Make "analysis disabled" the default (poll) - COMING SOON!

Feedback Suggestions
82

Not specifically refering to the referendum but the Tory infighting that led to Cameron calling the referendum.

So not weasel worded.

83

nooooooo! Not a referendum :wink:

1 Like
84

From my understanding of how things have been handled so far with regards to issues such as

“someone assuming bent 4 is dead, but theyre playing chinese rules => too bad you agreed when you accepted the game, it was your fault for not reading the game info”, and

“someone not realizing it was a blitz game => too bad you agreed when you accepted the game, it was your fault for not reading the game info”, and

“someone not realizing analysis was enabled => too bad you agreed when you accepted the game, it was your fault for not reading the game info”,

then the opposite should be true too. If you agree to disable analysis, then anything resembling analysis mode is against the agreement the players made when they posted/accepted the game. Essentially, anyone not available to the players in the OGS game itself, or specifically stated by the OGS rules (joseki books for corr games, etc), then it’s cheating.

2 Likes
87

cheating

This.

I can’t help but have a sense that underlying this debate about the analysis tool is a perception that having it on by default means that people who use it are somehow cheating…

But, as has been mentioned, use of analysis and/or AI is reflected in the rank so what’s the big deal?

2 Likes
88

In thinking about whether analysis off should be the default, I was struck by this argument.

“Their use of analysis is already reflected in their rank, so why do you care?”

That is quite compelling.

So I too find the question

"Their use of bots is already reflected in their rank, so why do you care?

fascinating.

Both give the player an advantage they would not have in real life. Why does one seem more acceptable?

I guess it is as someone pointed out: one helps you think more (analysis) so we don’t mind that as much as one that means you don’t have to think at all.

2 Likes
89

How do you get this from the poll?

It appears that nearly 60% of respondents would like analysis-off to be the default for live games, so the lean is in that direction :thinking:

90

So what about the fact (in correspondence games) one can place the stone without submitting.

At a stretch one could argue that being able to see one’s proposed stone on the board before clicking submit gives the player a slight advantage…

Will we be seeing next discussions about whether to not be able to do this, that once the stone is placed on the board it is submitted?

Because you don’t do that IRL…

2 Likes
91

That’s not actually true. Multiple questions were asked in the poll, and the options were biased towards “disable”.

It is only 38% to 34%. It is nowhere near 60% for one side.

1 Like
92

You need to add up 38% and 20% and that’s 58%.

2 Likes
93

We should make analysis-off the default for live/blitz, but leave it on by default for correspondence

We should make analysis-off the default for all games

It’s inappropriate to say these responses should be in the same bin.

94

It seems perfectly reasonable to think that the people in each of these bins thinks that analysis should be off by default for live/blitz.

4 Likes
95

I have misread the poll and may have misremembered an earlier status. Maybe we are closer to an analysis referendum than I originally thought.

96

  1. Why does a yes/no poll have more than two answers?

  2. Why did you not do something like…

  • Do you make use of analysis in OGS? Yes/No
  • Should OGS analysis be enabled by default? Yes/No
  • Do you feel that OGS Analysis has a significant impact on performance? Yes/No

or rather “Based on your personal experience, does analysis have a significant impact on OGS performance?” The results will be more useful when they’re parsed out like that and it’ll be easier to take the survey.

  1. You made the poll knowing what you wanted out of it, and the conclusions you’re drawing really need to be thrown out, not just because the results are not actionable at all.

  2. The only thing you can take away from this is it does look like no one “doesn’t care”…

  3. I’m willing to bet there were a significant amount of people that misunderstood the poll. You should really re-do the poll to make sure, especially since no bias or factor analysis, or anything to look at bias has been done. Because of point (4), the poll needs to be done properly.

1 Like
97

As long as it’s advisory :wink:

1 Like
98

It’s not really a yes/no question though. First, we have correspondence and live which are different. Second, people care to different extents about this, I bet a lot of people don’t care about this that much. How about 10 point scale?

Of course, we need to remember that 60 people browsing forums don’t represent thousands playing on OGS so it’s all just a rough estimation.

And I think the statement that original poll is rigged to get specific results is wrong and silly.

4 Likes
99

Polls are one of those impossible-to-get-right things :smiley:

That being said, although the title appears to be a yes/no question, actually the poll is about opinions on options, and this should be pretty obvious: the fact that the title question is yes/no is a problem only in a nit-picky sense.

In general the set of questions in a poll is a notoriously hard thing to get right (see the recent OGS annual poll for an example :smiley: )

This is obvious when you think about it: when you start a poll, you don’t know what thoughts are out there - after you conduct the poll and get the resulting feedback then it’s easy to see what better questions there would be.

So the answer to

“Why did you not do something like… [things]”

Is because

a) I didn’t know what I know now
b) I wasn’t even trying to ask some of those things - those are interesting to you, and may have come to light as interesting questions, but were not what I thought we needed to explore with that poll

2 Likes
100

Interesting replies… considering those weren’t my words, but those of a friend who is a working post-doc in statistics.

Just something to think about.

101

Yes - your suggestions had the smell of someone who knows what they are doing in polls :slight_smile:

However, I was answering your question about why mine was the way it was…

… I suspect you weren’t really interested in the answer though, you were trying to say “why don’t we do a better poll, designed this way?” :wink:

2 Likes
102

I said all I could previously and had made my case. The topic was generating a lot of heat, so I figured it would be worthwhile to have someone who knew what they were doing check it over and see if they agreed with you or yebellz, cuz I wasn’t so sure myself. Turns out he just didn’t like the poll itself ¯_(ツ)_/¯

I meant more that Alex’s responses were interesting (and rather strangely defensive), and I actually was interested in your answer, I just don’t have a response to it. I agree with what you said, and would have done no better myself in making one. Also, note that “interesting” doesn’t mean “bad”, or anything. It just means interesting. To me, anyway… I get that tone and whatnot doesn’t convey itself over internet forum text, so I apologize if it seemed like I was aggressive or anything.

If the analysis mode thing really is causing this much discussion, then his suggestion of “why don’t we do a better poll”, is what I inherently agree with. Especially because it’ll have a large impact on new players.

3 Likes
103

Thanks for your understanding and measured response, much appreciated.

We could do another poll for sure - the current one has elicited useful conversation that might inform a better one.

I’m not convinced that the proposed alternative is better. It starts with two premises.

  1. You have to have experience with using analysis to have a valid opinion on the topic
  2. Analysis has to demonstrably improve a person’s performance in order for it to be worth changing the default setting for it.

I don’t think either of these are obviously true.

“Based on your personal experience, does analysis have a significant impact on OGS performance?"

I don’t think this is the key question. It would be an interesting result, but we would need another poll after we find out the answer, which asks

“Given everything we know about analysis, should it be enabled by default for live/blitz?” and
“Given everything we know about analysis, should it be enabled by default for correspencence?”.

In fact, now that I typed those, maybe that is in fact the poll we need!

4 Likes