[rockeron]

I understand that there is no intended bias toward any partnership in the private virtual club games. Nevertheless, I have discover4ed a pattern of imbalance in the distribution of high card points that, while challenging, can become downright demoralizing when it is repeated game after game. In smaller games, it is always the side with access to making game that wins. Though great defense can sometimes yield a positive score, that happens infrequently. Has anyone else seen this pattern?

[hrothgar]

Hundreds of people claim to see hundreds of patterns
No one has ever been able to demonstrate anything

Perhaps you will be the first (though I doubt it)

[sfi]

rockeron, on 2020-November-01, 14:20, said:

I understand that there is no intended bias toward any partnership in the private virtual club games. Nevertheless, I have discover4ed a pattern of imbalance in the distribution of high card points that, while challenging, can become downright demoralizing when it is repeated game after game. In smaller games, it is always the side with access to making game that wins. Though great defense can sometimes yield a positive score, that happens infrequently. Has anyone else seen this pattern?

Hrothgar is right about the hands themselves being random, but there is definitely a bias in the scores. The correlation between a positive score and positive IMPs is extremely high.

The reason for this is that each board is scored against (typically) 16 other tables in BBO, where the standard is quite variable. If you bid and make a fairly straightforward game, there is enough randomness that you will likely pick up a few IMPs. Similarly, if you make a partscore you're doing well even if there might be a better place to play.

This does skew perceptions of results. If you are just playing at a casual table you can choose old vugraph deals (the table host selects Deal Source from the menu to do this). That leads to comparisons against two tables who at least in theory can play pretty good bridge.

[barmar]

sfi, on 2020-November-01, 16:46, said:

Hrothgar is right about the hands themselves being random, but there is definitely a bias in the scores. The correlation between a positive score and positive IMPs is extremely high.

The reason for this is that each board is scored against (typically) 16 other tables in BBO, where the standard is quite variable. If you bid and make a fairly straightforward game, there is enough randomness that you will likely pick up a few IMPs. Similarly, if you make a partscore you're doing well even if there might be a better place to play.

This does skew perceptions of results. If you are just playing at a casual table you can choose old vugraph deals (the table host selects Deal Source from the menu to do this). That leads to comparisons against two tables who at least in theory can play pretty good bridge.

The OP was about virtual club games, not casual tables.

In general, variance is higher when there are fewer comparisons, that's just a matter of statistics.

[johnu]

rockeron, on 2020-November-01, 14:20, said:

I understand that there is no intended bias toward any partnership in the private virtual club games.

You would be wrong. I accidentally got access to the list and your name was on it. I would have investigated further but I didn't want to get noticed and put on the list myself. You'll just have to live with it until whoever put you on the list decides to take you off the list.

[sfi]

barmar, on 2020-November-01, 17:46, said:

The OP was about virtual club games, not casual tables.

I did get that, but I don't know which clubs the OP is talking about and the observations from the casual tables may well apply.

[barmar]

I assumed he meant the games you find when you click on "Virtual Clubs" in the Competitive section. These are the online versions of f2f clubs that are organized by national bridge organizations.

[sfi]

barmar, on 2020-November-01, 18:22, said:

I assumed he meant the games you find when you click on "Virtual Clubs" in the Competitive section. These are the online versions of f2f clubs that are organized by national bridge organizations.

Sure, but there are many clubs with varying standards. If the standard of play is better, you will have less variance in results and be more likely to be rewarded for good decisions rather than for simply holding better cards.

[aawk]

If it feels imbalanced it doesn't mean it is.

For the next 1000 boards you play ad all your HCP together and dived it by 1000 and you see the outcome will be 10.

If that a problem start a practice table and deal 1000 boards and do the same.

In any sport we always look for a excuse why we lost and it can't be us.

[mycroft]

There could be an issue with some of the virtual clubs using pre-dealt hands (such as mine that use the Common Game hands). That puts the potential bias away from BBO's dealer and into the one used by the pre-dealer.

Note that I trust the Common Game's dealer even more than BBOs; their whole shtick is to get pros to analyze hands and play as a "serious fun" event, and if their dealer was proven to be non-random, the pros would disappear faster than nightlife when the lights come on, never to return.

But it is a concern with random club owner making up boards with their hand-hacked card shuffling program, or if they use van Staveren's dealer (not BigDeal) which is intended for analysis, not for explicit randomness. Hopefully, nobody does this.

But in general, the answer is "no there isn't". Humans are pattern-matchers, bridge players doubly so. There's *always* a pattern to find in a session of 18 boards. Frankly, there's seriously often a significant bias in the hands in a session of 18 boards. But over 1000 18-board sessions, it does even out (unless you're me. Always Pass As Dealer is a convention on my card (not partner's) (yes, that's a joke, but it's a very well-known one by my partners)).

[Huibertus]

[I have discover4ed a pattern of imbalance)

If that is the case surely you can share the pattern, other can verify it, and if you are right, you'll have found a bug in the random hand generator.

In that case you should be awarded life long free premium membership.

Until this reward is published I don't believe you. I believe that you don't realize that randomness does include long periods of below average strength hand for a few unhappy pairs out of loads of pairs, just as it does in real life.

[nige1]

Agree with sfi. Match-pointed pairs and Board-a-match are less susceptible to chance effects than total-points or imp/vp teams but luck is still a considerable factor. At imp scoring one swingy board can decide a match. At any form of scoring, in a mixed field, if you have the better cards, you will usually score well by making routine part-scores and games. When you meet poor opponents and they get good cards cards, you do well when they bid beyond their playing abilities.

[undoubling]

I suspect that you are confusing the BBO method of dealing cards with that you face in club games. Club games are NOTORIOUS for "flat hands" with "normal" distribution---------ie: very few 4-1 or 5-0 splits. And when those splits DO arise, we tend to blame our "bad" result on that!!! BBO presents truly RANDOM deals (just like at ANY ACBL tournament), with "bad splits" occurring at the "correct" percentage of times. This "true randomness" insults our skewed sense of "rightness"!! For those of us who play (or DID PLAY) in tournaments, we are far more "comfortable" with BBO hands-------although they are STILL not "likeable" for all of the inimical "bad breaks"!!!!

[barmar]

undoubling, on 2020-November-03, 18:02, said:

I suspect that you are confusing the BBO method of dealing cards with that you face in club games.

Are there still clubs that use hand dealing?