Player taking advantage of undo move to stall the game.

[paramesis]

One could make a relative privation argument for any problem someone raises on this site. The reporting system exists as the underutilized proper channel to raise this issue, and the votes this report is getting are an indication that at least a few other people are affected by it.

For any loophole or exploitable behavior that can be specifically defined, it would be better to have it patched by the site, or at least to put it on the radar. Relying on the moderation and reporting system will result in a lot of frustration and wasted time for both moderators and players, who will all have different ways of communicating about the issue.

[aaronkw]

For what it's worth, this bug has been reported and discussed in the past so I just wanted to post a reference to it here for visibility. https://boardgamearena.com/forum/viewto ... 45&t=25476 there is a known top ELO ranked player for Castles of Burgundy that exploits this constantly to stay at the top of the rankings.

I don't think it's a waste of time to fix. It absolutely undermines the platform. If someone can exploit this to never lose a game, what's the point of an ELO ranking at all then? Why even have it if anyone can exploit it to get to the top? Fixing bugs related to the integrity of wins and losses seems like a valuable use of time for developers of this platform.

[shrouded figure]

I'm surprised this only showed up now; I've commented on the possibility of players abusing the Undo feature since last year and no one said a thing.

https://boardgamearena.com/bug?id=13887

An easy fix could probably be for the inactivity timer to not count 'Undo' actions as an action, and continue to count for the 20 seconds until the timed-out player actually makes a real action.

[Sheba-leila]

Für mich ist es unakzeptabel wie es jetzt ist. Vor asmodee gab es das nicht.

[paramesis]

I'm surprised this only showed up now; I've commented on the possibility of players abusing the Undo feature since last year and no one said a thing.

https://boardgamearena.com/bug?id=13887

An easy fix could probably be for the inactivity timer to not count 'Undo' actions as an action, and continue to count for the 20 seconds until the timed-out player actually makes a real action.

The problem is that this doesn't just affect the framework undo, it is possible to exploit the 20 second last-chance timer in any situation where you can take a series of database actions and return to a previous state. For example, if you take a worker from Uxmal 2 in Tzolk'in, you can exchange resources back and forth indefinitely to avoid being skipped. Several games also have their own implemented undo functions so they can work in game states where multiple players are active or so you can take back one step at a time (Dice Hospital is an example of both), and there isn't anything that specifically identifies these to the framework as "Undo"

The solution I suggested in the report I posted is to not allow a player to cancel the 20 second timer more than once. That way you don't need to require any changes to game modules that have potential loops.

[shrouded figure]

The problem is that this doesn't just affect the framework undo, it is possible to exploit the 20 second last-chance timer in any situation where you can take a series of database actions and return to a previous state. For example, if you take a worker from Uxmal 2 in Tzolk'in, you can exchange resources back and forth indefinitely to avoid being skipped. Several games also have their own implemented undo functions so they can work in game states where multiple players are active or so you can take back one step at a time (Dice Hospital is an example of both), and there isn't anything that specifically identifies these to the framework as "Undo"

The solution I suggested in the report I posted is to not allow a player to cancel the 20 second timer more than once. That way you don't need to require any changes to game modules that have potential loops.

That's a very good point and a far better solution. I have upvoted your report.

[sourisdudesert]

This is a good point. We are working on a fix.

[sourisdudesert]

Fix we found:

First expel warning is 20 seconds. Second is 10s. Then 5s, 2s, 1s, and finally no warning at all (immediate expulsion).

Each time the turn of the player is really finished (= another player must play), this goes back to 20s.

Thus, any player who tries to build a "loop" will finally be expelled.

[Silene]

Fix we found:

First expel warning is 20 seconds. Second is 10s. Then 5s, 2s, 1s, and finally no warning at all (immediate expulsion).

Each time the turn of the player is really finished (= another player must play), this goes back to 20s.

Thus, any player who tries to build a "loop" will finally be expelled.

This sounds like it would lead to almost instant expulsion on games that have a lot of single actions.

It would go like this:
Expel-Action
--> 20sec timer starts
Player does 1 Action within 2 sec
--> 20sec-timer is cancelled after 2 sec
Next Expel-Action
--> 10sec timer starts
Player does next Action within 2 sec
--> 10sec-timer is cancelled after 2 sec
etc.

So the Player could be expelled within 5 sec or so.

But it would work if the timerwould keep running down while you'remaking moves, so that the 2nd Expel-Action would be possible only after the 20sec. That would give about 40 sec for finishing your turn. Still hard on some long-turn-games but propably doable.

[Sheba-leila]

Es gibt laut BGA 3 Spielgeschwindigkeiten (was nicht stimmt) sind alle uf langsam ausser die alten Spiele vor asmodee. Ich verstehe nicht wenn es 3 Stufen gibt ein Spieler der lange braucht in ein Spiel mit schneller Geschwindigkeit einsteigt. Ignoranz? Oder wie soll man das sonst verstehen?