Richard, maybe a double-game Twixt tournament is not such a good idea

I make some assumptions here. First, a tournament outcome of a clear winner is preferable to a tie. This is certainly debatable. There's no money at stake as far as I know. Players might not have much emotional investment in whether they win a tournament or not. I'm also a bit hazy with my statistics.

Twixt has two properties that might differentiate it from other games here: high volatility and rare draw outcomes. I intend to use the term volatility the same way the Glicko-2 rating system does. It refers to a player whose game results often differ from the result predicted by the difference in ratings compared to the opponent. It can also refer to the game itself. I believe the dense move branches generally lead to more mistakes made compared to other games. IMO Hex, by way of comparison, is less volatile, although draws are even rarer than with Twixt.

In the 66 League 1 championships that were single-game matches between each pair of opponents, there were no 2-way ties for first place. 4 of them ended in a 3-way tie, and 3 of those ties happened in the first 5 tournaments. I'm not sure why it turned out like that for those early tourneys. Maybe players were generally weaker then? At any rate, here we are in Championship 67 and a 2-way tie is the inevitable outcome. Of course Cosimo fully deserves credit for his achievement. I just wonder how many more 2-way ties we will see, if the double-game format continues.

Maybe a single-game format would be statistically more likely to end with a clear winner. There's also the increased workload that playing twice as many games incurs, which of course applies to all games. What does everyone else think?

I believe that Richard has moved to single round-robin for all tournaments. It was just an experiment.

Now I'm having my doubts..... I just noticed that a double round-robin started on September 1st.....

I agree with David Bush and I add some personal considerations.
In my opinion, the pie rule provides a good balance for Twixt. In other words, it is not that crucial to be the first to lead in Twixt. Thus, a double game is useless.
Furthermore, I found myself strained with so many games to play. I made a lot of mistakes, and not only in the games I lost, but also in the games in which I was able to recover. Probably, it was the worst competition that I have played in the last years (and, please, give more compliments to David Bush, because he deserved to be the only winner. See the Twixt Commentator for more details).
I left the Infinity competition, when I realised that I was obliged to play another match against the same opponent after some days. It is too much for my poor old brain.
I intend to abandon also these tournements as well, if this scheme of play will be applied again.

Maybe it would be better to reduce the number of players but keep the 2-round tournament.

It's your choice, but details would be appreciated.

We are in the Twixt forum. Are you talking about changing the Twixt championship only?

Cosimo and I have both presented arguments against the double-game format. We would be grateful for specific reasons why your suggestion would be better.

Do you propose to create more than three leagues with smaller tournaments? That would be a massive change from all the other championships.

I haven't played in the championship in many years. Maybe I will again, but I find that having to play 16 or so games at once is more pressure than I care for, and also too time-consuming. That said, I am going to make some comments, and some of this is based on my (extensive) knowledge of the history of Chess.

Firstly, to disagree with David, I don't see anything inherently wrong with double round robins compared with single round robins. I do not believe that they are any more likely to result in ties. Several of our single round robins ALSO have ended in ties. I might sight the great Chess World Championship Tournament won by Botvinnik, which was a QUINTUPLE round robin! But here's a difference: in Chess, you are never saddled with playing a dozen or so games at a time. Perhaps one or two per day, when live. Online Chess is different, of course, and I'll get to that.

Thus, my first comment is that, IF there is to be a double round robin championship, then at least DOUBLE the amount of time players get to analyze. Start with 48 hours per game, and consider adding 72 hours per move. This may seem excessive, but can be tweaked. As an observer, I must remark that players in all the sections seemed to PLOW through the many games as if they were being chased! (excepting a few who are still playing, taking their time) This can only result in more mistakes, reduce the quality of play, and put much more pressure on players. This is reflected somewhat in Cosimo's comment above. Keep in mind that there are only 3 tournaments per year - and rushing through this one will lead to a long wait for some anxious players (you know who you are). But rushing is not the players' fault, necessarily - since all games start at once, with the same 24 hour time limit. You are practically forced to play many simultaneously, and quickly.

So, if Richard wants to keep the double round robin, then making the sections smaller (so more separate categories) would help. So would increasing the time limits. Again - only 3 tournaments per year, so it makes sense if a single championship takes around 4 months.

I would also suggest other types of tournaments - and here is where I might draw some examples from online Chess. ONE type is single elimination, just like the NCAA Basketball Championship. They used to do that a lot, even in live tournaments. However, lots of folks didn't like the 1-game, you're OUT condition. Lately they've been using DOUBLE elimination, which may be better, though might be confusing. Also - there are Swiss Systems as used, for example, in LARGE tournaments. (heck - the Championship could include all 50 or so players this way!) Or there can be single eliminations based on MATCH PLAY (used in PGA Golf, and now online Chess) where players ARE eliminated but they must face a single opponent several times in a match.

OK, just some rambling comments from me, since I've been reading this thread with interest. Perhaps I've said something of use, and perhaps I'll re-enter the championship someday.
(meow)

MC, nearly all of your points pertain equally to chess or nearly all the other games here. This is the Twixt forum. Twixt differs from chess with regard to game outcome. Draws are almost nonexistent. As Cosimo pointed out, the pie rule serves to balance the game very well. That implies a single-game Twixt match between two players is more likely to produce a decisive result than a double-game match.

You mentioned the previous Twixt tournaments which were tied. So did I, in my first post above. 4 out of 66 were 3-way ties, and 3 of those were in the first 5. I suspect the players back then were closer to each other in playing ability. Since September of 2004, in 61 single-game tournaments, there has been one 3-way tie. Compare this to the only double-game tournament so far, which is a 2-way tie. Of course, it is difficult to draw strong conclusions from a single datum.

The point I have been trying to make, from the start of this thread, is that a single-game format is more likely to produce a clear winner. If you believe there is no inherent difference FOR TWIXT, you are welcome to your opinion, but I would appreciate hearing your specific reasons to counter my argument.

Maybe indeed in games where there is a swap rule there is no point in playing two games because they could be identical ?!

Well... yes, when there is a swap it's useless!

But I think for all games if there is a double round you need to reduce a lot the number of players in a same league.
8 should be a maximum, except for games like EWN where we don't care, and even 8 is still 14 games per player.

I also think the double-round-robin format is overkill, for all game types.

Fortunately AFAIK none of the current players are interested in duplicating moves, but it has happened in the past with a double-game Twixt tournament, and is a potential threat in the future for all games. This is one reason I pay a subscription, to have more vacation days to use as a defensive measure.

I guess it doesn't matter which forum we are in, as long as Richard knows about it. But if he wants input from all gamers, perhaps the general forum would be more appropriate.

Again I harp on my point that a single-game Twixt tournament is more likely to produce a clear winner.