A “relax game” button is an interesting suggestion to me.
But in the state of the things, we may talk on getting into some debates for weeks in the middle of an implementation launched but not yet fully rounded.
So I 'll push you using some custom games setting instead.
1 Like
I agree that’s an interesting setting, but I highly doubt it would ever gather enough popularity to justify being a quickmatch setting anyway (especially as we’re trying to keep the number of options low).
This seems fully within custom games territory.
2 Likes
Yeah, it could have been useful to have a question such as “Would you prefer that only one time-setting is retained, even if it isn’t your preferred option?'”.
7 Likes
I prefer the same but feel the Fischer time increments are a bit too low and should be adjusted upwards a bit, especially on 19x19.
The timeout rate on 19x19 blitz is not surprising somehow, 3s is not much time at all, and it takes time to click anywhere on such a large board even when just moving the cursor around.
I’m not sure of exact better increments, but something like or around Blitz +5s, Rapid +8-10s, Medium +15s feels better ?
Edit : Woops, sorry, didn’t mean to reply to the message you posted, @josjoba – it’s a nice idea for custom games, though – I prefer longer games like that sometimes even as an experienced player, and I’m fairly sure a custom game of 5:00+5x60s would be accepted by someone if created ^^
2 Likes
How about this complete redesign, a bit after the way the game difficulty is set.
I see on the ‘play’ page ( Play Go at online-go.com! | OGS ) 7 timing options as buttons. That’s a lot, and can lead to fragmentation, while on the other hand we are dealing with a continuum (time).
If the user had the ability to choose both an ideal time for the three variables “main time”, “Byo-Yomi time” and how many times the Byo-Yomi may be exceeded, and then for each the option for how far off this setting they would still accept a game, you have the interesting opportunity to calculate a better settings for those two players who are within each others timing bounds.
The players can then increase this margin, if they do not find a game quickly enough, until that they do. The benefits could be:
- People can play a wider variety of game timings (anything can be set up now).
- It is still a fairly neat interface I guess (three ideal settings and three times the margins on them). It follows the Kyu/Dan scheme a bit, also a margin around a central value (your own level).
- Perhaps most beautifully is that two persons who may want to play a different game from the usual, but who have to settle for something more popular, will now have a chance to be matched on the otherwise too rare occasions that they would both be looking.
Example: Someone wants a game with a 0 minute main time, but a 60 second Byo-Yomi, and 10 times over it, 0:0+10x60s. Since it is too rare to find anyone (let’s assume), he gives up and plays the closest 5:00+5x30s from the button.
Someone else may be in the same situation, but in this different interface he could set something like 2:00+5x90s as his ideal. It isn’t exactly the same, but more the same than the usual games they get. When they are both finally look, they can be matched and get a game like 1:00+7x75s. In cases like this like the average between 5 and 10 is the impossible 7.5, you could go with the weaker Kyu/Dan being benefitted, or if it is the same the one who played more/fewer games, etc.
The whole thing then becomes dynamic. I don’t know if I explained it well enough. For the sake of simplicity, you could always have three quick-fill buttons, which could say something like ‘Blitz’ / ‘Live’ / ‘Long’, set to values which will quickly yield a game, to still have that benefit.
You could go even more fancy by having an optional timer which slowly increases the margins on those settings, which could ensure you eventually will find a game. You could then break that down into by how far that should go, and how fast it goes.
A difference with costum games as it is now, is that it isn’t one side who defines the game and the other accepts, but rather a calculated middle.
EDIT: So if person A sets 0:00+10x60s as the ideal, with margins (±)4:00+(±)5x(±)30s, and person B the now usual 5:00+5x30s with margins (±)1:00+(±)0x(±)5s, they could end up matched as: 4:00+5x35s. If person C sets the same as A to get enough games just like A did, once their luck strikes they get their game 0:00+10x60s (and might play many more games happily ever after ;-).
1 Like
What if I don’t want to play a wide variety of game timings? 
Personally I’d rather play a game with a regular fixed time setting that I’m comfortable with, than a weird setting I’m uncomfortable with.
I feel like one advantage of picking a weird setting like a 6s or 9s byo-yomi, or having 2 periods instead of 3 is just that you can throw off your opponent if they forget that “usually” they have slightly more time and so they have to rush or just timeout.
I don’t see why I would want to agree to time settings like that just to get a game.
They should do that 
Like custom games make sense if you really really want to play
and nothing else.
But personally I would play a standard time setting regularly than a unfamiliar time control every game.
3 Likes
Hi shinuito. In this design (see above) you would just set the margins to zero. Then you would get your game exactly. Since there would be so much freedom, there is less reason to offer so many standard buttons (now 7). Fewer standard options reduces the fragmentation, while people who want different can search for what they want, while also still accepting the standard if they want that.
