Tash Kalar move timer when undoing moves.

[Quinarbre]

By design an Undone move is the same as a done move in regards to the red bar timer. In other words, you can expend all of your total game time on one turn. Does this, in practice, nullify the turn timer for T-K?

Exactly.

Considering this is the only Undo (to my knowledge) on BGA...

Well, some games allow you to undo a move you haven't completed yet : for instance in Stone Age you can take back your workers if you've placed too many of them. I think that's dealt with on client side so it does not affect the turn timer. In TK it's the same when you're offered to move any piece on the board : you can select and de-select a piece without commiting to anything server-side. But for more complex moves, the rules and cards need to be checked so I have to make a server call.

In some other cases (Twin Tin Bots, Through the Ages) you can restart your turn altogether, but the two seem to work in a very different way and I cannot check how the red bar behaves for them right now (TtA is too long, and it's hard to find opponents for TTB).

[EllieK]

If you catch me online sometime and want to check the red bar timer in Twin Tin, let me know. I'd be happy to assist.

[Issity]

If players want to play game thinking thoroughly every move no matter how long it takes then they should choose slow or turn based game speed.

Short thinking time is not equal to not interesting or not competitive game (see 1 minute chess for example).

Both players should play on equal terms. When one player has more time for his move then the game is not fair.
Undoing move should not affect timer in any way.

I didn't play Tash Kalar so I don't know how undo and redo work but from Quinarbre's description it seems they are broken (in terms of time management).
If undo and redo is counted as a move and therefore resets the red bar then it adds you more time.
In each game player is given base time allotment and then additional time after every move. So if undo button works like a move then it increases time allotment, right? If not, then it shouldn't also affect red time bar.

[EllieK]

I didn't play Tash Kalar so I don't know how undo and redo work but from Quinarbre's description it seems they are broken (in terms of time management).
If undo and redo is counted as a move and therefore resets the red bar then it adds you more time.

Quinarbre says that for his implementation of T-K the red bar timer functions differently than in all other games. Once you understand that, it.....

something

[Kirian42]

A small note about the Undo in TK: the ability to undo is in fact in the actual rulebook--it's something you're supposed to be able to do in real life even!

I do wish other games had confirm/undo choices as well, for actions that can be done without revealing other information. Thanks for programming TK to have Undo, Quinarbre!

[CB Droege]

The way I see it is this:

The ability to expel people after the red bar is up is there to get rid of AFK folks: people who have completely vanished and are clearly nolonger playing the game, it is not for getting rid of 'slow' people, that's what the overall turn-timer is for. Therefore, ANY interaction with the board should reset that red bar timer, even if the interaction is essentially meaningless. It lets you know that the person is not AFK, and that's enough.

I think it should go further, and reset that bar anytime the player interacts with the BGA interface at all, even just for a chat.

[EllieK]

The ability to expel people after the red bar is up is there to get rid of AFK folks: people who have completely vanished and are clearly no longer playing the game, it is not for getting rid of 'slow' people, that's what the overall turn-timer is for.

I agree, CB Droege, that one should not expel someone in the middle of a game just because they're playing slowly. It's a courtesy I extend to others and I hope that others extend it to me. In my experience, I haven't played a game with anyone who just up and expels a player once the red-bar timers is full.

But are you saying that if a player has 36 minutes on the game timer, that you sit and wait 36 minutes before expelling them? Or do you wait for some minutes after the red-bar timer is expired and then expel them? If you don't wait the full 36 minutes, how long do you wait? When does one know that a player has completely vanished? If there is 36 minutes on the timer and a player returns after 34 minutes, did they completely vanish? Or did they just vanish for 34 minutes?

