Long robot delays

[dartagnan9]

I experience a systematic pattern of bidding/play delay in the robots, and I'm sure it can't be just me, because of the pattern.

 

- If robot in any direction is the declarer, there will be multiple long pauses doing absolutely nothing - sometimes it seems like 30 seconds. And these pauses come from both offense and defense.

- However, if I 'm the declarer, there are significantly fewer pauses from all players.

 

These things are computer programs. There's no earthly need for them to take 30 seconds to figure out a bid or play. And, in any case, that's not what's happening, because sometimes there'll be that long delay even when they only have one possible card-play!

 

Especially when playing the race, it's extremely infuriating. You see two minutes left on the clock, and you're in a contract that will likely give you the tournament. Then you watch as a defending robot hesitates for 30 seconds over whether to play the 2 clubs or the 2 clubs (clubs were led, and he only has one!) You anxiously look at the clock ticking away, and, there it goes - hand is unfinished and you don't win.

 

It would be nice if this could be explained by the developers. The only likely explanation I can imagine is that the delays are programmed. The play speeds up dramatically in the second half of each hand.

 

When I've reported this in the past, I've been told it's due to delay on my side of the network. But that wouldn't explain the pattern I see. Why does it always delay dramatically, but only in the first half of the hand? And why is the pattern different, depending whether a robot is defending against me, or against my robot partner (this is where the long pauses come in.?)

 

If a developer is reading this, please give a response. Otherwise I'm going to get kicked out of my own house for screaming at my computer :)

 

Keith

[arigreen]

You are saying that the longest pauses occur when the human is dummy, including the case when there is a singleton to be played? This seems unlikely to me. Has anyone else experienced unusually long pauses when human is dummy?

[Bbradley62]

Has anyone else experienced unusually long pauses when human is dummy?

Yes. Of course, those pauses are frequently followed by rapid-fire playing of the rest of the hand. Since I only play duplicate games, not races, the pause is not long enough to bother me, but I'd say that 15-20 seconds is not uncommon.

[bid_em_up]

Has anyone else experienced unusually long pauses when human is dummy?

All the time.

[jamegumb]

Has anyone else experienced unusually long pauses when human is dummy?

Yes. Of course, those pauses are frequently followed by rapid-fire playing of the rest of the hand. Since I only play duplicate games, not races, the pause is not long enough to bother me, but I'd say that 15-20 seconds is not uncommon.

I've always thought that the pause was to determine how the rest of the hand was played. That is, the long pause is almost like a 15-second switch to "auto pilot" for all of the robots.

[dartagnan9]

Well, it would be nice if a developer could explain it. It's certainly true that it's only in the first half, before it speeds up faster than the eye can see! But I doubt it's because it's trying to figure out the play. Because it also happens when the GIB defender is humming and hawing about playing a card. He doesn't have nearly as much to figure out as the declarer, yet can take just as long to decide on playing in a situation where it obviously doesn't matter which card he plays!

 

Thanks, Keith

[barmar]

GIB doesn't know that it doesn't matter. It still performs a full simulation of each possibility. It only figures out that it doesn't matter after doing all the computations and getting the same results for each card.

[655321]

I just had an infuriating robot event (Best Hand $1 Event #3299) where the robot stopped for more than a minute on several occasions, once deciding whether or not to follow suit with its only remaining trump, another time after I had drawn trumps and also played another round so the robots could see each other show out, and was cashing winners from dummy (dummy only had winners).

 

Are there server issues at the moment that could cause this?

[uday]

Not that we know of; plenty of robot capacity, etc. How's your connection to the site in general ?

[655321]

No connection problems at all in general.

 

I just played another robot event where the robots played like speed demons! The earlier slow event was just one of those things I guess.

 

(Both times I was using the web version)

[dartagnan9]

I don't mean to be overly feisty, but I'm convinced it has nothing to do with the user's network. While the GIB is hesitating to play, I can often see somebody else's score increasing. I couldn't receive that information without network latency, but then not receive the information that makes the card play.

 

Also the dynamics of how this happen also argue against it being anything to do with network latency. It never ever happens once the play has speeded up. Only in the first half of the hand - but even their, it's frequently not due to any difficulty in making a choice of play when the play is very obvious.

 

Can I suggest that the developer try it himself? I don't know if the server is local to you, so suggest playing where it's not - i.e. at a cafe, or something.

 

Are you able to capture debugging information about the time each GIB play takes? That would show you the same thing. Maybe that would help understand what's happening while it delays.

