Most active partiicpants of twitter chat announced via Kneaver use to be counted by tweets. All kind of tweets, retweets being taken into account. This mean that someone automatically retweeting everything would be the top participant. It happend once. But also we need to raise the challenge. I call this new measure “Impact” before I get a better name. The idea is to identify people bringing consistently good answers as well as contributing to the ambiance independantly of their influence beforehand (ie of their klout score). At the end of the day this should stay an amusement reusing Kneaver’s real time analyzer. This is how it’s implemented today but it’s open to changes. +10 If the tweet is a direct response to a question, it should start with Axx or Qxx +5 If the tweet is a reply to a direct response as above +2 for other tweets on top of above tweets you get bonuses: +1 for each retweets you get +1 for a retweet you do That’s the state today. For the future: I wish I could track replies to replies to replies but it becomes very confusing quikcly I was tempted not to count retweets at all but this would discourage people to build on top of theirs answers which is not the idea of a chat. I was also tempted to not take into account the retweets some people are doing of the questions themselves but again they are bringing value since they retweet it to their own followers who might not be aware of the chat and doing so they bring more participants in the git. In this direction we could add some bonuses to people who have followers in the chat. This would give an advantage to tweeps with tons of followers hence creating a barrier to new entrants. Again not something we wish to encourage. I’m also able to analyze manual retweets more subtly by identifying what was added before or after the tweet. RT, +1, +1000 would lead to different scores but also being able to identify people who add value while retweeting. I’m not the only one doing that. I suggest that more than 2 words nefore or after the RT should be considered as an answer on its own. I could also put a bonus on the first answer to a question or if this first answer comes very quickly after the question as been issued. This could brings some excitement. Why not also adding bonus to submission of pictures, videos, gif. May be this is more something the host could adjust. Since we also have the semantic analysis being done I imagine we could gibe a bonus to first person introducing the term becoming the most used term of the chat. Finally we could take into account bonus given at host’own discretion by emitting a tweet with #bonus hashtag + mention of the participant. Any suggestion is welcome. For me it’s a proof a concept, I’m not doing large development for that but surely I take the occasion to make it more robust. The same could be used on facebook, yammer, google+ etc possibly concurrently. NB: I took the option not to take into account hosts or guests of chats. Hosts could want different choices when they produce recaps and analytics themselves.