Missing Scores

[Dave Flower]

Two weeks ago at Banbury (EBU) we tried remote scoring with results transmitted to a central computer from small units on the tables. Regretably, something went wrong, and the results for the last two rounds are missing, although our scorer is trying to recover them (and has some personal score sheets).

If all the results cannot be recovered, should average plus be awarded for the remaining scores ?

Dave Flower

[Balrog]

In article <20972383.1135.1336148537947.JavaMail.geo-discussion-
forums@vbuo17>, DavJF...@btinternet.com says...

>
> Two weeks ago at Banbury (EBU) we tried remote scoring with results transmitted to a central computer from small units on the tables. Regretably, something went wrong, and the results for the last two rounds are missing, although our scorer is trying to recover them (and has some personal score sheets).
>
> If all the results cannot be recovered, should average plus be awarded for the remaining scores ?
>
> Dave Flower

The person who programmed the devices should be fired. That sort of
thing can't happen if they're programmed competently. Each transmision
of a core should be similar to a banking transaction in that it either
succeeds on both ends or fails on both ends.

First of all, the devices on the tables should have enough memory (a
tiny bit) needed to record at least one entire session and should be
able to resend any number of scores until its memory is cleared. Send
once and forget is ridiculous beyond all comprehension.

Secondly, the server should acknowledge receipt of a score when one is
sent and the devices on the tables should expect the acknowledgment.
Thus, not receiving acknowledgement within a reasonable period of time
after sending should signal an error.

This is not rocket science. It's basic required property of a database
transaction called atomicity that was documented at least 35 years ago
and has been in use ever since.

[dak...@aol.com]

If all the results cannot be recovered, should average plus be
awarded for the remaining scores ?

Dave Flower

***
Shouldn't No-Play be the choice for the unrecovered scores?
Hate that thought, but that at least keeps the rightly scored
valuable.

[blackshoe]

I don't think that's legal.

[Barry Margolin]

In article
<12434666.2127.1336171584027.JavaMail.geo-discussion-forums@vbpz13>,

Probably true.

Although no matter what we do, the results of this game are going to be
pretty meaningless, if a large proportion of the pairs have artificial
scores for 2 rounds.

--
Barry Margolin
Arlington, MA

[Douglas Newlands]

On 5/05/12 2:22 AM, Dave Flower wrote:
> Two weeks ago at Banbury (EBU) we tried remote scoring with results transmitted to a central computer from small units on the tables. Regretably, something went wrong, and the results for the last two rounds are missing, although our scorer is trying to recover them (and has some personal score sheets).
>
> If all the results cannot be recovered, should average plus be awarded for the remaining scores ?

The manual for the combination we use (ASE8 + bridgemates), indicates
the scores are kept in 5 places and that it is possible to recover
using any copy. Maybe a close reading of the manuals is in order.

doug

[Bob]

Blackshoe argued that everyone gets A+. But the logic of these feel-
good scores is that they rob the people who don't get them. Usually
they don't notice if there is just one, but if you give too many, then
the unfairness can become obvious. If you are running a multi-site
event and your site is way over-represented in the winners, the other
sites are likely to notice. So you should give no play for the scores
you don't recover.

[derek]

On May 4, 6:49 pm, Balrog <brog...@rogoff-darrow.net> wrote:

> The person who programmed the devices should be fired. That sort of
> thing can't happen if they're programmed competently. Each transmision
> of a core should be similar to a banking transaction in that it either
> succeeds on both ends or fails on both ends.

Duh? Besides the fact that this _isn't_ as serious as banking,
whatever makes you think it's a programming error? I've seen it
happen, and it came down to user error.

All of your suggestions appear to be in use on Bridgemates.

[Dave Flower]

On Saturday, 5 May 2012 03:07:06 UTC+1, derek wrote:
> On May 4, 6:49 pm, Balrog

On Saturday, 5 May 2012 03:07:06 UTC+1, derek wrote:
> On May 4, 6:49 pm, Balrog

For the record, we were not using Bridgemates, but another product (I don't know the name) which uses touch-sensitive screens

Dave Flower

[Bertel Lund Hansen]

Balrog skrev:

> The person who programmed the devices should be fired.

People who jump to conclusions, should be fired.

--
Bertel, Denmark
http://bridge.lundhansen.dk/

[Balrog]

In article <9f029ecd-20f1-4907-b0ba-
a88d84...@o6g2000vbz.googlegroups.com>, de...@pointerstop.ca says...

