What with most people staying at home to hide from the Covid-19 virus, there has been a good deal of interest in moving some tournaments that used to be played face to face to the internet. Challonge provides internet web services for tournaments of various kinds, and is particularly suitable for the new virtual tourneys. It has, among other functions, a “bracket generator“. I took a look at the kind of bracket it generates.
Various aspects of the Challonge product cannot be seen without buying a paid membership, and I have not done that. But I was able to generate a few brackets. I was particularly interested in looking at their double-elimination brackets.
The first thing to note is that Challonge brackets are not shifted. A 64 DE Challonge bracket plays out in 12 rounds (or 13 with a possible recharge), on the pattern A.B.|.C.|.D.|.E.|.F.X.R, not in 10 or 11 rounds with A.B.C.D.|.E.F.|.X.R. This is to be expected, I suppose, as most of the world seems blithely ignorant of bracket shifting. This fact, in itself, makes it unsuitable, to my way of thinking, for most applications – once you’ve begun to enjoy the benefits of a shifted bracket, it’s hard to go back. At least the Challonge system does, I believe, make the recharge round optional.
Apart from their failure to offer shifts, the brackets are not bad. I expected to find some sub-optimal drops, but that’s not the case. I noticed that the drops were spaced a little differently than the ones I use for my printable bracket. Like my drops, they managed to prevent any rematches occurring before round M. I ran a simulation (at luck = 3) on the Challonge bracket, expecting to find that it produced a few more rematches in rounds M and N, but that’s not the case. The Challonge bracket produced 2.9% rematches in round M, and 7.2% in round N, as compared with 12.2% in M and 2.9% in N for my bracket – a slight improvement, overall. In looking closely, I see now that made my drops without considering that M is a consolidation round. So I’ll go back and revise the drops on my unshifted 64DE – I’m not sure I can make them better than the drops in the Challonge bracket, but I should certainly make them at least as good.
The lack of shifted brackets makes the Challonge product a doubtful choice for running double-elimination tourneys. But apart from that rather serious problem, the Challonge product looks pretty good to me. I was expecting to be able to show that their drops could be improved at least a little, but I found instead that it was tourneygeek’s own drops that could be better.