Bridgemate Accuracy

[mrdct]

Once the live scoring got going a few days into the WMSG, it was evident that quite a lot of Bridgemate errors were creeping into the live scorecards. I doubt this has anything to do with the Bridgemates themselves and was more an issue with the accuracy of the data entry on the units.

 

In Beijing were players entering the results themselves or were there recorders at the tables doing it?

[rogerclee]

In Beijing were players entering the results themselves or were there recorders at the tables doing it?

Results in the pairs and in the round robin of the teams (for the juniors) were entered by the players. Results in the knockout stages of the teams were entered by recorders.

 

The bridgemates were having problems during the knockout stages of the teams as well as the pairs. Not really sure what the deal was, but at my table, the recorder entered 3N= for +140 (as it was displayed on the screen) and sort of just shrugged.

[OleBerg]

Bridgemate is being used quite a lot in Denmark.

 

I have been TD in a number of smaller and larger tournaments. I have never experienced, or heard of, a bridgemate that gave a wrong result. It is always the players that have made a mistake.

 

Sometimes there is problems with the wireless connection, or the software supporting the bridgemate, but the never with the bridgemate itself.

 

And they are really efficient. For a 20-table seroius tournament, only one person is required to be TD, manege results and run with cards.

 

The handling of mistakes made by players entering incorrect information is so easily solved (with the danish software), that penalty points are not ever given for these mistakes. The same goes for players sitting in the wrong direction, or playing wrong cards etc. (The usual 60/40 rules apply, when a board cannot be played.)

 

Bridgemates are a big plus. Only heard complains are that they are ugly.

[skjaeran]

100% agree with what Ole Berg wrote above - we've got exactly the same experience in Norway where Bridgemates are used extencively.

[mrdct]

The reality is, however, that they were less than successful in Beijing.

 

I've not used Bridgemates before, but with other electronic scoring units that I've used one of the NS players enters the score and then one of EW players has to check it and press OK. Under that method, I've hardly ever seen any errors.

 

I think if recorders are being used they really should be cross-checking every board with what is on an actual player scoresheet. Moreover, when an event is down to the final there should be ample resources to cross-check player scorecards, BBO results and Bridgemate results in real time.

 

It was really unfortunate that in the final 16 boards of the seniors final there were no less that three errors resulting in USA's score being represented about 17 imps too high.

[syl]

As far as I have seen it crosschecking IS required using Bridgemates.

Thats another issue that some players are lazy to control and let NS to do it alone, others quicly hit double OK so "doing" the input and the checking the same time.

[NickRW]

I've not used Bridgemates before, but with other electronic scoring units that I've used one of the NS players enters the score and then one of EW players has to check it and press OK. Under that method, I've hardly ever seen any errors.

 

I think if recorders are being used they really should be cross-checking every board with what is on an actual player scoresheet. Moreover, when an event is down to the final there should be ample resources to cross-check player scorecards, BBO results and Bridgemate results in real time.

 

It was really unfortunate that in the final 16 boards of the seniors final there were no less that three errors resulting in USA's score being represented about 17 imps too high.

I'd be very wary of attributing this problem to the Bridgemates. I've never had recorders do the job, but I have used them with N entering the score and E checking - providing that both parties take care to actually do what they are supposed to, there is no problem.

 

What can be a problem is people not using them as they should be - but that would potentially be a problem whatever hardware/software choices were made.

 

Nick

[blackshoe]

Heh. I will be very surprised if a session occurs - using written scoring, either pickups or travellers - where I don't get at least one scoring error. People seem to think the sole purpose of the "check the score" procedure is to give the TD a souvenir of their initiials. :wacko:

[Fluffy]

Bridgemates worked quite well, they are clearly a step forward for duplicate bridge. I only miss not being able to see what other tables have scored, althou I have played some tournaments where bridgemate showed them, so I guess this is just a matter of configuration (and learning how to use it).

[shintaro]

:)

 

Here In the England North enters the Contract and then + - or = then EAST MUST check before ok 'ing we have had no Major problems that I know of

 

:wacko:

[skjaeran]

The reality is, however, that they were less than successful in Beijing.

 

I've not used Bridgemates before, but with other electronic scoring units that I've used one of the NS players enters the score and then one of EW players has to check it and press OK. Under that method, I've hardly ever seen any errors.

 

I think if recorders are being used they really should be cross-checking every board with what is on an actual player scoresheet. Moreover, when an event is down to the final there should be ample resources to cross-check player scorecards, BBO results and Bridgemate results in real time.

 

It was really unfortunate that in the final 16 boards of the seniors final there were no less that three errors resulting in USA's score being represented about 17 imps too high.

I was open team coach for our open team in the European Teams Championships in Pau this summer, and as such recorded about 2/3 of our teams matches.

 

I remember having 2-3 results in the wrong direction (of something like 400 boards). This surely was due to an error somewhere between my input and the results being imported to the result file. I can say this for sure, since we entered the whole auction and all the cards played up to a claim or the last trick (unless getting lost halfway through the hand), and know for sure that i had these things right.