Duh! Of course it's not as serious as banking but the same basic
principles apply. I wouldn't describe this situation as the result of a
programming error. It's more like a hopelessly bad software design and
implementation effort.

We're all human beings and human beings make mistakes. That's just a
fact of life. Robust software products allow users to recover from
mistakes. The simplest example is the "undo" function present in pretty
much every editor currently in existence.

[Balrog]

In article <4fa54967$0$295$1472...@news.sunsite.dk>,
kanon...@lundhansen.dk says...

>
> Balrog skrev:
>
> > The person who programmed the devices should be fired.
>
> People who jump to conclusions, should be fired.

My conclusion is based on 32 years of experience as a software technical
writer and programmer. This disaster is the result of a hopelessly bad
software design and implementation effort and no competent software
professional in the world would disagree with that.

Now who's jumping to conclusions?

[Bertel Lund Hansen]

Balrog skrev:

> My conclusion is based on 32 years of experience as a software technical
> writer and programmer.

I don't see how that knowledge can help you make the right
diagnosis about something that you have had no chance of
investigating, and you didn't ask even one follow-up question.

[Steve Foster]

Dave Flower wrote:

> For the record, we were not using Bridgemates, but another product (I
> don't know the name) which uses touch-sensitive screens

That sounds like Jannersten's BridgeScorers (red plastic units with a
flip cover?).

--
Steve Foster
For SSL Certificates, Domains, etc, visit.:
https://netshop.virtual-isp.net

[Steve Foster]

We don't have sufficient information to draw that conclusion.

> Now who's jumping to conclusions?

You are (still).

[Steve Foster]

Dave Flower wrote:

> Two weeks ago at Banbury (EBU) we tried remote scoring with results
> transmitted to a central computer from small units on the tables.
> Regretably, something went wrong, and the results for the last two
> rounds are missing, although our scorer is trying to recover them
> (and has some personal score sheets).

From a cursory check of the BridgeScorer web site, the results should
probably still be in the individual scoring devices.

> If all the results cannot be recovered, should average plus be
> awarded for the remaining scores ?

Yes. L12C2.

[Steve Foster]

Bob wrote:

> On May 4, 12:22 pm, Dave Flower <DavJFlo...@btinternet.com> wrote:
> > Two weeks ago at Banbury (EBU) we tried remote scoring with results
> > transmitted to a central computer from small units on the tables.
> > Regretably, something went wrong, and the results for the last two
> > rounds are missing, although our scorer is trying to recover them
> > (and has some personal score sheets).
> >
> > If all the results cannot be recovered, should average plus be
> > awarded for the remaining scores ?
> >
> > Dave Flower
>
> Blackshoe argued that everyone gets A+. But the logic of these feel-
> good scores is that they rob the people who don't get them.

The primary goal is restitution for the damaged pairs.

If we could do so without potentially damaging the field, that would be
really nice, but I don't know of a way to do so (and no A/A isn't it).

> Usually
> they don't notice if there is just one, but if you give too many, then
> the unfairness can become obvious. If you are running a multi-site
> event and your site is way over-represented in the winners, the other
> sites are likely to notice. So you should give no play for the scores
> you don't recover.

Your method unfairly penalises the competitors without results (and
isn't currently legal).

[Barry Margolin]

In article <MPG.2a0e18ea2...@news.giganews.com>,

Balrog <bro...@rogoff-darrow.net> wrote:

> The person who programmed the devices should be fired.

Maybe, but unless the devices were designed by someone at the club, I
don't see how the TD can make this happen.

And it still doesn't help him score this particular game.

[Dave Flower]

Spot on

Dave Flower

[blackshoe]

On Friday, May 4, 2012 10:05:54 PM UTC-4, Bob wrote:

> Blackshoe argued that everyone gets A+. But the logic of these feel-
> good scores is that they rob the people who don't get them. Usually
> they don't notice if there is just one, but if you give too many, then
> the unfairness can become obvious. If you are running a multi-site
> event and your site is way over-represented in the winners, the other
> sites are likely to notice. So you should give no play for the scores
> you don't recover.

First, the only thing I've said previously in this thread is that "not played" is not legal in this case. As for A+/A+, that's not a "feel-good" score, it's what the law requires. This business of an ArtAS "robbing" the other players is bullshit. What you're suggesting is that there is a concept (which is commonly called "field protection") that requires that the impact of a ruling at a particular table on the rest of the field be considered. There is no such concept in the law. Finally, as I said, "no play" is illegal, so no, you should *not* give that for the scores you don't recover.

