<figure><imgsrc="../.gitbook/assets/gif/avatar-plugin.gif"alt=""><figcaption><p>Make your selection and commit. Good things will come.</p></figcaption></figure>
- Only allocate to assets that actually qualify for DF. The "Assets that qualify" section below has details.
- Allocate to assets with high DCV. Predictoor data feeds are a good bet.
- Lock more OCEAN.
- Lock OCEAN for longer. For example, a 4-year lock is 4x more reward than 1-year.
- Refresh your OCEAN locks often. This is needed since veOCEAN degrades with time.
- Frequently claim your OCEAN rewards then lock it. Be careful: too often may make gas fees > OCEAN rewards.
- Publish your own assets; you get 2x rewards for such assets. [Here are details](https://blog.oceanprotocol.com/data-farming-publisher-rewards-f2639525e508).
You can delegate your veOCEAN from one wallet to another. This is handy if you have a hardware wallet holding veOCEAN, and you want a "hot" software wallet to automatically re-allocate more often. There are other uses too.
The asset may be of any type — dataset, data feed, a C2D algorithm, or other. It may be fixed price or free. If fixed price, any token of exchange is alright (OCEAN, H2O, USDC, etc). If free, then gas used for consume is the "volume".
* Have been created by Ocean smart contracts [deployed](https://github.com/oceanprotocol/contracts/blob/main/addresses/address.json) by OPF to [production networks](../discover/networks/README.md)
1. Distribute OCEAN across each asset **based on rank**: highest-DCV asset gets the most OCEAN, second-highest-DCV gets second-most, etc. [Here are details](https://blog.oceanprotocol.com/data-farming-df22-completed-df23-started-reward-function-tuned-ffd4359657ee).
1. For each asset and each veOCEAN holder:
1. If the holder is a publisher, 2x the effective allocation
1. Baseline rewards = (% allocation in asset) \* (OCEAN for an asset)
1. Bound rewards by asset’s `DCV` \* `DCV_multiplier`.
Step 2.4 is to prevent wash consume. `DCV_multiplier` is a function of fees, with a value of 0.201 for Predictoor feeds, and 0.001 for other data assets.
When DF restarts each Thursday, the global allocation towards each asset resets back to zero and then starts counting up again until the end of the round. After allocating your veOCEAN to different datasets, it will take time for your voting power to be accounted for.