[shintaro]

;)

 

Hi

 

I cannot believe the Bridgemastes can be at Fault they work on the GIGO principle; far more likely is the Human error either with the Scoring Software or Scoring input

 

 

:rolleyes:

[dburn]

I cannot believe the Bridgemastes can be at Fault they work on the GIGO principle; far more likely is the Human error either with the Scoring Software or Scoring input

I was also in Pau, and had only one occasion to query a result - a board had been scored as flat when in fact there had been a swing of 15 IMPs.

 

Investigation revealed that the official scorers were checking the inputs from the Bridgemates as they arrived at the scoring computers. Seeing that England had recorded 600 in both rooms when the contract was 3NT, the scorer assumed that the compass position had been entered incorrectly in one room and amended the result accordingly. In fact, both original entries were correct - the contract had been made in one room and gone down six in the other.

 

Whereas it is certainly possible for machines to make mistakes, in the vast majority of cases errors are due to what is known in the technical support business as PEBCAK - either at the card table or in the scoring room.

[Walddk]

The trouble in Beijing was that in many matches they used inexperienced local recorders, some of them not even bridge players. One of the common errors was that a lead was entered as the contract.

 

Example: 4♥ by West was entered as 3♠ by North, because the lead was ♠3. For obvious reasons that caused a few misscores and, as a result, as many corrections.

 

Roland

[whereagles]

I use bridgemates at the local club. They do fine, though there are some annoying issues to solve before they start working :)

[Codo]

At our local club we use them too and there had beebn problems.

 

I do remember sone hands where 3 NT scored 400 for some N/S but 600 for some others.

The bridgemates have to "talk" to some other programs, so my guess is that problems can arose there.

 

But I join the chorus that 99 % of all mistakes are human and that bridgemates are a great improvement.

 

And for the scores of the other tables: This is optionla and can be allowed or forbidden from the TD before the round starts.

[Gerben42]

Chtonic would not be surprised that it's the humans creating the errors.

[matmat]

someone needs to make a system where two players, say, North and East have to enter the score.

If the scores don't match, they need to try again.

 

(two keypads, not one with a signoff...)

[shintaro]

:)

 

Dburn as a member of our L&E would know better than I as to whether we have had ANY scoring problems with the Bridgemates we use in EBU Land or whether ANY errors have always been down to a combination of 'Input Error/Operator Error or Scoring Programme Error'

 

I cannot see ANY computer/keyboard doing anything other than what it is told to.

 

;)

[mrdct]

There's no point blaming operator error if there aren't processes in place to ensure accuracy. I would've thought that when a World Championship is down to the final, there would be resources floating around to check and recheck each score as they go.

 

If we are going to rely on these things for accurate live scoring, there must controls in place to ensure that when West presses "OK" the result on the screen is indeed the correct score. The obvious control is EW get penalised 3 IMP for any incorrect scores which should be a good incentive to not press "OK" unless it's right.

[MFA]

I think the Bridgemate desperately needs a feature, where one could easily check all the scores of the session.

 

Our team in Beijing had a long, continuous struggle after the matches with correcting mistakes with the Bridgemates. Just as we had had in Pau. This is not fun. If it were possible to run through the scores at the end of the session, things would be much easier.

[MFA]

The obvious control is EW get penalised 3 IMP for any incorrect scores which should be a good incentive to not press "OK" unless it's right.

... and what would the incentive then be to NS to enter the correct result in the first place? :rolleyes:

 

Introducing penalty points is a delicate matter. The margin in the women's final was 1 imp and in the seniors' 2 imps just to give an example. The world champions shouldn't be just the ones who are better at producing the correct running score for the public.

[mrdct]

Perhaps the Bridgemates could be reprogrammed so that NS enter the result and click OK and then hand the unit to EW who enter the result and click OK. The score will only validate (and update itself on a live scores website where applicable) if NS & EW enter exactly the same contact, declarer and tricks. If EW enter something different, the result is cleared and the unit goes back to NS to try again.

[cherdano]

If we are going to rely on these things for accurate live scoring, there must controls in place to ensure that when West presses "OK" the result on the screen is indeed the correct score. The obvious control is EW get penalised 3 IMP for any incorrect scores which should be a good incentive to not press "OK" unless it's right.

There must be a way to describe mrdct that is worthy of being added to Josh's signature but I can't quite come up with it...

[mrdct]

I assume you are looking for a descriptor such as "strikingly handsome".

 

Don't get me wrong here, I think Bridgemates and other handheld scoring units are great but through one or more of player apathy, recorder inexperience, counter-intuitive software and inadequate data validation, they were a failure at Beijing because at the end of the day bridge fans simply could not rely on the website scores. This may not be a big deal for many, but some of us were having a punt here and there, sometime in-play, and it makes it kind of hard when the live scores can't be trusted.

 

I'm simply searching for a better mouse trap.