Lichess Pools

@PhillipTheTank

And that's exactly what I mentioned in my original post.

lichess DID do that. They just weren't very popular.

Maybe it would be different now, maybe not.

I know the lichess implementation was different than ICC's; whether it was so "flawed" is another question entirely.

A hugely important difference is that on ICC the tournaments have always been rather unpopular and sparsely attended.

Here they've always been wildly popular, and they do offer an auto-pairing with no choice of color or opponent. They just do it based on tournament standing instead of rating.

There are several reasons pools failed here; maybe they would work now, but maybe not.

If enough people ask for it, then maybe you'll get your wish :)

No, lichess didn't do that. Lichess made a mess of it. The problems with lichess's implementation were many. The main two problems were:

1) Players were not paired immediately. They sat in a queue for sometimes more than a minute until the next "wave pairing" came along. This was terrible.

2) Players who were waiting to play in the pool were literally visible to potential players who hadn't joined yet, allowing not only choosing of opponents, but actively discouraging people from joining the pools under various circumstances (such as not enough people at your rating, not enough players at all, individuals you don't want to risk being paired against).

The proper implementation would fix both of those major issues, which literally crippled the pools. As soon as two players close enough in rating have entered, they should be paired instantly, and the list of players waiting to play should not be visible to anyone -- not even the number of players.

Whether this implementation was "so flawed" is not even a question. It was patently obvious.

I'm rather pessimistic about this ever happening the right way because of the twin reasons of:

1) now the terminology is all screwed up and it's almost impossible to even have the conversation.
2) every time you get past that, you have to fight through the mess of the previous horrible implementation.

It's doomed, and I think that's tremendously sad.

ICC's auto-pairing pool is STILL the best thing in online chess.

I subscribed to ICC a few times for a few months just because of their pools system.

I agree with @PhillipTheTank description of the facts.

My opinion is that "ICC pools" are a far more accurate way to rank competitive players from best to worst than rating-based leaderboards are, and that they guarantee fast, close pairings against highly competitive opponents.

*shrug*

I guess we'll just have to disagree somewhat.

I acknowledged there were multiple reasons the pools implementation here wasn't as popular.

Let's not get all rosy about ICC's, though. It's simply not true that you get instantly paired all the time.

I and a friend of mine still have ICC accounts, although we don't use them much anymore, and I've had to wait 30+ seconds multiple times in the last 20 or so pool pairings I've attempted (and this is getting more and more common as ICC's player base gets smaller and smaller).

My friend's experience has been similar.

The visibility into the current pool players was a problem, undoubtedly.

I just don't think it was the main reason the pools died.

The wave pairings might have been a problem, but I'm not so sure about that either. With enough players the waves would come quickly; the big problem was just not enough players.

At any rate, I agree (and have always agreed) that ICC's implementation is better; we'll just have to disagree on the degree to which each of these factors contributed to the pools' demise.

It is true that you get paired instantly when someone is available. That is the key. Even if there were people available in lichess's implementation, you sometimes had to wait a minute for the "wave pairing" to happen. It was awful.

And yes, the point is with lichess's much larger population, the wait times would be nearly nothing, so citing ICC's dwindling population just helps that argument.

Sure, I'm happy to agree to disagree.