Recording Problem?

[DenisO]

There seems to be a problem with recording the USBF Team trials to the archive in the BBO Bridge Library. The round of 16 is currently underway and we have nothing for session 3 or 4. There is also effectively nothing for session 2 (only a title - file is empty). Even the 1st session is flawed with lots of hieroglyphics in the file and the result bears no resembance to the score on the USBF site.

 

I wonder if there is a software problem? or is it an operator issue as we have the same very experienced operator who did the Cavendish but there were also problems in recording that - see my post on that topic.

 

Denis

[nikos59]

Hi Denis

 

Apparently the 1st and 2nd segment of the Round of 16 were

transmitted as a single file (30 boards). I was watching,

hence I have it but I also see that there is a file of 71K

under "... 1st session" in the archive page of BBO.

This one may well be corrupted and need manual editing.

I guess that the culprit here are the Japanese characters

of the Tokio Vugraph which for some reason spilled over

(commentary always spills over, it seems).

 

In the Vugraph page there is an entry for

2005 Vanderbilt SF ...

which in reality is the 3rd segment of R16, match

Simson v Lynch. The score is approximately the same

as the one in Usbf site.

 

The 4th segment seems to be lost forever, though

-unless a gentle member of the BBO community

has it on his/her hard disk and is willing to send

it to me at sarant-at-pt-dot-lu.

 

For my part, I have almost ready the Vugraphs

of the previous events (African Zonal, Zone 7,

Tokio) and I am going to upload them some time

today -or tonight.

 

Nikos

 

PS If you have 4th seg. or R-16, please send!

[Walddk]

We have asked the operator (open room is the master table, so there is really only one person involved) to label each segment properly. If they do, there should not be any problems with storing the data. However, as you may have noticed, they crashed on several occasions and that may have been a contributory factor in the missing info.

 

I don't know enough about that topic though. Uday?

 

Roland

[uday]

The disconnections by the client should not have affected anything about the saving. The save logic is simple: when a vug table closes (whether deliberately or via disconnection) save the vug file, overwriting the existing file as needed.

[DenisO]

Well Uday, these operators in Houston seem to have found a way to sabotage pretty well everything including the Save function. I'm watching boards 61 to 75 of the QFinals and they've just got about everything wrong that might be possible- wrong team names, wrong player names and wrong scores. On top of that, boards that I've just watched with bidding and play aren't being saved to my hard disk. I don't think it can get any worse :unsure:

 

Surely with five or ten minutes preparation before the broadcast, they could get the data and file naming sorted out?

[mrdct]

Surely with five or ten minutes preparation before the broadcast, they could get the data and file naming sorted out?

The problem is that tournament organisers quite often refuse to release the hand data files to the vugraph operator until just seconds before play starts (and in some cases after play starts). This can often put the operator into a bit of a panic where their priority is to get the show up and running by quickly clicking through all of the options and accepting the defaults for event name ("Untitled"), team names ("Team 1" and "Team 2"), carry-forward scores ("0-0"), etc.

 

My advice to vugraph operators is "don't panic". Even if bidding has already started when you get the hand data file, load the file up in a careful and considered manner as viewers will always be happier to miss part or all of the 1st board of a segment than have wrong team names and carried forward results for an entire segment. In practice, if the auction is over before you get time to load the hand data file, you can just enter the contract as a single bid (but please alert it as such so people will know) and carry-on. Similarly if the auction is over and play has already started, you can simply enter the result for that board in "movie".

 

The reluctance of some tournament organisers to allow the hand data files to be loaded earlier may be due, in part, to a design problem with BBO that displays the first hand in the hand data file to the world-wide audience immediately upon commencement of the presentation even when that is not the hand on the table about to be played. It would be much better if no hand gets displayed to the audience until the operator "redeals" to the first board to be played.

 

The other related BBO software deficiency is the inability to edit team names and carry-forward scores after a presentation has started. Fred has told me that such functionality has been developed, but its stability hasn't been properly tested yet; but hopefully we will see that soon.

