Incremental Time Controls

[Stewart Reuben]

An increment of 30 seconds per move means that the players, generally, use their time more efficiently. It seems to me that about one hour less thinking time for the whole game is just as good. 40 moves in 2 hours followed by all in one, with no increment leads to inferior chess. There are too many ridiculous time scrambles. 40/100, all in 30 + 30 seconds results in better chess.
When the game gets to be seven hours, then stamina becomes too big a factor. Six hours probably leads to better chess.
When it is two rounds per day, 40 moves in 90 minutes + 30minutes for the remaining moves + 30 seconds seems to be too difficult to fit in.
Thus we end up with All in 90 + 30 seconds per move. That is rather fast.

I believe the introduction of the 30 second increment is one reason why the standard of international play has improved.

[Roger de Coverly]

When it is two rounds per day, 40 moves in 90 minutes + 30minutes for the remaining moves + 30 seconds seems to be too difficult to fit in.

If you have a Friday evening round, I would agree. If it's a Bank Holiday weekend or a working week tournament, I don't think playing from 10.00 to 3.00 and 4.00 to 9.00 (or half an hour earlier) is particularly difficult to schedule. The parallel schedule for 90 30 being 10.00 to 2.00 and 3.00 to 7.00. In the brief period where 90 30 was outlawed for Norm tournaments, I'm aware the e2e4 ran at least one event with five hour scheduling. The timings weren't popular because of people finishing their morning games at or around the 40 move four hour mark and having to wait around for the next round.

[Stewart Reuben]

That may work for a round robin. But a 5 hour session for 60 moves with play in the second of the rounds only 6 hours after the start of the first one, runs the risk of a very long game meaning there is inadequate time between rounds. The pairings may be held up.
Thus people have inadequate time for preparation. Also some people like to eat in between, or after the game. Of course, if you run a tournament to simulate weekend Swiss, with 3 games on Saturday and just 10 minutes after the announcement of the pairings before start of play, then it is fine.
If the game finishes at 9pm, then eating an evening meal, before or afterwards is very problematic.

I have the ambition to have a Swiss where each round is of two games against the sme opponent, played on the same day. Probably 7 rounds, that is 14 games. Then 40/90, all in 30 + 30 seconds would be possible. The players in each match could choose their time of play at their convenience, depending on when the first game finished. No such event has ever been organised anywhere in the world - yet.

[Roger de Coverly]

That may work for a round robin. But a 5 hour session for 60 moves with play in the second of the rounds only 6 hours after the start of the first one, runs the risk of a very long game meaning there is inadequate time between rounds. The pairings may be held up.

Two of this year's Easter Congresses, namely Cardiff and Southend used five hour increment sessions. With a short turnaround between rounds, it's as well to trust computer generated pairings to save time.

[Stewart Reuben]

Roger >With a short turnaround between rounds, it's as well to trust computer generated pairings to save time.<

It is generally best to rely on computer generated Seeded Swiss pairings at all times, rather than on fallible human beings. There is no good Accelerated Swiss Pairing System, but that isn't used in Cardiff or Southend. If a game is unfinished in a tournament, a computer won't be able to make the pairings. Of course the human can decide to assume it is a draw or temporarily adjudicate the game for the purpose of pairings. I remember doing that for the early Lloyds Bank Masters when there were two rounds some days.