The Good Sport "feature"

Announcements OGS Announcements
1

Superseded by The "Good Sport feature" has been removed

Hello OGS!

There’s been some (in hindsight obvious) concern about the “Good Sport” feature, what it is and how it will be used.

For those that haven’t encountered it yet, when a game ends you’ll now see a thumbs up thumbs down prompt below the rematch button that looks something like this:

image

What is this?

It’s a low-key feedback system. Was your opponent a good sport - that is to say, did they play their game without throwing a tantrum, abandoning, delaying your game, calling you rude names? Sounds like a good sport, give them a thumbs up. Did they throw a bit of a fit, abandon your game, tried to stall with the intention of wasting your time or getting you to resign, call you rude names? Feel free to report the egregious cases, but there are a lot of times where it doesn’t meet the threshold of a full on report, plus our volunteer moderators are already overloaded and might not get to your report in a timely manner, so maybe just give them a thumbs down in some cases.

Will my opponent know if I gave them a thumbs up or thumbs down?

No

Will this result in a “good sport rating” on my profile?

No, there will be no indication to other players about how good or bad of a sport your opponents have rated you, it’s not a popularity contest kind of a thing.

What about trolls who give everyone a thumbs down?

I hope it’s cathartic for them. Also, feedback from folks who just give an excessive number of thumbs down (or thumbs up for that matter) is easy to scale appropriately towards zero.

What effect will this have on my account, how the community perceives me, ability to play games, or anything else?

No effect.

How will it be used then?

A good question! But not one that I can actually answer right now because this is a bit of an experiment, so we don’t fully know yet. Dealing with cheaters and jerks that ruin the fun for those around them is an ongoing challenge. Chess.com has one of the best and most robust anti-cheat and moderation system in place, and this is one of many data points they collect, so I thought to myself, “well that’s interesting, they likely do that for a reason, maybe we should do that too”. I suspect there’s two benefits that are realized with the system: 1) It probably has a bit of a psychological effect to discourage people being unsavory to begin with, if so that’s a net positive. 2) It can likely be used as one of many indicators when reviewing reports and cases, prioritizing them, and potentially reorienting or suspending problem players quicker.

Another effect that I’m hopeful for is is the possibility of it reducing the number of reports that are below the threshold of being actionable while still allowing us to capture that feedback for an aggregate action. For instance, if someone loses connection and someone reports them for abandoning the game, that’s not actionable - we give the player the benefit of the doubt that they just had crummy internet problems. However, if this happens consistently, then they need a talking to that that sort of behavior ruins the fun for other players and they need to resign properly. Those sorts of cases tend to require multiple reports and/or diligent moderators, and so can go unnoticed or unhandled longer than anyone would like, a quicker aggregate view of feedback might help shine the spotlight on those situations sooner and with less manual research for our moderator team - that’s a hope at any rate.

So there you have it, I’m sorry I didn’t announce it earlier, it’s one of those things that starts off as more of an experiment and isn’t exactly some big feature so I didn’t really think to make some big forum announcement about it. In hind sight though of course I needed to write this up to alleviate concerns.

– anoek
20 Likes
2

Step 1: Collect data
Step 2: ???
Step 3: Profit

9 Likes
3

A good and fair summary :joy:

4 Likes
4

Will the mods be able to see those thumbs up / down stats somewhere?

5

In an aggregate form, but not yet, we need to filter and compute and in general figure out how to get good signals from it.

6

@anoek I really appreciate this informative post. (I gave you a :heart: up above).

I’ll add that the OGS site is awesome. I’m a newbie, but already have become a major fan in just a couple of months. To demonstrate my support I just became an Aji monthly subscriber.

I understand this initiative to implement a thumbs up/down poll is well intended and (now) more fully explained.

Nonetheless, I will not be responding to the thumbs up/down poll. Simply put, I do not want to be put in the position of rating other OGS players. This would fall outside my comfort level… And I feel that doing so would make me a “bad sport.”

In my short tenure I’ve already had a good number of cancellations and no-shows. Irritating, yes, but these aren’t violations, just annoyances.

If someone ever violates hard boundaries of behavior, I’ll decide whether to report them. If the mods are overwhelmed with a backlog of reports, so be it. (They’re only volunteers not paid servants.)

Meanwhile, I’ve found this community to be friendly, warm, and welcoming. This, without our being asked to rate each others’ sportsmanship.