[derek]

On May 5, 2:10 pm, Balrog <brog...@rogoff-darrow.net> wrote:
> In article <9f029ecd-20f1-4907-b0ba-
> a88d84741...@o6g2000vbz.googlegroups.com>, de...@pointerstop.ca says...

Except that you're jumping to conclusions without any evidence that
there's a software problem (and if you insist on claiming that your 32
years of experience give you the expertise to do so, I'll argue that I
have more experience than you). I've seen this happen (with
Bridgemates - admittedly the OP has said they _aren't_ Bridgemates,
but I expect all vendors to be able to meet Bridgemate's abilities, or
they'll never break into the market), and it was user error - it was
also recoverable, because all the data _is_ available. The mere fact
that this seems to be the first time the club's used the devices would
suggest that the scorer isn't fully conversant with the software's
capabilities.

I'll grant you that the software could almost certainly have used a
better technical writer - not because I know the particular devices
and their software, but because ime support documentation is usually
woefully inadequate, and when present is frequently written by
developers who don't actually understand users.

[Adam Beneschan]

On Saturday, May 5, 2012 10:18:12 AM UTC-7, Balrog wrote:
> >
> > Balrog skrev:
> >
> > > The person who programmed the devices should be fired.
> >
> > People who jump to conclusions, should be fired.
>
> My conclusion is based on 32 years of experience as a software technical
> writer and programmer. This disaster is the result of a hopelessly bad
> software design and implementation effort and no competent software
> professional in the world would disagree with that.

Of course, even with that in mind, maybe it's not the programmer who should be fired. A programmer with 32 years experience like you (and me) should probably be expected to think about things like how to prevent irrecoverable disasters when users make simple mistakes (of course, none of us can do anything about those determinedly idiotic users who make series of mistakes). But if the programmer involved had, perhaps, just a couple years' experience, then the fault is probably with his management. Sometimes management doesn't spell out the requirements well enough. Sometimes they're more concerned about getting something out the door quickly. I'm reminded of a Dilbert cartoon, when after he warns that the product has tons of problems and the documentation is a work of fiction, his boss suggests that the proper course of action is "Ship it and hope somebody writes a Dummies book about it". So who knows. I tend to agree with you that this is a flaw in the software assuming that the users didn't do anything horrendously stupid.

-- Adam

[Fred.]

My own conclusion, based on 34 years in distributed financial
systems and medical information systems, is that a rather small
percentage of system failures are due to incompetent or
slovenly work. Even given incompetence on the part of a
programmer, I would expect failure to check for delivery to
show up in any adequate system test. In the absence of adequate
system test, given the error rates in normal communications networks,
I would expect it to show up in the field too often to be ignored. Unfortunately, successfully delivered data can still be lost further
down the line due to a more subtle error or hardware failure.

One of the realities of the computer business is that to have a
quality product one needs to manage and control errors. The
sort of intolerance you exhibit causes people to conceal problems
and, thus, render them unmanageable. If you were working for me,
I would definitely take action to keep your attitude from damaging my
organization.

Fred.

[Dave Flower]

Well, this thread has generated more heat than light; what I was looking for is a concensus as to how the missing results should be scored, and few posters have even addressed the issue

Dave Flower

[Fred.]

I think there was as much light as heat in my response. I do
regret that it didn't shine on your issue. I wasn't neglecting
it. I just had no idea. You would think, however, that some sort
of precedent would have come up in manual scoring. Maybe the
North's havent' been doing so badly with their pencils.

My understanding from the rules is tha same as yours. Both sides
are non-offending and would receive average plus, with the standard
exception that if they are doing better than 60% they would get
their average on the boards.

But, at some point this all becomes meaningless, and the only
sensible option is to declare no contest, appologize, and refund
the entries. After all, it was the organizers' equiptment which
failed. I have no idea how to make this call and perhaps there
needs to be some guidance somewhere.

Fred.

[David Stevenson]

blackshoe wrote

It's not legal to publish results with two rounds of scores missing
either. There is no legal solution.

I am afraid No Play sounds right to me - and I have probably made more
posts telling people not to use No Play than anyone.

I must say that I have the same worries as other people. We knew
immediately that BridgeMates were not involved because you cannot lose
the data unless the computer goes bang *and* the server goes bang *and*
**every single Bridgemate** goes bang. While it is not at all obvious
who is at fault I suggest you ask for your money back and get
BridgeMates.