 

Thanks, Keith

 

P.s. I just want to reiterate: I'm convinced this is a software issue, not a network issue. Or, at the very least, if it's a network issue, it's not dependent on the user's network, since I experience the same delays even on a very fast T1 line at a university. Same delay whether I'm playing on slow WiFi at Starbucks, or my much faster connection at home.

[peachy]

You are saying that the longest pauses occur when the human is dummy, including the case when there is a singleton to be played? This seems unlikely to me. Has anyone else experienced unusually long pauses when human is dummy?

I see completely unnecessary LONG pauses very often when I am dummy.

I use the old version and have excellent connection.

[JoAnneM]

A little off topic but how do I see other players' scores while I am playing?

[jillybean]

Has anyone else experienced unusually long pauses when human is dummy?

Yes when playing with 2or3 GIB at a table.

[jdonn]

No connection problems at all in general.

 

I just played another robot event where the robots played like speed demons!  The earlier slow event was just one of those things I guess.

 

(Both times I was using the web version)

I also have very inconsistent results, sometimes it pauses for a minute or two (literally) on a single card, sometimes it fires through the entire hand practically instantly.

[TimG]

I do find that GIB pauses when declarer (presumably to think things over) and that he play post-pause often goes very quickly. I seriously doubt that I have ever experienced a 60+ second pause, probably never even a 30+ second pause.

 

I don't doubt that such pauses have ever occurred. But, I suspect that the perception of a long pause is often influenced by the otherwise usually speedy play of a robot game.

[calm01]

Uday, Ari and team.

 

The large potential value of this forum to Bridgebase is again demonstrated by the willingness of the members to spend time, effort and risk of embarrassment communicating with the support team. And the quality of that members communication is of a high standard.

 

Whatever the outcome of this particular problem, I know how difficult it can be for one group of people to have empathy with another groups experience.

 

There is always a difference in:

 

- viewing point in space,

- viewing point in time,

- knowledge,

- roles,

- effort required to communicate,

- openness to being wrong and authority required to communicate,

- ...

 

This is probably the support teams best recent opportunity to demonstrate rapport with its forum members.

 

Fail and the forum world will judge unfairly - and the forum feedback will diminish; succeed and the world will love you for a while - and the forum feedback will improve further.

 

I am betting you will select the latter path as your future prospects partly depends o the nature of your rapport. And rational support teams usually act in their long term best interests.

[jdonn]

I do find that GIB pauses when declarer (presumably to think things over) and that he play post-pause often goes very quickly.  I seriously doubt that I have ever experienced a 60+ second pause, probably never even a 30+ second pause.

 

I don't doubt that such pauses have ever occurred.  But, I suspect that the perception of a long pause is often influenced by the otherwise usually speedy play of a robot game.

FWIW I can guarantee I have had pauses as long as 2 minutes before, although it's not common.

[jillybean]

Just now I am playing with 3 GIB at a table in the MBC

GIB is playing trump from hand, A then K, Qx is left on the table at this stage - pause......... eventualy x is played. I dont know how long, but a noticeable delay.

Earlier today I gave up playing with GIB because of this kind of delay.

[jillybean]

I am still having problems with GIB. Has BBO responded to the complaints in another thread, I may have missed a response. Or has the problem gone away for others?

[dartagnan9]

The problem is still there, and I'm not aware that BB responded. I didn't expect them to, which is disappointing. However, I think something has changed - they're faster than they were, although still the same occasional meaningless pause (e.g. when playing the last card!) It's just not as bad as it was a few weeks ago, it seems to me.

[dartagnan9]

GIB doesn't know that it doesn't matter. It still performs a full simulation of each possibility. It only figures out that it doesn't matter after doing all the computations and getting the same results for each card.

Then shouldn't there be a simple coding step that says: oh, I only have one card in that suit. Therefore no need to do any simulations?

[jaapfr]

JaapFr

 

It is sure a server problem. I travel around the world. At home as in some hotels all bids and plays by robots are performed in a split second. Now here in Indonesia as an example all similar robot delays as you all mention (although connection is not too bad. It has to do with available bandwith and that can differ each second.

So or accept these pauses or increase your bandwith.

ps Indeed these pauses are rather frustrating:):(

[Bbradley62]

Looks like JaapFr's server causes a 4+ year delay :-) Is this a necro record?

[jaapfr]

Looks like JaapFr's server causes a 4+ year delay :-) Is this a necro record?

 

lol got it:)