How can BGA improve Real-Time matchmaking?

[Perduicitte]

You can see a live list of all real time & turn based games here:
https://en.boardgamearena.com/gameinprogress

Tremendous, thank you!
So... data supports my impression. I don't know how to post a graph in the forums, but the webpage's information can be easily exported to excel and graphed. It looks more like this:

I
I
IIII:::::,,,,,,,,............................................................... a drastic drop.

4 games > 100 tables
6 games > 50 tables
13 games > 25 tables
24 game > 10 tables
33 games > 5 tables
126 games >1 table
409 games have 0 table, no play.

Granted, this is a Thursday morning in North America, afternoon in Europe. People have, presumably, more available time on weekends to play at all, and/or play longer games.

Cheers,

me, with the help of Smeg's direction towards data (yay!)

PS: Stuff by Bez, thanks for generating ideas! Yay!

[Perduicitte]

@Stuff by Bez, I re-read your comment and have one insight:

I honestly wish that there was a UX person employed at BGA. As an example, I hate the way that there are at least 5 totally different ways of placing stones in very similar abstract games. Do you get confirmation? Do you have to click at the top, or again on the piece? Do you get a short time limit? Whilst the time limit is going, do you click to confirm, or cancel?

These types of improvements have been well handled, in my experience, through the bugs reports / suggestions system https://boardgamearena.com/bugs

Cheers!

me

[Fletcheese]

The reason that different games often have very different UX is very simple - they were made by different people. The BGA framework is very loose and you end up writing a lot of custom code for any game. That's what allows for such accurate rules-driven games on the platform. Should the framework have more boilerplate for things like placing stones? Perhaps but realistically it will likely never happen.

[Jellby]

So... data supports my impression. I don't know how to post a graph in the forums, but the webpage's information can be easily exported to excel and graphed. It looks more like this:

I
I
IIII:::::,,,,,,,,............................................................... a drastic drop.

4 games > 100 tables
6 games > 50 tables
13 games > 25 tables
24 game > 10 tables
33 games > 5 tables
126 games >1 table
409 games have 0 table, no play.

Yep, pretty much an exponential decay, as expected for many natural and statistical phenomena. Interestingly, the "lifetime" seems to be close to 42 , i.e. the tail distribution is roughly proportional to exp(-x/42).

[Stuff by Bez]

The reason that different games often have very different UX is very simple - they were made by different people. The BGA framework is very loose and you end up writing a lot of custom code for any game. That's what allows for such accurate rules-driven games on the platform. Should the framework have more boilerplate for things like placing stones? Perhaps but realistically it will likely never happen.

Why not?

If there are games that share similar mechanisms (e.g. trick taking games, or abstract tile-placement games) why not have some sort of 'best practices' guide to ensure that developers know what is expected?

[Fletcheese]

The reason that different games often have very different UX is very simple - they were made by different people. The BGA framework is very loose and you end up writing a lot of custom code for any game. That's what allows for such accurate rules-driven games on the platform. Should the framework have more boilerplate for things like placing stones? Perhaps but realistically it will likely never happen.

Why not?

If there are games that share similar mechanisms (e.g. trick taking games, or abstract tile-placement games) why not have some sort of 'best practices' guide to ensure that developers know what is expected?

There are best practices guides. Have you done much software development? BGA has hundreds of independent developers each with completely different experience levels in the BGA framework and development as a whole. Learning a new framework is a significant undertaking even when you have a lot of past experience. You are constantly trying to make thing work, learning alternate ways you could have made them work, and determining which option may or may not have been better based on numerous factors that you might not even be aware of until much later. Development is by necessity complicated. The goal of a framework is to make it less complicated, but that's not always so simple.

If they want to update the framework, then any update needs to work with historic games or they need to introduce some legacy versioning where old games are running on a different version than newer games. Plus any such update costs time and money. It's really hard to fight against "if it aint broke, dont fix it". "If it aint ideal, improve it" is not a common cliche for a reason.

[ufm]

You can see a live list of all real time & turn based games here:
https://en.boardgamearena.com/gameinprogress

This is not a valid data as real-time games end faster than turn-based (which is obvious).
For example, Piraten Kapern has been played 7500 times in real-time in 2023 May alone.
Turn-based games are less than half of that number (currently 3427).
So, 700+ games per day has been played in real-time.
Yet there are 0 ongoing real-time tables 'right now'.
(FYI, developers and publishers can check the number of games played from the audience statistics)

[ufm]

Anyway, being a subsidiary of Asmodee does not mean BGA can spend money as they wish.
They started to expand their staff list only recently.
For example: https://boardgamearena.com/forum/viewtopic.php?t=29161
I highly doubt investing in marketing would lead to profit instead of net loss, assuming they even have funds available for marketing.
Introducing new players to niche games by marketing for a long time is impossible as even many games advertised with front page news and social media posts die out eventually.

UI issues are practically unsolvable as Fletcheese has pointed out.
Also, don't forget most games are developed by volunteers. They are not obliged to maintain the game forever.

In contrast, I also think queueing real-time games have been worsened after introducing new lobby.
Even the total number of games played in BGA stays the same, the number of games played in most games are dropping as queueing multiple games or multiple settings in a single game is hard or impossible in new lobby.
Abolishing Arena mode is too far though.

[Jellby]

This is not a valid data as real-time games end faster than turn-based (which is obvious).

I think here we're only looking at real-time games, so what happens with turn-based is irrelevant (except for the minority of tables that start as real-time and are switched to turn-based).

For example, Piraten Kapern has been played 7500 times in real-time in 2023 May alone.
[...]
So, 700+ games per day has been played in real-time.
Yet there are 0 ongoing real-time tables 'right now'.

That's the main problem of the list, which is actually the same reason why comparing real-time and turn-based is not possible/reliable. The data in the list would correlate more with "total time spent on real-time tables" than with "number of real-time tables started", so long games are overrepresented with respect to short games.

[ufm]

This is not a valid data as real-time games end faster than turn-based (which is obvious).

I think here we're only looking at real-time games, so what happens with turn-based is irrelevant (except for the minority of tables that start as real-time and are switched to turn-based).

It is also relevant, as real-time mode is much more favored in some games (mainly trick-taking and draw-and-discard games).