Bridge Hands on Bridgebase.

[smerriman]

No, I made no such specific claim. You interpreted my statement that the differences between the challenges I played on that occasion were obvious as amounting to a claim that I could, and would, do it over and over again like a magic trick.

To be fair, given the way you said 'obvious' and 'clear difference', the most logical interpretation was that you believed you would continue to be able to see such a difference, even if you didn't specifically state it. Though, you actually did, as far as I can tell, when you said:

 

As I have said, you will see the difference most obviously in just declare hands

It's not a magic trick, but if there were *any* discernable difference at all, the above test would prove or disprove it - even if you can't predict every single one, you would be able to perform significantly better than statistics would indicate if random.

 

If you can, you've convinced me. If you can't, then you can at least conclude that your 'intensified suspicions' that raised 'the strong possibility' based on what you thought were discernable differences were all cognitive errors.

 

So, did you look at the hands I posted above? Could be a coincidence, but 3 out of the 4 IMP hands were game score contracts, whereas 4/4 of the MP hands were part score contracts. One of the trends I said one might see to tell one set from another in one of my earlier posts. Doesn’t prove anything, but certainly doesn’t disprove my hypothesis.

The main reason I stopped playing MP + IMP daylongs a while ago was because the number of flat boards made it futile in aiming for a top score - I found it too frustrating that they were largely dependent on luck vs skill (the complete opposite of your experience). But you're right; a one-off example doesn't prove or disprove anything; a large set does for the reasons mentioned above.

[smerriman]

Last I knew, smerriman could access my tourneys and could verify that the two sets were the only two run, as I said they were.

I can't - barmar has a script set up that emails me results by participants during BBO events, but not in general.

[smerriman]

By the way, I looked at the hands you posted.

 

You're saying set 1 was from an IMPs challenge, where hands are biased to make them have more IMP variance?

 

Board 1 - both beginners and experts alike would take a flat 10 tricks. No decisions to be made.

Board 2 - given the friendly trump break, both beginners and experts alike would take a flat 11 tricks. No decisions to be made.

Board 4 - given the club lead, both beginners and experts alike would take a flat 9 tricks. No decisions to be made.

 

I guess it's possible an absolute beginner might find a way to go down 3 on board 3, but everyone else should go down 2 and score a fraction of an IMP at most.

 

As for the MP set:

 

Board 1 - no decisions

Board 2 - after the first two tricks I guess the only decision is how to play trumps, but it turns out everything you try works equally, so no variance

Board 4 - again there's only one way to play this at any form of scoring

 

Board 3 appears to be the only one of the 8 boards where you actually have anything to think about.. still looks pretty flat but even if there's a bit of variance, it'll affect both forms of scoring equally.

 

Perhaps it wasn't the greatest example for demonstrating your point :)

 

If you're no longer looking at variance, but think that the chances of a game being makeable is more likely in IMPs than MPs (you referred to the number of games and partscores, rather than variance), then that can be analysed.. but let's stick to one thing at a time.

[pilowsky]

I just played a 16 board tournament.

Every single board had a hand with a singleton or a void - nine had more than one.

Therefore the Earth is flat and Santa Claus is green.

[Douglas43]

I believe Santa Claus was traditionally portrayed in a green costume before a commercial organisation reclothed him in corporate colours, so Pilowsky could be onto something...

[mythdoc]

Beautiful day in southeastern US. Woke up today ready to download and post sets of hands, then read smerriman’s post 78. In an nutshell, he draws a completely contrary set of inferences compared to the inferences I drew from the same hands. Firstly, thank you for finally looking at them. In the time that was spent writing posts insisting I said what I didn’t say, and demanding I submit to tests to satisfy your idea of mathematical and statistical rigor, you could have looked at them and responded 20 times over. But, finally you did. And that told me what I needed to know, which is that it will be pointless posting additional sets of hands and then arguing over inferences from those.