--
David Stevenson Bridge RTFLB Cats Railways
Liverpool, England, UK bluejak on BBO Mbl: +44 7778 409 955
<webj...@googlemail.com> EBL TD Tel: +44 151 677 7412
bluejak666 on Skype Bridgepage: http://blakjak.org/brg_menu.htm

[David Stevenson]

Douglas Newlands wrote

They are quite obviously using an inferior product, not BridgeMates.

[Paul Hightower]

"Dave Flower" <DavJF...@btinternet.com> wrote in message
news:20972383.1135.1336148537947.JavaMail.geo-discussion-forums@vbuo17...

Two weeks ago at Banbury (EBU) we tried remote scoring with results
transmitted to a central computer from small units on the tables.
Regretably, something went wrong, and the results for the last two rounds
are missing, although our scorer is trying to recover them (and has some
personal score sheets).

---------------------------------------------------------------------------
Two weeks ago? What happened at the time? In an American club, results are
announced at the end of the game, recaps are printed and players are advised
to "check your scores." Corrections can be entered directly into the PC.
The Bridgemates are little more than a device for saving the director the
effort of retrieving pick-up slips and manually entering data -- they do not
control the final outcome in any way.

[Barry Margolin]

In article <joh9st$846$1...@dont-email.me>,

But if the results don't actually get transmitted, or the software loses
them, as apparently happened in the above case, that time-saver becomes
a big time sink.

Based on the ensuing discussion, it seems like they weren't using
BridgePad or BridgeMate, they must have been using something new that
hasn't had the bugs worked out of it yet.

[Steve Foster]

Barry Margolin wrote:

> Based on the ensuing discussion, it seems like they weren't using
> BridgePad or BridgeMate, they must have been using something new that
> hasn't had the bugs worked out of it yet.

The EBU tried out the devices (as possible replacements for BM1s) at a
minor national event in late 2010.

They identified a number of areas that were a problem for English
bridge playing during the first session (a notable surprise to the
continental vendor was that we run swiss events that get bigger partway
through [as contestants are turfed from a primary event]).

The vendor (unwisely in my opinion) tried to make some changes
overnight before the second session, but landed up making things worse,
and their use had to be abandoned partway through the second session
and the event reverted to paper score recording.

The EBU has since upgraded to BM2s.

[David Stevenson]

Dave Flower wrote

>Well, this thread has generated more heat than light; what I was
>looking for is a concensus as to how the missing results should be
>scored, and few posters have even addressed the issue

It is a long time ago, but either I answered you or I agreed with
someone else's answer so did not add a me2.

[Steve Willner]

On 2012-05-04 12:22 PM, Dave Flower wrote:
> If all the results cannot be recovered, should average plus be
> awarded for the remaining scores ?

I responded privately to Dave, but I'll also post here even though it's
an old thread. If the fault is the organizers', I think "not played" is
the most suitable score to assign. There's precedent in one of the WBF
pair games (Lille?), where an error in assignments caused the movement
to collapse about halfway through a session.

I'm slightly surprised the EBU White Book doesn't cover this situation.
Maybe the next edition will do so. I have a vague memory about
something to the effect of "not more than 2(?) avg+ scores to be
assigned," but maybe that's about an unrelated situation or is just
another memory failure on my part.

--
Help keep our newsgroup healthy; please don't feed the trolls.
Steve Willner Phone 617-495-7123 swil...@nhcc.net
Cambridge, MA 02138 USA

[David Stevenson]

Steve Willner wrote

>On 2012-05-04 12:22 PM, Dave Flower wrote:
>> If all the results cannot be recovered, should average plus be
>> awarded for the remaining scores ?
>
>I responded privately to Dave, but I'll also post here even though it's
>an old thread. If the fault is the organizers', I think "not played"
>is the most suitable score to assign. There's precedent in one of the
>WBF pair games (Lille?), where an error in assignments caused the
>movement to collapse about halfway through a session.
>
>I'm slightly surprised the EBU White Book doesn't cover this situation.
>Maybe the next edition will do so. I have a vague memory about
>something to the effect of "not more than 2(?) avg+ scores to be
>assigned," but maybe that's about an unrelated situation or is just
>another memory failure on my part.

I think you will find that is a personal opinion from the Chairman of
the WBFLC but not an official opinion.