Evening Club Bridge Is Dying


It’s no secret that attendance at club games in the evening has been declining. For four consecutive weeks in May and June the Soledad Thursday night game in San Diego dipped to four tables, hovering at financial break even and at risk of statistically fluctuating under three tables. In 2008 this game averaged 9.4 tables. Old timers mention it averaging 13–16 tables as recently as 2000, just a few years before I arrived in San Diego.

I had heard rumors that a similar decline was happening across the country. But we all live in the bubble of our own locality and memories just ten years distant grow fuzzy. So I asked Bruce Knoll at ACBL headquarters to run a few simple database queries for me. He forwarded my request to Richard Oshlag who sent me the data shown in the table below which I augmented by computing the columns with blue column headers.

Several things are readily apparent. First, annual table count at clubs has been remarkably consistent for 25 years, averaging 2.35 million, despite the steadily aging membership. Second, the percentage of the table count from evening games has fallen monotonically from 43.75% to just 16.47% as illustrated in the plot below. Simple linear extrapolation (red line) predicts the total annihilation of evening club bridge in 15 years. But I predict it will take less than half that time for nearly all evening club games to crash through the danger zone of four or fewer tables.

Percentage of club table count in the evening session from 1991-2015 with projection to 2030

Third, online bridge is doing quite well, gaining steadily, though it is showing signs of leveling off. The online evening percentage has stabilized at 35%. The online numbers are from ACBL sanctioned Bridge Base Online (BBO) events. The events are the hourly Speedballs (IMP Pairs), hourly Individuals, Robot Duplicates, and Instant Tournaments. The typical time commitment is about one hour.

Total club table count from 1991-2014 and ACBL sanctioned BBO table count from 1999-2014

When evening table counts from clubs and BBO are combined, the total evening table count is actually increasing. However, a one hour commitment online is only about one third of time commitment of playing at a club game (though online play is faster). The final table column adjusts for this by adding the evening club table count and one third of the evening online table count. This Evening Adjusted Tables is holding steady at about 750,000 tables. In essence we are seeing players split what might have been a once a week evening visit to a club into three online events, perhaps one per night.

But the stability of the Evening Adjusted table count is less encouraging than it may seem. Uday Ivatury, CEO of BBO, kindly provided me more detailed statistics. A stunning 55% of ACBL sanctioned BBO table count are robot games where one human is paired with a robot and faces robot opponents. In homage to Robert Putnam, it appears we are clicking alone. And though the robot games are nominally one hour events, most players burn through them in a half hour. Between 2013 and 2014 robot games grew in every hourly time slot except the wee hours of 1-5 am MST.

Uday provided hour by hour trends for ACBL sanction games with all human participants. He reports, “a significant decrease in hour 00 [midnight MST] and 01; virtually shut down in the wee hours of 02–05; decreased slightly in 10–12 (lunch) and 16–23 (late afternoon / evening); flattish in 08–10 (morning) and 13–15 (afternoon); and increased significantly in 06–07 (early morning).”

The period over which the ACBL can easily query the table counts corresponds almost exactly to how long I have been an ACBL member. I remember when evening bridge was healthy and afternoon bridge had a dopey reputation. Most of the people who played in the evening were twice my age but they still had jobs. Then my involvement with startup companies had me skimming the Bridge Bulletin more often than playing sanctioned games for about a decade. When I arrived in San Diego in 2004, evening games were still reasonably healthy. Since then evening attendance has nearly halved.

I’m a night owl and evening bridge has always seemed more social to me. I also spend a lot of time in front of a computer and strongly prefer that bridge be an opportunity to get away from computers. Daytime bridge as it is today seems too much like a job, but one where the money flows in the opposite direction. Our local D22 regionals all run on the 10:30 am and 3:15 pm schedule. There is barely time for lunch, which comes two hours too late, let alone any time to socialize. Everyone vanishes instantly after the afternoon session. Evening bridge at the D22 regionals is a ghost town, even at half price.

The ACBL seems to have defaulted into a plan of hoping to replace deceased members with recent retirees. Perhaps this is working well enough because total club table count isn’t dropping despite a mean membership age that is rising 0.4 years / year according to the Quarterly Unit Information Package (QUIP) reports provided to unit presidents. But if the evening game dies, the ACBL will be committed to this default course of action. Most working age people will have no face to face playing option, even if they should somehow be introduced to the game. And even if they do find an evening game, they will always feel like they are on the periphery of bridge. This can’t be good.

Year TotalClub

Tables

EveningClub

Tables

EveningClub

Percent

TotalOnline

Tables

EveningOnline

Tables

EveningOnline

Percent

EveningTotal

Tables

EveningAdjusted

Tables

1991 2,293,822 1,003,755 43.76
1992 2,387,116 1,013,664 42.46
1993 2,425,386 995,283 41.04
1994 2,430,160 962,827 39.62
1995 2,419,214 927,294 38.33
1996 2,419,825 893,148 36.91
1997 2,391,177 859,340 35.94
1998 2,298,183 795,851 34.63
1999 2,204,318 744,616 33.78 3,872 819 21.15 748,488 745,907
2000 2,154,227 699,736 32.48 23,713 5,355 22.58 723,449 707,640
2001 2,299,310 729,359 31.72 80,204 30,972 38.62 809,563 756,093
2002 2,379,048 734,086 30.86 136,591 44,998 32.94 870,677 779,616
2003 2,285,731 696,524 30.47 141,470 36,550 25.84 837,994 743,681
2004 2,321,791 677,815 29.19 156,335 58,010 37.11 834,150 729,926
2005 2,311,188 653,522 28.28 212,950 87,204 40.95 866,472 724,506
2006 2,349,811 632,942 26.94 238,851 95,616 40.03 871,793 712,559
2007 2,336,144 602,898 25.81 335,975 128,321 38.19 938,872 714,889
2008 2,364,155 582,911 24.66 470,854 170,746 36.26 1,053,765 739,863
2009 2,446,271 577,717 23.62 584,679 207,670 35.52 1,162,396 772,610
2010 2,412,263 541,547 22.45 709,333 249,526 35.18 1,250,880 777,991
2011 2,401,360 510,706 21.27 819,333 285,532 34.85 1,330,039 783,817
2012 2,391,837 479,527 20.05 905,890 319,875 35.31 1,385,417 781,490
2013 2,370,713 449,698 18.97 942,040 333,577 35.41 1,391,738 763,711
2014 2,333,013 418,027 17.92 1,007,856 353,437 35.07 1,425,883 753,979
2015 962,242 158,450 16.47 414,454 146,152 35.26 572,904 296,601

This data is also available as a tab delimited text file which can loaded into Microsoft Excel or otherwise easily processed.