting in the DF Main launch post, section 4.4 “DF Evolution”: “With each weekly cycle, the DF core team may tune the Reward Function or make other changes based on learnings. After DF Main is shipped, the DF core team will expand scope. Current plans are … DF Crunch Program — Kaggle-style data science competitions. Initially, there will be a weekly reward to the best predictor of OCEAN.”
[3] Expectation setting in “DF Main is Here” post, section 4.4: “Competition Data Farming. At some point, some of the Ocean-run competitions (e.g. Predict-ETH) can get streamlined & automated enough to put into weekly DF ops.”
[4] Where in DF should Challenge DF fit? One constraint, out of our commitment to the OCEAN community, was “don’t touch passive DF”, which has 50% of the weekly DF rewards budget. In contrast, Active DF was always meant to evolve, all towards driving DCV. This is why we made Challenge DF a new substream of Active DF.
[5] Challenge DF’s README may evolve week-to-week. Don’t expect this blog post to reflect those changes. Rather, treat the README as the “source of truth” for competition