I think you might be able to appreciate that a system where you specify not only 3 main times but margins around which you will accept time deviations is a much more complicated system than simply
even if it
If we’re worried that just having 6 or 7 buttons to click is too complicated, then this new system would be above the level of most users.
They might not even learn how to use it for example.
and yet they are playing/learning Go.
Hi again shinuito.
It is true that it might be slightly more complicated. However it is in principle a similar setting as to choose how far off your competence level you want to accept a game (Kyu/Dan). If someone can do that, then can’t they also set their ideal total match time, their preferred Byo-Yomi time, and the margins they would still be okay with ? If they cannot or don’t want to bother, there can still be the standard choice button, right ? I also think that if people are already playing Go online (as someone else also writes), which is a lot of clicking and also a (very) challenging and deep strategic game (uff lol), then is this really too difficult ?
If I see 6 standard buttons around quite similar time settings as it is now, then it seems to almost be an attempt to allow infinite possibilities. Unfortunately these 6 choices may cause fragmentation on the one hand, while still not truly allowing everything users might want ? Will it be 10 buttons soon ? Why these choices, not others, etc. If you go to a system which can allow anything (as proposed) you achieve that flexibility goal, and you can reduce the amount of standard games from 7 to 4 for example (Blitz, Live, Long, Correspondence, but each just one), and that might reduce fragmentation. Hence there isn’t much of a loss ?
If too many people chose costum settings and can no longer find each other, then they can just set margins wider again, or go for a standard choice.
Personally I was quite confused by the buttons at first, because I had no idea what some code like 5m+5x30s was supposed to mean. I had to search around on the internet, and eventually found the answer and with reading forum posts I eventually confirmed what it means. I think this is where the interface might be simplified: a bit longer explanation on what those codes mean, perhaps in a pop-up ?
Have a great day. It was just a proposal, just an idea, that’s all : ).
Compare with the announced times on the Play page:
- Blitz: ~15 minutes
- Rapid: ~25 minutes
- Long: ~40 minutes.
The ratio is:
- Blitz: (7+5/60)/15 = 47%
- Rapid: 64%
- Long: 63%
Considering that people may timeout or resign early, I think a correct ratio should be around 75%. So maybe something like
- Blitz: 30s+5s
- Rapid: 5m+7s
- Long: 10m+12s.
6 Likes
The point is more so that if having a few buttons to choose from is confusing, either because of the time controls etc, and requires reading and asking on a forum to understand then
then this sort of a thing sounds even more difficult to grasp.
Discussion is always useful, and aspects of different ideas can always be used even if the idea as a whole isn’t used!
^^ It’s not out of the question so far that some sort of merging of the two could happen.
The announced times correspond closer to 90th percentile times, it’s meant so you can budget your time accordingly.
2 Likes
I’m looking at +5s and +7s right now. +10s seems to line up quite well with the byo-yomi counterpart and the majority of folks say it’s not too fast, so I’m inclined to leave that one.
4 Likes
That is true I think, it is overall more complicated.
Downsides …
- A bit more complex interface.
- Users have a (likely) need to understand how game times end up being computed, which is a burden of learning & explaining.
- To quickly adjust to unusual settings when the game starts, if you had set margins on your game (as you mentioned). I thought you could flash the setting on the screen for a few seconds, but if it keeps changing from game to game, it may be hard to adjust with your eyes on the board.
- Could draw people away from the more common time settings, making it harder for them to integrate with other players. Your “just get used to it” argument.
1 Like
Ok all, seeing as the polls aligned with the data in terms of the blitz and rapid fischer times being a bit too fast, I’ve crunched some numbers, analyzed the suggestions in this thread, looked into my crystal ball, and everything lines up pretty well for trying +5s and +7s for blitz and rapid fischer time increments respectively. I’ve gone ahead and updated those settings, so we’ll see how those work out over the next few weeks. Thanks again to everyone who participated in this poll and discussion!
11 Likes
Just a quick reply/glance at this as I was surprised when reading this, currently 43% are not interested in fischer long clocks, 33% think + 10s is fine, 24% think it’s too fast, so it’s about a 57/43 split and not a large majority – though this isn’t really a complaint.
Personally I’d try playing +12s Fischer for long settings if they were there, to see if it is a bit more evenly paced but probably rarely or never use +10s on 19x19, from experience with it, but not sure how others feel about it in specific if it’s closer to 50/50 or if it makes a large difference, and it’s not difficult to find or create a custom game with higher Fischer increment either way.
The new +5s and +7s increments look nice, and interesting to try !
@fuseki3 @mirthturtle you said you prefer custom games. Could you share why and what time controls please?
I prefer to create custom games mostly because of the Ranked, Disable Analysis, and Handicap options, whereas I’m not sure what an Automatch will give me. And I like having presets for my favourite time settings.
1 Like
Thanks, and what are your favorite time settings?