[cloa513]

You see lots of funny events like claims that should not be accepted. Unbelievably bad plays. I don't blame Vugraph operators being volunteers and the software is awkward like you can't correct a mistake in the claim or play easily. Certainly there bidding misunderstanding but some events defy that understanding. All you can think if it real then it is extreme time pressure that drives

For example I saw a which something like this- messed who was dealer it was South. and hearts broke nicely and then East in the only way of going down lead to Q in West and then back to the king and then they claim to make North held J10x in spades (with South naturally holding a normal good opening card with 6 diamonds AQJxxx .

---

Any chance there will visual and verbal analysis software to greatly automate the process that will do away with the operator error?

[barmar]

Sometimes the play goes very quickly and the operator can't keep up or makes a mistake. If they lose track of the individual plays, they will just wait until the hand is completed and use "Claim" to set the final result.

A number of years ago we experimented with a computer-vision setup that replaced the human computer operator, but it didn't work very well. CV software has improved, and AI could probably help it, but I don't think there are any current projects to attempt this.

[mycroft]

One joy of the current delayed vugraph is that if they make a mistake (in the auction, say) and it's too late to fix when they notice it, the only correction that can be applied is a "claim" with a "correct score". Before, same problem, but they'd have 50 people commenting that "you have an extra/missing pass in there" and were less likely to get caught by it.

I assume that's not something BBO doesn't know, in spades, since the New Orleans trial. How it gets dealt with is yet another question needing to be answered in the process of creating "delayed vugraph".

[johnu]

barmar, on 2023-April-06, 11:40, said:

Sometimes the play goes very quickly and the operator can't keep up or makes a mistake.

How many years of programming do you think it will take to enable undo for Vugraph operators?

[barmar]

johnu, on 2023-April-06, 16:03, said:

How many years of programming do you think it will take to enable undo for Vugraph operators?

It's been there forever. But if you're busy undoing and fixing, it's hard to keep track of the current play.

If the players stop to think, you can use that time to fix things up. But if the play is going quickly you're screwed, so you just wait for the end and enter the final result.

Undoes are used most often during the bidding. It doesn't go so quickly (especially when screens are in use). And you can see the entire bidding record.

[pescetom]

barmar, on 2023-April-14, 14:45, said:

It's been there forever.

But they still have to be accepted, which is kind of weird (I always wondered by whom).

[diana_eva]

pescetom, on 2023-April-18, 10:23, said:

But they still have to be accepted, which is kind of weird (I always wondered by whom).

They don't have to be accepted. Operator clicks undo, undo, undo like at a teaching table which takes effect immediately. Kibitzers see Undo accepted because that's the default message (teacher or operator unilaterally undo-ing is an exception)