 

At some, but not all, Australian events the players are required to be seated 5 minutes before the official start time to go over pre-alerts and to listen to any special instructions from the CTD. If you are in discussions with tournament organisers about vugraphing their event, I strong recommend that you encourage them to include such playing conditions in the regulations.

 

As part of the preparations for the BBO vugraph coverage of the World Youth Teams Championships from Sydney in August, the line-up sheets for the broadcast tables will be pre-printed with the precise answers to all of the start-up questions so that the naming of segments and teams will have a consistent look and feel for the audience. In the knock-out stages this will include the official carry-forward scores also.

[nikos59]

Yes, taking into account the level of competition, this is

probably the worst fiasco in BBO Vugraph history;

(if the records of some seaside festival are mangled,

no big harm is done; but of the US championships???)

[DenisO]

Thanks mrcdt for useful reply - perhaps Roland or Uday could convey the "don't panic" advice to the Houston operators because since I started the thread these broadcasts (as Nikos says) have become a fiasco - it shouldn't happen for such a major championship

 

There is no record of either match for segment 7 in Bridge Library - I saw (just before going to bed) that there was a hand rotation problem with Seg 7 for the Ekeblad/Cayne match -was it not sorted out? And for seg 8 we only have half a match from presumably two which were broadcast - I suspect both matches had exactly the same file name so one has been overwritten :rolleyes:

[mrdct]

Be careful using the "f" word. I'm sure the operators are doing the best they can and, believe me, it is not as easy at it looks. Unless you are onsite, you wont have the full story about other problems that they may be encountering.

 

I think in my time organising, operating, commentating and watching BBO vugraph presentations I've just about seen it all in terms of what can go wrong; and there are lots and lots of things that can go wrong that even perfect preparation won't be able to handle.

 

Like all bridge fans, I hope they can get it all sorted out for the rest of the event.

 

And, Nikos, what have you got against the Seaside Festival? I'm sure it's a fine and prestigious event. In any case, the USBF event is only selecting the USA's "B" team.

[DenisO]

Agree with you mrdct - shouldn't have said that . One gets a different perspective from 4000+ miles away. I can appreciate on-site problems but I find it difficult to understand why different matches are given the same file name and why operators don't have the match data (team and player names +scores) ready on a piece of paper to type into their laptop - or are they only given a few seconds notice as to which matches are to be broadcast?

[mrdct]

I'll see your 4000 miles and raise it to my 9000 miles from Houston!

 

In most serious events (such as the Seaside Festival) captains are required to complete line-up sheets. When everything is working smoothly one of the tournament officals can pass the line-up sheet, or a copy thereof, to the vugraph operator. But of course often the CTD has other issues on his mind like getting the correct boards out and chasing up late players.

 

Having the match data on a piece of paper at your fingertips is certainly the way to go, but still wont work if the players sit the wrong way.

 

In multi-match coverage that I've run, I've always included the table number of the match I'm covering to avoid any confusion with the LIN file names.

[shoeless]

Well this thread certainly gets me excited about being an operator. Groan!!!!!! Since last nights difficulties with hands (switched 90 degrees it seems) I am wondering who checks and when the integrity of the dup file disk for vugraph is checked. Time for you to write the operators training manual David - I will buy the first copy.

[inquiry]

Well this thread certainly gets me excited about being an operator. Groan!!!!!! Since last nights difficulties with hands (switched 90 degrees it seems) I am wondering who checks and when the integrity of the dup file disk for vugraph is checked. Time for you to write the operators training manual David - I will buy the first copy.

Actually, a very good operator user manual already exist, it is 10 pages long, clearly stated, even has screenshots of what the screen look like and what the operator should do. I can't imagine a better one is ever needed.

 

Ben

[mrdct]

I am wondering who checks and when the integrity of the dup file disk for vugraph is checked.