I’m confident this community spirit can be maintained without our doing so. :folded_hands:

5 Likes
7

This is not what I would call a good sport, but a neutral sport. A good sport in my book is going beyond normal and expected behaviour, e.g in football kicking the ball out of play because opposing player injured even though referee didn’t call it.

5 Likes
8

I’m glad to be vindicated. I hypothesized that this was the reason in my first post in the discussion thread on this topic. One trouble with this feature is that it exacerbates an already large non-reporting problem. I estimate from experience that at least 2/3s of score cheating incidents and maybe as much as 90% of escaping incidents already go unreported. In addition, if this feature succeeds in reducing reports, it will thereby undermine the educative value of a warning. Escapers and stallers, among others, will be able to escape and stall more games until they are warned.

Even worse, there are huge categories of illegitimate complaints that will lead to down-votes. For example, large numbers of players complain about and sometimes even report their opponents for stalling because they play “too slowly,” or because they won’t resign when they are far behind in score. Many, maybe most, of these will give a down-vote that is illegitimate. This also applies to the mass of players who think their opponent is botting, or who get mad because their opponent refuses an undo or doesn’t give a greeting. This is all noise, and I don’t believe you have a means of separating the noise from the signal.

4 Likes
9

Definitely not. I expect many people will fall into the category of never using it and that is 100% fine.

We have the the best Go community out there I’d say. If this “feature” diminishes that in any way I’ll cut it in a heart beat. We’re great without it, I thought it was worth experimenting with to make us even greater but if that’s not the outcome, then it gets cut.

As it stands right now, I’d say the overall sentiment on this feature is quite negative, at least on the forums from those that care one way or another. Part of that is undoubtedly the lack of good communication about the feature and my intentions with it, but it might also just be fundamentally something we don’t want to collectively do.

My current thinking is to monitor this thread and see if the explanation helps ease any concerns, and sooner rather than later do some preliminary analysis of the data and see if it’s at all potentially useful. If it is, I’ll try and convince y’all with some numbers that hey this will be a net positive, but ultimately if we collectively feel it’s a net negative we’ll just remove it.

8 Likes
10

That’d definitely be a good sport, but I think many games go by without such an extreme opportunity to give someone a chance to show if they are a “good sport” or a “bad sport”.

We’re never going to get everyone aligned on some manifest of criteria about who is a good sport vs bad sport, it’s subjective on purpose, it’s a gut check.

4 Likes
11

I too view these as problems, but I come to a different conclusion regarding the possible effect of giving a low key gut check way of providing that feedback. We can discuss more in an internal setting if you’d like.

2 Likes
12

I rarely commented on feature-related posts, and generally consider myself more of an observer. However, I feel, the reason why it became a useful indicator for botting is due to the observer effect, people feel they are being watched by a system, and most crucially, those who act normal, and those who would like to avoid any contact to avoid attention.

The second one is that due to global conflict and geopolitical pressure building up, and screen name often tell their origin and preference, and people can check the opponent’s flags and profile easily, I suspect the first signal shown will be an indication of the ethnicity/demographic of particular groups, and this might even have a gender bias as well, or targeting minority groups. And I personally think this is not a good sign or data to collect, and can go down a rabbithole that you don’t want to go down.

7 Likes
13

Well, consider this:

  1. You have to play a game against someone to mark them as a good sport or not, so it’s not like you can just create a bunch of bots to go try and flag someone you don’t like as a super-duper-bad-sport.

  2. It’s not a public number, if anything it’ll draw attention to the nefarious individual(s) and they’ll the ones getting the attention of the moderators not their would-be victims. After all, looking at who is being thumbs-downed is only one way to look at the data, there’s also who is doing the thumbs-downing and who are they thumbs-downing.

This isn’t lost on me and you’re not the only one to share concerns about this. Though my intentions are pure, it’s clear the concern for either misuse or a side effect of cultivating a negative aspect in the community is quite real and perhaps justified.

4 Likes
14

I think you’ve nailed it. Indeed, there is little shared understanding of what constitutes a “good sport.”

If the specific problem to be addressed is made clearer, perhaps a less subjective question can be designed to provide useful information for use by the moderators?

Meanwhile, not quite knowing what its ultimate use will be leaves folks scratching our heads a bit. :man_shrugging:

1 Like
15

Will we (players) be able to see the number of up/down thumbs for ourselves?
(I already understand that we won’t see who gave us these, which ofc is good.)