 

I thought mycroft’s post 58 was very interesting, wherein the “corporation test” was cited. I think it’s a test that many, many corporations have failed in this internet age. That has left many of us skeptical of how mainframes may be used to adjust the environment is subtle ways, to substitute a plausible enough (or indeed hyper-plausible) reality in the place of actual reality. Why do they do this? Many reasons, but money tends to be the common factor. I am speaking of the googles, facebooks, and Amazon’s of the world, of course. Their algorithms are their private property. It would be very hard indeed to know anything specific without access to these programs, but the effects on people have been pernicious in many obvious cases. And these corporations often succeed because their participants like to be manipulated in the exact ways the computers manipulate them. Go figure!

 

This is different. I don’t sense any potentially pernicious effects. My hypothesis, if found to be true, would not, as I have said above, be of real consequence from my point of view. I think others would care very much if BBO was using any kind of process to make daylongs more engaging, or to reduce the effect of bad robot play, or whatever. Anything that altered 100% random hands. Which probably explains the ferocity with which the very idea has been met. Which, in turn, leaves us with the irony that I, who harbor suspicions, would not really care if they were found to be true, whereas others, who, as we have seen, dismiss such suspicions with supercilious condescension, would be absolutely irate if they were found to be true. Again, go figure.

 

Lastly, I know I am not the only BBO user to have had these suspicions. I know it because others have posted in other threads and because (other) others have made themselves known to me directly. I apologize to them, because I don’t think I did a good enough job passing along their experiences and their concerns. These concerns will persist after this thread dies down, until either BBO becomes more transparent and definitive as to their processes, or until the processes themselves change. But I’m going to stop writing about it, at least for now. Thanks.

[mycroft]

If that's what you got out of my corporation test, we clearly don't see the world the same way.

 

Google, Facebook, et al weighs *everything* on "how much would we like to get away with this?" vs "how much does it cost us if people find out?" vs "how likely is it that people find out?" - even their anti-snitching policies are weighed on this balance. It's just that they've found, and repeatedly proven, that it doesn't cost them enough when they're caught to make up for doing it anyway.

 

When you're a multi-billion dollar corporation whose job is to sell eyeballs to companies ("you are not the user. you are the product"), and those eyeballs are effectively "everyone", you have that leeway. When you're a 10 employee company and your eyeballs are a very limited, very entitled, very opinionated, and over the last year, increasingly very suspicious self-selected set, the costs are (relatively) higher, the risks are existential, and the rewards are - what? Getting people to play tourneys so they can *reduce* their ad revenue, 20 rupees a shot?

 

And really, I don't think people who raise this have any idea *how* difficult it would be to "make the hands interesting" in a way that wasn't obvious. Unless you want everything hand-selected (and I can think of one job suggested by users to BBO that would be worse than this, but even that wouldn't be much worse), you'd effectively have to write a computer program that can play bridge, with all the common bidding systems, well enough to judge "matchpoint hands" from "IMP hands". If I suggested the one they have is good enough on these forums, I'd be laughed clear into the ocean. So...

 

And again, with all the Ph.Ds in statistics lying around, some of whom have friends who actually said "yes" to "do you want to spend some of your Copious Free Time cracking an encryption system for a game you've never heard of and certainly don't play?" - what's the chance someone in the boardroom said "We know what we're doing, and we're better at it than those people. It'll be fine, we'll never be caught?" (actually, pretty high. People Are Stupid when it comes to numbers past 2^100. Even people who know better, even professional cryptographers and statisticians. But still, even if they did it, with all the anti-cheating statistics and ACBL hand record analysis happening, surely someone would have said "we really need to pull that out RIGHT NOW before we're caught. Because we're going to be caught"?

 

Others are saying this can't be because tests have in fact been done with lots of framing, and everything's passed. I don't remember this one exactly, and I'm not saying this, but the chance is miniscule. I'm saying this won't be because there is literally no one who would benefit from this; it's a massive enterprise to do, and at least as massive an enterprise to do undetectably; and an entire company to blow up if it came out.

