Killing for frowning is using analysis in live games in the analogy, though obviously I don’t think it’s as bad as murder. My point was that when people have grown up in a certain culture, they might not realise that something normal in that culture is abnormal in the wider world. OGS is the only server I can think of with analysis board even possible in live games, let alone enabled as the default. The only time you do analysis of variations during a live game in real life is in teaching games or casual games in a club between friends, it would be out of the question in any serious tournament, amateur or professional (nowadays, I know things were different 100+ years ago in games spanning many days such as between the Japanese go houses, or Shusai’s famous myoshu, those are more like correspondence games) . That’s not to say the rest of the Go world sans-OGS is necessarily right, if you want to argue live go with analysis should be the new normal then that’s a position you can argue for, but it should be from the informed position of its current abnormality and the reasons for that. OGS, as the introduction to Go for many Western players, has a responsibility to set a good example and expectations about culture and etiquette.
People played live go on the internet without analysis for about 20 years before OGS added it, and continue to on other servers. I’m not even sure if the change on OGS was a conscious design choice, or an accident from the developers not being/asking experienced go players, but I suspect the latter. I think it is better if beginners realise that you are supposed to read in your head. Exploring variations on a board is a valuable part of learning, particularly when guided by a teacher. But that’s for practice mode, not game mode. And with my choice of off by default for live, it’s still an option to turn on analysis if you want it, it’s just discouraged a little bit by not being the default: “I want to play a practice mode game, not a normal game”. I don’t see how the difficulty of reading ladders is a bug, anymore than finding good moves can be hard is a bug, it’s part of the skills of Go. In fact on a computer it’s quite a bit easier to read ladders than in real life, as you can trace them out with your mouse (if not on touchscreen). Tracing out a ladder with your finger just above the board is not polite in real life games, though I’ve seen it happen!
I don’t have a strong opinion either way, but would tend to allowing live ranked analysis. I could see the argument “as most live games are without analysis allowing in some ranked games means the ranks are wrong”, but don’t find it persuasive as I think it places too much emphasis on preserving the apparent sanctity of the ranking system at the expense of people’s freedom to do things which aren’t bad. We don’t, can’t and shouldn’t stop people playing when they are tried, drunk, just been to go camp or drilled some tsumego so their performance will differ from the average so I don’t see allowing a mix of analysis on or off games in the same ranked account is a problem. If someone wanted to make separate accounts they could, but not forced to.
Do we know if this change is actually being considered, or if it is merely within the realm of “things you would like from OGS but will never happen” ?
I feel like the poll result shows a very large majority in favor of disabling analysis at least in live games (even without accounting for the status quo bias highlighted by Uberdude).
I question the assumption that the analysis tool is used in live games to a significant extent (which I will define for the purpose of this post as more than 20% of non-teaching games). Of course, it is perfectly reasonable to object to its use in even one live game as a matter of principle. I understand and agree with that. However, the general tone of this thread is that its use in live games is a major problem, even though we do not seem to have any evidence of that.
I am skeptical for two reasons. First, I have always watched numerous games on OGS (especially DDK and SDK), and the large majority of them proceed at a pretty fast pace, even when there are generous time settings. Obviously, on those occasions when a player takes more time to think (say a minute), they could be analyzing, but that is not very common in my experience. And that much time is very rarely used in 9x9.
Second, the typical time settings do not give much scope for use of analysis. The principle opportunity would be during main time in 19x19 games. However, that comes early in the game, when complicated reading usually isn’t necessary.
Lest I be misunderstood, I favor disabling analysis in live games, even by default. I just don’t see any evidence that use of analysis in live games is a major problem.
I wasn’t debating whether its use is good or bad, which seems to be the focus of the Reddit discussion. I was questioning whether it occurs to a significant extent in live games.
I don’t think you can use speed of play as a measure of this - I can use the analysis tool as fast as I can think, and I can think faster if I’m using the analysis tool.
It’s literally a fraction of a second to pull it up and visualise a variation that is too hard to hold in your head.
(This said, I don’t play a lot of live games, but when I do… )
Nonetheless, I was hoping to provide input that the 20% might be on the low side of the estimate, because it was premised on the basis of playing speed. The main point I was making is that you can play quickly while still using the analysis tool to advantage … which in turn means that the % might be higher than you think…
I think you have misunderstood what the 20% represents. That is not my guess about the amount that exists. It was intended as a rather arbitrary benchmark for what could be considered a significant problem. If I had set a higher number it would make it more difficult to prove there is a problem. So I was being generous by setting a lower number. The benchmark isn’t about how many exist, it is about what percentage constitutes a problem. Some people might consider 1 case a significant problem, and others might not care if it were 100%. If you think there is more than 20% than you think it is a significant problem per the benchmark, or else the benchmark should be changed. As for me, my position is that there is no data. I’m not even decided about what constitutes a significant problem in terms of percentage of use.
But I do think you postulated that the number of people using analysis is low, based on observation of playing speed. I think it’s higher than you would suggest, because you can use and benefit from the analysis tool at reasonable speed.
Conrad_Melville: my point about culture and setting expectations of beginners is that the problem of analysis enabled by default in live is not simply based on its use, which is some unknown percentage we are talking about. I see it as a problem that a beginner who learns on OGS thinks that simply the ability to open an analysis board to help with reading is a normal thing to do in live games, even if they don’t actually do it often or at all. So to gather statistics on that would be a survey of attitudes rather than behaviour.
Also, I don’t think we need to prove the current situation is a significant problem for making the change to be a good idea: it’s a small change in terms of development effort, and it doesn’t take away the ability to play live with analysis from those who wish to so I don’t see it having a downside.
I don’t disagree. And to repeat myself, I favor disabling analysis in live games. My post addressed what I consider, generally speaking, the overly dramatic tone of the thread.
Again, I don’t disagree. You have stated it with the proper tone here. A good idea for improvement doesn’t need to solve a major problem and doesn’t need to be boosted as if it does, when there is no data indicating a major problem exists.
TBH i dont see any reason why can’t we change the current "disable analysis [ ]* " box to “enable analysis”.
There is big psychological difference between asking if people want to enable something, or disable something. “Could you tell us what happened” is entirely different question with different attitude than “do you want to use your right to remain silent” - even if you’re basically asking the same thing.
I dont really care if it would be checked or unchecked by default when making custom challenges or custom tournaments, ppl can check/uncheck that if they want.
All i’m asking that analysis would be enabled for ladders and corr-AST’s, since analysing the game is (/has traditionally been) big part of correspondence go.
)