1 Like
16

I’m keying off of chess.com here - I think the subjective gut check phrasing here, “is your opponent a good sport?” is actually important and correct. Despite a lot of chat about moderation and whatnot, this isn’t some trigger mechanism in and of itself I don’t think. It’s not like this is 3 strikes and your out or anything like that. It should be viewed much more casually - did you enjoy your game with this player or did they try and ruin your fun? Were the pretty cool, or were they kinda not? Are they naughty or are they nice? That kind of gut check.

Yep fair.

There’s a lot of chatter above is regarding the potential moderation side of things, but honestly I think all of that is secondary, I highly suspect that just the presence of the question - even if the data is thrown away and never used - probably has a net positive impact on by reducing the negative interactions that occur. The observer effect as @Counting_Zenist points out. Almost like those empty camera shells in stores. Plus it can be an outlet for people who would otherwise take it out on their opponent - I would much rather have them give a thumbs down rather than call their opponent a nasty name.

So time will tell if it’s more useful than just an outlet for emotions, but I suspect we’ll have the primary net positive effect of encouraging people to be nice and some secondary moderation uses.

I don’t think that’d be a good idea.

4 Likes
17

Yes, opportunities for good sportsmanship in Go are rarer, e.g. giving an undo when you were losing up until the blunder they asked to undo, so by my definition I’d only be giving “good sport” for like 1% of games, whereas the vast majority of “neutral sport” players I had a perfectly nice game with might be >90%, with 5% of bad sports being jerks. But you want me to give the thumbs up to that 90% majority of decent players who I would answer yes to “Was this player not a jerk?” or “Was this player satisfactory?” or “Was this a pleasant interaction?”. So perhaps the wording needs adjusting to get the results you want, if my interpretation of “good sport” is not uncommon.

3 Likes
18

I’d like to point out that while this sounds reasonable, it should be noted that many choices of other successful people/endeavors/companies/etc might seem useful, because those ventures are already successful, but that doesn’t mean that they are in any way part of that apparent success.

A good and recent example was that trend of moving away from physical buttons and indicators in car dashboards and moving all that functionality to the main tablet screen, which was a tactic employed by a successful company (in this case Tesla) and everyone jumped on the bandwagon. Later the other companies realised that this was actually a bad idea and rolled it back and that Tesla’s cars where not successful because of that implementation, but despite that.

Let’s not fall for the most basic “fashion/influencer statement concept” where someone successful and famous promotes a ridiculous piece of merchandice and people go “well that’s interesting, they likely do that for a reason, maybe we should buy that too”.

No, we shouldn’t, unless it is something we honestly like or find a use for.

You can already report those people and they know it and they do not care, even though a report to the moderators can have real repercussions to their account.
Why would a thumps up/down functionality, which will not even have a visual representation or warning, be a deterrant?

Couldn’t that be solved with an automated “internal counter of reports against that user”? :thinking:
Why shift that onus to the playerbase, especially considering that pressing the thumbs down might not even have anything to do with someone being reportable or unsavory in the first place?

Something more neutral and more constructive could be the question:
Did you have fun?
And if someone presses no, then it could open up a small feedback message pop-up as to communicate why?

4 Likes
19

As @JethOrensin suggests, perhaps confusing correlation with causation? :wink:

Screenshot_20240423_062356_Vivaldi Browser
Screenshot_20240423_062356_Vivaldi Browser1306×1301 185 KB

4 Likes
21

It’s precisely because it’s behind the scenes, they wouldn’t hesitate to judge or have existing “bias” against. And it won’t be just on one individual levels, there are nations/regions currently at war with each other and they don’t just have one single player in one region/nation, they collectively won’t give each up thumb up (not because they are good sports or not)) and very high chance just to express their feeling in their real life toward each other on a national level, and I predict they will be significant as a whole, and especially if they can get the attention of mods, to make the opposite sides as of a hard time as possible

And even if they are only after a match, a lot of games are automatic pairing, so the other sign might be players collectively just escape the matches facing certain demographic or rivals to avoid to have to think about thumbs up or down, thus less cooperation and games which shouldn’t brought real world sentiment into this. Also, mods are people as well, what would they feel and think about OGS, when collective regions in conflict or regional tensions get high indications behind the scene? Do they selectively pick those to deal with first? What positions would you put mods into, selective targeting a particular demographic or groups of people?

4 Likes