I see it like this:

Among the players, the red-bar timers is an agreed upon convention, whereby it is known to all, that if you are unable to reconnect or play, the others may assume you are disconnected and have completely vanished, no need to wait for the overall timer to expire. If I am playing a fast-turn game and I lose connectivity for several minutes, I assume that other players will begin to suspect that I have "completely vanished and [am] clearly no longer playing the game." They may wait a few minutes beyond the red-bar timer's end before expelling me, but I know (and they know) what we were getting into with a 'fast' game. If I experienced many disconnections I would play a slow timer, thereby providing ample time to reconnect.

The red-bar timer provides a convention agreed upon by all players for determining when a player has vanished and is no longer playing.

[Quinarbre]

Well with the new system, this discussion is moot.
Another question arises though, on which I'd like the community's input : what to do with zombies (players which are temporarily or permanently disconnected from a game) ?

In the current implementation, zombie players just skip their turn.
In High Form this is not a problem, since it always leaves room for the opponent to claim a task (there are never more than 2 destruction tasks offered simultaneously, the only really blocking offers would be 2 destructions + isolation).
In Deathmatch it's a totally different matter : if your opponent places no tokens, your score can't go up anymore.
When Melee will be ready this will be important, because the remaining players may want to continue, but they'll end up scoring in their worst color, which will probably be the one from the player who left.

Whatever form one is playing, the game in itself is never deadlocked : you can always summon creatures until your deck is empty (and you can do it pretty fast if you're the only one playing).

So what should I do ? Keep the statu quo ? Have zombies place random tokens ? I can have a different behaviour according to the form played.

[Sawyer Einherjar]

Considering the zombie will not summon his cards, as this will complicate the coding to a whole new level with all the card options, it might be annoying that the zombie will not gain any heroics.

Therefore when placing pieces at random i think it would be nice it places 1 common and 1 heroic each turn. Only heroics might be a bit to easy for gaining points, while only commons might make it to easy but also longer as you need 2 for one point. With the extra rule in melee where you should be able to still get a point for a single common if you have multiple of different colors you should still be able to score a nice 2 points of the zombie color in a turn. And maybe it would be nice to have the zombie use the flare effects: place an extra common/heroic piece or gain an action (to place another piece) that way if the two pieces placed get destroyed instantly all the time by the player after the zombie there is a higher rate of surviving pieces for others. Still in a 4p melee game the 3rd player after the zombie will have the biggest disadvantage.

Then where to place these is another point. If always adjecent to a previous placed piece it might become favourable for certain players due to position. But having them being placed as close to the center tile as possible might have the same problem. this latter might however prevent that a big blob of farmable pieces is created in a single spot. Instead pieces might get scattered around a bit more creating more balance for getting points from the zombie color while looking at it through melee glasses (and how i think it works based on the rules as i don't own the game myself yet ). By having the pieces placed in a more circular way around the center i think you can decrease the disadvantage of being the 3rd player after a zombie.

Anyway just some thoughts ^^

Cheers,
Sawyer

PS: just had another thought. MB you can consider that the zombie player discards one of his cards each turn (as if it was played) and then let it place a single common and a second piece for which the rank is based on the first card in the hand of the zombie player. That way there is a higher change for 1 common and 1 heroic (as i believe all decks have more heroic summons then commons) and a lesser chance for 2 commons.

Anyhow my coding skills are based on VBA so no clue what you can do with the BGA codes and how much work it will be ^^. i'm still amazed at how you've coded for recognizing all possible positions of the formations, i'm clueless on how'd you do that except writing lots of if then statments but that's probs inefficient .

[Quinarbre]

Using discard to simulate summoning and make the game progress is a good idea, but I'm also concerned with a player being temporarily disconnected and finding his deck depleted upon his return.
And most of all I want to keep the zombie code really straightforward, which is another reason why simply skipping remains a very good option.

Actually, coding for rotations and mirrors of the patterns was the easy part (it's a single page of code with matrix-based geometric transformations) ; coding all the effects and conditions was way more tricky.
The really messy ones were the Summoner and the Assassin : if you summon them on a square which fits two patterns, it becomes quite hard to tell what they can, must and can't do.