The wot solution is that there is no singular metric for validation. Yes each can be gamed, but wot developers also have to accomidate the current gamable metrics while also forgetting about outdated (bevause of gameability) metrics. The recycling multiply-coexisting and changing algorithms is what makes the system robust.
Discussion
The core problem is that
1. Bootstrapping is hard
2. Responding to reputation abuse is not real time
I'll agree timescale response is something to be aware of, but that would also mean that one should be highly suspicious of any account's rate of score acquisition.
Account aging is a thing. Traditional social media platforms already try and somewhat fail to solve this, with Nostr you have less signals to act on and slower response since it is decentralized.
Got me there, but my core point still stands. You can have multiple wot algorithms with contradictory metrics. Users can decide which they want at a given time. Nothing beats irl with trust, and then general perceived interaction history as a second. I think it would be interesting to have many users all engaging with rotating differing wot metrics over time. You'll likely still have your 1 or 2 degrees of separation, while the users on the periphery of your network will recycle depending on which metric your using.
If you make a map as the sole point of truth, it can also be used by anyone to traverse in. Would be an interesting experiment to see how gamable it is to have rotating & contextual maps. The users don't care how objective the wot is anyway, so long as it effectively filters their experience.
The map does not react fast enough to abuse.
Is it not possible to separate the different components of a monolithic wot into multiple ones? Of course, some get outdated but if there are 20 different metrics, often in conflict or deliberately worsened - you'll have something quite difficult to game.
The problem is what differentiates a real user from a bot. No WoT algorithm can solve for the lack of that difference if it exists.
i'm not sure if this is really true. wot algos can factor in many things like zaps and how often you interact with them. yes its not perfect but it can be pretty good.
You definitely can't filter out all bots. Some users even explicitly care to interact/follow bots. You can, however, filter on credibility. But like irl social credibility, there are different interactions you can work with to establish it. It can be context dependent, but i think even some sort of random rotation between them would be effective. I don't have evidence, for these claims but I think it would be something interesting to play around with, especially in a space where so many are searching for "the best wot algorithm"
I agree with this, value is subjective, so if computer is helping with discretion (or rather, keeping track of past discretionary decisions), it is best when it reflects real life.