Overall Decline in Quality

[EllieK]

As someone who loves BGA and only ever says glowing things about BGA, I am left with an uneasy feeling about what feels like an overall decline in quality. Over the past many month I have watched what was once a great gaming site decline into a pool of unorganized implementations. First a little bit at a time but now in great waves.

For example, the Innovation interface was upgraded with fancy new buttons at the top of the screen but now games won't end correctly and the player log that actually tells how your opponent beat you is missing, just gone. I wanted to file a bug report but I could not reload the end game screen to see what my table ID was. Instead I received 'red-bar' errors. That inconvenience didn't matter, however, because the page to enter bug reports threw a bunch of 'red-bar' errors and crashed.

Not looking good. Definitely moving in the wrong direction. My guess is it's time for a bad news update on why things are broken.

[Clockpusher]

Funny I was just coming to find out the exact some explanations about why the Innovation UX is different (and worse)

[Evilgenius1979]

I have noticed and posted about various issues with the site and unfortunately never get any sort of response. The stability of the site of several weeks has been very poor. It's a glaring problem when feedback like this just goes unanswered. I didn't renew my previous premium membership as I wasn't prepared to contribute further until some decent feedback is given to the playerbase.

I don't want every post to be negative towards the team, as this is a fantastic platform. I would like it to be constructive and building towards a solution. I would be interested if official stats could be posted on the amount of income the site receives and the number of people supporting it so we could in reality see if there is a general issue around support and resourcing vs. the perception that little due care is taken in managing the site and the premium member contributions.

I reviewed the release notes section of the forum: https://boardgamearena.com/forum/viewforum.php?f=5
It is never updated, I think that says something as well.

[FrankJones]

My thoughts:

For $3.50 a month USD, I find the website offers a lot for very little:

1) 1400+ games to choose from.
2) No restrictions on starting or joining games.
3) End-game stats
4) Individual statistics and winners statistics and average statistics for all games.
5) Unlimited viewing of game replays of any game played by any player, ever.
6) Dozens of forums to discuss various topics.
7) Useful friend features to chat with players outside of games.
8) Different options for play: Arena, tournaments, ranked games, friendly games, solo games.
9) Elo leaderboards for lifetime rankings.
10) Achievement badges that I find appealing to chase after.
11) Responsive developers who actually address specific bugs and issues for specific games, a very different experience from my time spent on other platforms.
12) Hotseat mode, which makes it very easy to learn a game in a PvP environment rather than just Solo mode or viewing the tutorial.

For me, this is worth $50 a month easily, even as is with the various problems.

For $3.50 a month, without knowing the details of their finances, I'm guessing BGA is not swimming in revenue.
Those 12 things in that list cost time, money, resources.

The site would improve if it had more revenue, which in turn would generate more premium memberships, and allow them to charge more, which would in turn create an upward trend in all those this, thus creating a cycle or more members, more revenue, and more improvements.

Perhaps the site feels as though it cannot do much more with its existing revenue. But perhaps most players feel they cannot upgrade to premium or renew their premium without seeing improvements.

In a catch-22 of that sort, the onus should be on the company to take the first step, invest money, accept the risk, and use that investment well to ensure a return on the investment and the betterment of the product.

Addressing the years-long overall site suggestions with hundreds or thousands of upvotes would be a good place to start. That, and better server stability.

[BarnardsStar]

I think the site should seriously considering allowing users to contribute to the general code base. I don't mean open-sourcing the whole dang thing or even part of it, but rather... consider that quite a few of the 1400 games here were essentially donated by people who wanted to see implementations. I'm sure there are coders among us who would like to similarly help out on the base site, perhaps in exchange for extremely modest compensation (along the lines of "a month of premium for every ticket closed" or maybe "special flair on their avatar"). I think it would be reasonable to have to agree to an NDA sort of thingy in order to gain access, etc. And BGA staff would obviously have the last say on what contributions actually get accepted.

But you have an avid fan base, who are perhaps growing a little disillusioned, but who still want to see the site succeed and thrive, and would love the chance to do their part, fix a bug or two, and make the site a better place. Not everyone knows software engineering, and I'm aware I'm talking about a probably small number of people, but it's not zero, and they/we could help.

[EllieK]

Addressing the years-long overall site suggestions with hundreds or thousands of upvotes would be a good place to start. That, and better server stability.

Based on my experiences on BGA, it would appear that the game state database, the information that allows BGA to recreate a game turn by turn, has some structural issues that slow things down. On other sites you can select a game and a turn and the system can immediately show you the view of the game at that point. BGA's need to play 'through' the game until it reaches your turn, tells me that the structure of the turn table is very likely dependent on externalities that require resources to recreate. Considering that the 'Game Turn' is a fundamental data structure, I would assume that the process of rebuilding it is daunting. To quote the Mighty Mighty Boss Tones, "That's the impression that I get."

[FrankJones]

Yes, that is an interesting point. Especially since the speed varies by game. For some games, clicking "go to specific move fast" will go from start to move 376 in a few seconds; other games seem to take a lot longer.


Wouldn't it take up a lot of extra storage space to store the game state for every move number?

I assumed that's why the game replays work the way they do: Access the moves list and replay through them one-by-one to arrive at any desired state.

This is definitely a nuisance when overshooting the intended move and not being able to go backwards without going back to the beginning.

[GTSchemer]

Yes, that is an interesting point. Especially since the speed varies by game. For some games, clicking "go to specific move fast" will go from start to move 376 in a few seconds; other games seem to take a lot longer.


Wouldn't it take up a lot of extra storage space to store the game state for every move number?

I assumed that's why the game replays work the way they do: Access the moves list and replay through them one-by-one to arrive at any desired state.

This is definitely a nuisance when overshooting the intended move and not being able to go backwards without going back to the beginning.

About the replay speed, it depends on implementation. One of the prolific developers (Tisaac?) even wrote a guide several years ago about how to make replays faster when you develop a game module: https://bga-devs.github.io/blog/posts/a ... play-mode/

Speaking as a programmer, it could indeed take up a ton of space to store the state at every move number. As an example, let's take Twilight Imperium 4th Edition, an especially long and complex game. I will give approximate numbers but hopefully they convey the concept. Near the start of the game, the database manager reports 368 KB (Kilobytes) total storage used for that specific instance of the game. This means a single game, at that moment in time, takes only 0.36 MB (Megabytes) of space.

However, a 4-player game that ended recently had over 1500 "moves", in BGA terminology. These are moments when the state of the game might change slightly, notifications happen, whatever. Now 368 KB times 1500 = a massive 539 MB of space. Note that is Megabytes, for one single game.

Instead of storing this huge amount of data, BGA just tracks the notifications throughout the game so they can be replayed to construct the state. It does take time to replay though. I tried to make fast replay as fast as possible, but I just tested and it was about 46 seconds using "Go to end (fast)" for a 1500+ move replay of Twilight Imperium 4th Edition.

[FrankJones]

Yeah, half a gig to store the results of one game is not feasible.

I also consider 46 seconds to be pretty good for advancing 1500 moves.

[Respublik]

About the replay speed, it depends on implementation. One of the prolific developers (Tisaac?) even wrote a guide several years ago about how to make replays faster when you develop a game module: https://bga-devs.github.io/blog/posts/a ... play-mode/

Given your knowledge of this, is replay/are replays being implemented by the individual game developers or handled site-wide? In the past, I've submitted bug reports on replay speed varying depending on players' preferences during play, but I'm now wondering whether those should have been directed to the individual game bug sections instead..?