Sadly, a lot of the time the dup file is not checked by anyone. This is virtually always completely beyond the control of BBO vugraph operators as a number of event organisers and bridge administrators simply don't understand the technology or don't understand what things increase, decrease or have no effect on hand data security.

 

When I'm working with reasonable tournament convenors and directors, the BBO vugraph operator is considered a trusted member of the tournament staff and is given reasonably free access to hand data files. In such circumstances, I like to personally check that the dup file opens OK in BBO (whilst offline of course) and then cycle through each hand in the set to make sure there are no hands with incomplete data and, if I've been given one, check the first and last boards of the segment to the hand-record.

 

At the Australian National Open Teams, about an hour before play each day all of the hand data files for the day were copied onto my USB memory stick and were either checked by me or by the official scorer who had all of the hand-records on his machine ready to upload onto the web at the end of each session. The USB memory stick was then retained by the official scorer who gave it back to me a few minutes before each segment.

 

If for so-called "security" reasons you can't get the hand data file a reasonable time before the segment starts, the next best thing is to have a good talk to the person who is handling the data files (sometimes the CTD himself, but often the person in charge of board duplication) and make sure you are comfortable that he has appropriate process and controls in place to ensure the accuracy of the hand data files, particularly if the hand data files are being converted from another format and/or split into smaller files for each segment.

 

If you can't do any of the above, just pray!

[uday]

Would someone mind explaining the process by which the files with the hand records (the ones handed to the operator) are created? How is this tied into the duplication of boards?

[mrdct]

Would someone mind explaining the process by which the files with the hand records (the ones handed to the operator) are created?  How is this tied into the duplication of boards?

It may work differently from country to country, but in Australia the process is:

 

1. A computer dealing program approved by the national body generates hand data files which are not necessarily in the "dup" format used by BBO.

 

2. The hand data files are sent (usually by email) to one of the various individuals, bridge clubs and private companies that have dealing machines. For major events with a large volume of boards there are basically two private companies in Australia that do commercial board dealing.

 

3. The organisation to which board dealing has been contracted-out physically deals the boards, often usually using casual junior labour. For some events there may be board dealing on-site with only a day or two of hands dealt in advance.

 

4. Printed hand-records with Deep Finesse analysis are produced from the original hand data files. Usually at or about this time, and sometime using the software that does the Deep Finesse analysis, the original hand data file (which I think in Australia is often a "bri" file is converted to a "dup" file which is what BBO needs. This a is point of the process where catastrophic failure (i.e. an event leading to cancellation of a BBO vugraph presentation) has been known to occur.

 

5. The "dup" files for the event, which are hopefully clearly named so different segments wont get mixed up, are saved on a disk and given to a tournament official who eventually hands them over to the BBO vugraph operator.

 

For smaller events, such as a single head-to-head knock-out final, the process can be streamlined in a number of ways. For example, the original hand data file can be generated in "dup" format and if the dealing machine supports that format (which many machine in Australia don't) the boards can be dealt from that file and the hand records can be generated from that file. In theory this method is fool-proof as the "dup" given to BBO will be identical to the one used to deal the boards and hasn't gone through any conversion process that may stufft things up.

[DenisO]

Nice to see big improvement for the semi-finals - for segments 2 to 4 we have recordings of both matches in the vugraph archive :P A pity about seg 1 where one match has been overwritten. I hope that the improvement is not by pure chance and that the operators now realise the importance of naming their files correctly. It seems such a shame to put so much effort by operators and commentators into a broadcast and then to needlessly trash it :(

[Gerben42]

Some insider information:

 

The best vugraph operator in the US (recommended by Fred) offered her services just for "expenses paid" but the USBF said they would manage. Guess they didn't. I hope they learned from this, also I can imagine that Fred will not be thrilled by all this...

 

On a side note, I think you can trust the operator with the board files before the match so he can check them. He is in principle on the TD's staff.

[mrdct]

Obviously I'm not at the venue, but based on my fairly extensive experience with running BBO vugraph presentations, the frequent shut-downs and restarts can only be explained by a poor internet connection onsite; possibly wireless.