 

"The race is not always to the swift, nor the battle to the strong, but that's the way to bet." -- Damon Runyon

[smerriman]

That was precisely the reason I said you would have to be the one to do the tests yourself. If anyone else did, myself included, you're probably not going to believe them.

 

Am I biased towards believing the hands are random? Yes, absolutely, given the immense number of tests I have done over the years, and the fact I don't believe it's possible for BBO to bias them. I won't deny that at all.

 

I still decided to look at your hands in an attempt to be as unbiased as possible, and am honestly perplexed at where variance could come from in the hands I listed as flat.

[hrothgar]

 

I still decided to look at your hands in an attempt to be as unbiased as possible, and

am honestly perplexed at where variance could come from in the hands I listed as flat.

 

It seems obvious that mythdoc lacks the courage of his convictions and is seizing this to avoid needing to validate his claims

[mythdoc]

That was precisely the reason I said you would have to be the one to do the tests yourself. If anyone else did, myself included, you're probably not going to believe them.

 

Am I biased towards believing the hands are random? Yes, absolutely, given the immense number of tests I have done over the years, and the fact I don't believe it's possible for BBO to bias them. I won't deny that at all.

 

I still decided to look at your hands in an attempt to be as unbiased as possible, and am honestly perplexed at where variance could come from in the hands I listed as flat.

 

Thank you for your reply.

[mythdoc]

If that's what you got out of my corporation test, we clearly don't see the world the same way.

 

Google, Facebook, et al weighs *everything* on "how much would we like to get away with this?" vs "how much does it cost us if people find out?" vs "how likely is it that people find out?" - even their anti-snitching policies are weighed on this balance. It's just that they've found, and repeatedly proven, that it doesn't cost them enough when they're caught to make up for doing it anyway.

 

When you're a multi-billion dollar corporation whose job is to sell eyeballs to companies ("you are not the user. you are the product"), and those eyeballs are effectively "everyone", you have that leeway. When you're a 10 employee company and your eyeballs are a very limited, very entitled, very opinionated, and over the last year, increasingly very suspicious self-selected set, the costs are (relatively) higher, the risks are existential, and the rewards are - what? Getting people to play tourneys so they can *reduce* their ad revenue, 20 rupees a shot?

 

And really, I don't think people who raise this have any idea *how* difficult it would be to "make the hands interesting" in a way that wasn't obvious. Unless you want everything hand-selected (and I can think of one job suggested by users to BBO that would be worse than this, but even that wouldn't be much worse), you'd effectively have to write a computer program that can play bridge, with all the common bidding systems, well enough to judge "matchpoint hands" from "IMP hands". If I suggested the one they have is good enough on these forums, I'd be laughed clear into the ocean. So...

 

And again, with all the Ph.Ds in statistics lying around, some of whom have friends who actually said "yes" to "do you want to spend some of your Copious Free Time cracking an encryption system for a game you've never heard of and certainly don't play?" - what's the chance someone in the boardroom said "We know what we're doing, and we're better at it than those people. It'll be fine, we'll never be caught?" (actually, pretty high. People Are Stupid when it comes to numbers past 2^100. Even people who know better, even professional cryptographers and statisticians. But still, even if they did it, with all the anti-cheating statistics and ACBL hand record analysis happening, surely someone would have said "we really need to pull that out RIGHT NOW before we're caught. Because we're going to be caught"?

 

Others are saying this can't be because tests have in fact been done with lots of framing, and everything's passed. I don't remember this one exactly, and I'm not saying this, but the chance is miniscule. I'm saying this won't be because there is literally no one who would benefit from this; it's a massive enterprise to do, and at least as massive an enterprise to do undetectably; and an entire company to blow up if it came out.

 

"The race is not always to the swift, nor the battle to the strong, but that's the way to bet." -- Damon Runyon

 

Thank you, Mycroft, for your reply.