I think that language may have been deleted between the 4th and the 5th editions, but I’ll let someone else do the legwork.
I’ve been working on a novel, Bill, maybe if the technology refresh program is successful I’ll have more time to work on it. (It has nothing to do with chess, though I do have the beginnings of an outline for one about 3 chess players vying for the suddenly-vacated world title.)
Susan agrees with me that the G/25;d54 is probably a typo, because the TD has run many events at G/25;d5. He’s being contacted.
If you can find the TLA, you might not even need to contact the TD. For that matter, even if the TLA says G/25 d/54, you could look at the scheduled round times, and if it looks as though none of the rounds could possibly have finished on time at that control, it would probably be safe …
Only around 20% of the events rated by US Chess have a TLA, and I don’t think this organizer’s events are among them.
George RR Martin’s short story, “Unsound Variations”, has characters based on a number of Chicago-area chess figures in it, but I don’t think either of us are among them.
I’d like to think the characters in the book I’m working on are made out of whole cloth, but that’s almost never true, especially for first novels, or so my creative writing teacher at Northwestern said.
The host site should be able to handle triple-occurrence and 50-move claims without the aid of the TD. In fact, the host should act as the TD in such cases.
It may be wise, though, to change (or at least bend) the triple-occurrence rule a bit, to accommodate the inherent differences between online and OTB play.
There could be a “Draw” button, on the screen at all times for the player on move, which could double as a “draw claim” button and a “draw offer” button. A player who wanted to claim or offer a draw would have to click on this button before playing his move. The host site would then act as follows:
If the position on the board, without a move by the player, is the third occurrence of the position, or if the 50-move rule has been satisfied, the host would immediately declare a draw, end the game, and inform both players.
Otherwise, a message like “You have requested a draw, now please make your move” would appear, and the host would continue to wait for the player to move.
After the player moves, if the new position now on the board satisfies the triple-occurrence or 50-move rule, the host would immediately declare the draw, as above.
Otherwise, a message will appear on the opponent’s screen, “Your opponent has requested a draw. Click the Draw button to accept, or make a move to decline.”
If the opponent accepts, the game is over. Otherwise, the game continues.
Some OTB purists may squawk that the above procedure blurs the distinction between a draw claim and a draw offer. I say, it does indeed, but so what? Under the rules, any draw claim is also a draw offer anyway. If a player wants a draw, he presses the button, otherwise he doesn’t. The player doesn’t even need to know whether he is claiming or offering. Nor does he need to know whether he is claiming by the triple-occurrence rule or the 50-move rule. If either claim is valid, game is over.
I think this would be an extremely clean way for an online host to handle draw claims and draw offers. Do any of the sites already handle it this way? Confession: I have never played chess online.
I believe a total time of 5 to 10 minutes has been the definition of Blitz since that ratings system debuted in 2013, leaving 11-29 as the domain for quick-only.
Non-integer time controls are something that parser should probably flag as an error, even if there are clocks that can utilize them. I’ll run that past the ED.
I believe the question was whether mm>=10 is required for on-line quick which is how someone interpreted the G/10;d1 as the “minimum” example. I doubt that was the intention; G/8;d3 is, and should be, a perfectly reasonable quick time control. Absent any other guidance, I would think that the mm>=5, mm+ss<=29 should still govern the range for on-line quick. An interesting question is whether you would see any of those goofy time controls with mm=5 and a bigger ss—something like G/5;d10 rather than G/10;d5 seems to make sense only if you are trying to design a time control to obviate the need for scorekeeping, which isn’t relevant for on-line chess.
G/5;d10 seems like a reasonable time control to me, but there was a time when I played 10 second chess, so it seems like 10 second chess with a little extra thinking time.
Again, my understanding (having written the parsing code) is that Blitz goes from 5 to 10 minutes, quick-only from 11-29 minutes, regardless of whether it is OTB or online.
I’ve asked the ED for guidance, including on the non-integer issue, it is being passed on to the RC Liaisons.
I believe the current parsing code would either ignore the non-integer portion of the time control or round down. Either way, G/25;d4.9 would be a total time < 30 minutes.
G/5;d5.1 is potentially messier, though ignoring non-integer portions or truncating down is straight-forward.
There are some events where the delay in the time control has been listed as “d.5” instead of the usual “d5”. I’m assuming these events were played with a 5 second delay. Does the current programming recognize the “.5” as 5 seconds?
There are also some events where the delay in the time control has been listed as “d05” instead of the usual “d5”. I’m assuming these events were played with a 5 second delay. Does the current programming recognizes the “05” as 5 seconds?
I’m not sure how it would interpret ‘d.5’ or ‘+.5’, or, for that matter how it would interpret G/.5.
I’m pretty sure that ‘d05’ would be the same as ‘d5’.
I need to run some tests to make sure the parser handles the online-slow time controls properly, since we don’t have the dual-rating zone, I can add that to the tests, though depending on how quickly I get back guidance from the ED/RC Liaisons that might change what tests need to be run.
As Bill has pointed out, banning non-integers in time controls or a statement saying non-integers in time controls are always truncated to integers would clarify matters.
Does the current programming recognize the phrase “no delay”, even if it’s in the next sentence and after a period, as “d0”? Would it still recognize “no delay” if misspelled as “no daley”? Would it still recognize any phrase that might be in there, such as “none of this new-fangled extra-time stuff”?
Somehow I suspect that the answer to at least the 3rd question would be No. Every so often, human intervention is required.
Aha, so in the case of just a single control, it will accept only “Gmds” or “Gmis” or “Gmincs” or Gm+s", where m and s are integers, the other letters must be exactly as shown except that they may be uppercase or lowercase, and there may be slashes, commas, semicolons, or spaces between the elements.
Bill, to be honest I haven’t looked at the code much in several years, I don’t really remember exactly what it can and cannot do. It’s worked well enough for somewhere around 200,000 sections since 2013 that I don’t remember the last time someone couldn’t get their valid time control entered, even ones that some might find bizarre.