 

It may well be that the majority of the problems have nothing to do with the BBO operators but are more related to venue selection.

 

Can anyone who is actually onsite give me a detail description of the entire set-up.

[uday]

I'm not onsite, although I now wish otherwise.

 

The playing site is a hotel. The operators use hotel-supplied wireless connections. The connections are sometimes spotty.

[mrdct]

The operators use hotel-supplied wireless connections. The connections are sometimes spotty.

Do we know if a single wireless connection is being shared via a switch or hub to the various vugraph operators at the tables, or does each vugraph operator have its own wireless connection to the hotel's access point(s)?

 

In my experience, wireless generally only delivers a consistently stable internet connection for the 3+ hours straight you need for BBO when you actually have line-of-sight to the access point. Some venues will have better quality network infrastructure than others, but it's always a good idea to check first.

[sdoty]

You'll have to excuse me for mushing responses to various posts into one reply:

 

1. In Houston each of the vugraph computers was connecting to the hotel's wireless service independently. They didn't have much (if any) control over the signal strength/reliability; the tables were in individual rooms along a hallway on one of the guest floors, and they had to make do with whatever the signal was like in those rooms. When the connections turned out to be so spotty the first day, Fred thought that dial-up connections might be a better arrangement; unfortunately there wasn't really a head operator in place to implement that change.

 

 

2. The broadcast from the USBF Trials in Houston was staffed entirely by inexperienced volunteers (well, except for the first day when I was there). I don't say inexperienced in at all a negative way, I simply mean that for most of them this was their first ever broadcast. I believe there was a relatively high turnover rate since many individuals only worked for a few sessions -- and quite reasonably so, given that they were volunteering their time to do vugraph when they could have been playing at the concurrent Regional.

 

One thing that it's important for us to keep in mind is that it's extremely difficult (some might say impossible) for people to anticipate problems that they are simply unaware of.

 

Most people are blissfully ignorant (and again, I don't mean that negatively) of the technical side of the broadcasts... the hand logging, the file duplication, the ramifications of making an error when setting up the initial table specifications. Even an experienced Vugraph operator (I use that term loosely, since one can be in charge of organization/set-up and leave the actual play entry to other individuals) can run into problems; it takes quite a bit of knowledge to resolve anything more complicated than a lost internet connection successfully while a broadcast is in progress.

 

The staff at Houston put in a great deal of effort and, in my opinion, salvaged an impressive amount given a) the cascade of problems they had to deal with and b.) their newness to the undertaking.

 

At the Nationals the files go through one of the head tournament directors. I'm not positive that he checks the files beforehand, but I would say that at those events he's the one in a position to do so. That has certainly not been a responsibility shouldered by the operators. Anyhow, no, he was not working at the Trials in Houston, and there wasn't anyone else filling that position. Hence.... there were a few issues that might have been otherwise avoidable.

 

I do regret if the USBF organizers were unaware of just how much goes on behind the scenes of a broadcast; hopefully there will be a trained chief operator in place at next year's trials.

 

 

3. QUOTE (uday @ May 19 2005, 10:08 AM)

Would someone mind explaining the process by which the files with the hand records (the ones handed to the operator) are created? How is this tied into the duplication of boards?

 

I think mrdct gave a good overview. The only difference is that here (if my understanding of the process is correct, which is a fairly safe bet but not 100%) everything stems from the same original .dup file; the hand records and board duplication (which the ACBL does 'in-house' with duplicating machines) are based on that file, and the vugraph broadcast loads the boards from that same file. As described at the end of mrdct's post. :D

 

 

4. In general no, the operators here aren't considered in the same light as tournament staff (unless of course they are actually a director). By and large the operators are charged with showing up a bit before gametime and broadcasting from their table, and that is the extent of it. Which is about right, since with the exception of me they're generally new each tournament. My point is that there has not traditionally been a dedicated 'chief vugraph operator' at ACBL events to handle all of the various aspects of organizing/running a broadcast.

 

Susan