Why we don't display native chain TVL under the main list

[0xngmi]

Initially we counted matic staking for polygon and the tokens locked on the polygon bridge towards polygon's TVL, along with other bridges (eg: xdai) or staking protocols (eg: stacks). However someone pointed out that this was creating unfair comparisons between chains because:

• We were counting the tokens that were bridged from ethereum to polygon as polygon tvl, but we weren't counting the binance-pegged tokens on bsc or the tokens that have been created on ethereum (which could count towards ethereum tvl)
• We counted matic staking but not eth2 staking or staking on other networks

To solve this unfairness we had to make the decision of whether to count everything (although choosing how to count native ethereum tokens would be hard) or not count anything.

We asked our community and the most common answer we got is that most people wanted to use defillama to check how much defi adoption there is on a chain, so they wanted to see the TVL of the applications built on the chain, rather than how much money is staked.

Due to this, we removed native staking and bridges from the total TVL charts, while keeping the projects on the list. However, people kept getting confused about the polygon listing (which tracked bridged tvl and staked matic) and the tvl chart, which tracked the TVL of all the defi dapps on polygon, so we removed the listing from the main list, and along with it we removed all the other similar ones like STX to keep it fair.

After removing these from the main list, we moved these to a different "Chain" section, where you could see all these chain-native listings. However, users kept getting confused so we removed that as well. Here I ran out of ideas on how to display these while preventing people from getting confused, so we haven't been added them to any other list. You can still find them by searching for them or through their URL.

This process was very public, most of the discussions happened on twitter and no decision was made on closed group chats.

[0xngmi]

Apart from the points mentioned before, there are also the following issues when counting TVL on bridges:

• There are inconsistencies between bridged tokens and tokens issued on the chain. For example USDT is directly issued on Ethereum and Tron, but on chains like avax or arbitrum it is bridged. If we count bridged tokens we'd be counting USDT on some chains but not in others.
• Most bridges only work one way, you can port tokens from Ethereum to the other chain but you can't port tokens issued on the other chain to Ethereum. Because of this some tokens such as QUICK chose to issue all their supply on Ethereum and then port it to Polygon to use it there. For all intents and purposes QUICK is a Polygon token, however it will show up as bridged. This creates inconsistencies since we'd count this as bridged but another token that has been directly issued on the chain (eg: JOE) wouldn't be counted.
• It's very easy to end up incorrectly counting unissued tokens. In the case of QUICK the whole supply was bridged but part of the supply remained locked and waiting to be dispersed later, clearly these tokens don't exist yet, however they are being counted as tvl. It's also possible to easily game bridged TVL by creating a token with a large FDV and bridging all of it. Keeping this in check is very hard and would require us to police tokens.