[jdonn]

"Given the scenario you propose, what need is there for any involvement by BBO?

Just set up your own paypal accounts and use them to transfer money back and forth. (I assume that if you trust people not to cheat, you also trust them not to welch on a deal) "

Of course people could do that, but to play for money on BBO would have at least three nice advantages. Avoiding the hefty paypal transaction costs, BBO keeping track of the money swings, and BBO actually performing the transactions.

"Gibs do not cut into your equity. If gib was a human, would you say your human partner cuts into your equity? It probably does but you can't get around having to play with a partner. The fact that it's gib and not a human has no relevance. "

Of course I would say that my human partner cuts into my equity, if I was forced to play with him instead of who I want. You are right, it's not relevent whether it's GIB or a human partner. What is relevent is that it's a partner forced upon us instead of chosen.

"We would probably have more fun if hamman and meckstroth were our partners. Unfortunately, that would not be a very profitable scenario. I'm not sure who you'd like to play money bridge against, but the better the players the worse off you are. The fact that gib plays pretty well and is not playing for money seems like a good situation. If you want to play with 3 other people who play very well and are playing for money, I would say that you are not in it for the profit."

That would be a great scenario, and would not impact profitability at all. If I was playing against you and we both use GIBs, they would essentially cancel and it's you versus me. If I was playing against you and we each had hamman or meckstroth as partners, they would still essentially cancel (not trying to begin a debate about those particular players), and it's you versus me. The profitability would hardly be impacted at all, and it would be a great deal more fun.

Sure, the GIB option should be there too. But why not a real people option? The apt comparison here is to poker, specifically home games. People have home games and invite their friends, they are trying to have fun and make money both. They want to play with who they want to play with, not have some imaginary player thrust into their game. If they bar a weaker player from coming to their game because he isn't as fun, that doesn't mean they aren't in it for the profit, it just means they give weight to other factors as well.

[awm]

While I agree with Josh that some people would enjoy a "four humans to a table" money game, I'm not convinced there's no market for the GIB-partner approach BBO is using.

The fact is, while some of us prefer to play with/against friends (including both Josh Donn and myself), there are a lot of people who play on bridgebase by simply joining a random table. Some of these people are trying to make new friends, but a lot of them are just trying to play a few hands of bridge and don't care so much who they're playing with or against. Adding money to the mix could be even more fun for these folks (and make some money for BBO). In fact if I had the choice between playing for money at a table with three "randoms" I've never met, and playing with a GIB against one "random" and GIB partner, I would prefer the latter. This way my partner's a bit predictable, and there's less chance that I'd be saddled with a bad "random" partner or that my "random" opponents are actually not random to each other (they don't have to be cheating, an established partnership can usually beat up on pickups quite legitimately).

Of course, personally I'm more interested in playing a fun game with friends, and/or practicing with my favorite partners, so I don't see myself playing a lot of money bridge with the GIBs. I expect Josh Donn is in the same situation. That doesn't mean there isn't a market for money bridge though, just that Josh and I aren't part of that market as things are currently set up.

[hrothgar]

jdonn, on Mar 11 2006, 12:12 AM, said:

Of course people could do that, but to play for money on BBO would have at least three nice advantages. Avoiding the hefty paypal transaction costs, BBO keeping track of the money swings, and BBO actually performing the transactions.

I see no reason to assume that BBO is going be able to provide money transfers any cheaper than Paypal. (If BBO is able to do so, they really might want to consider a different business model)

Assuming that you're playing a reasonable number of boards in fixed partnerships, there's not all that much data to track. I suspect all the effort involved in using a paper and pencil would be amply compensated by avoiding the BBO's rake...

Mind you, I'm not suggesting that you actual have money games using BBO's rubber bridge tables. Rather, I am noting that if I were prioritizing development resources, I wouldn't be focusing on a developing new fee based services that have free substitutes.

The two player/two GIB systems permits players to compete even if they don't have a trust relationship. There isn't a substitute available for this. I think that it makes more sense as a business venture.

[uday]

Quote

see no reason to assume that BBO is going be able to provide money transfers any cheaper than Paypal

We can't

[barmar]

jikl, on Mar 6 2006, 11:32 PM, said:

Saying the skill vs luck thing is so misguided. There are people that garner a living from rubber bridge. One very good Australian player kept statistics over a long period of time (decades), he would average nearly 2 points a rubber. Therefore for every rubber he would gain 2*stakes/100. On average then, this means about $60 an hour ($10 per hundred). Back to the luck thing, some days he would lose a lot, but the key is the average. On an average (8 hour) day this is $480 per day.

Sean

Sure, if you play bridge for hours, the skill factor will overtake the luck factor. But when the opponents make a vulnerable grand slam against you it will take you quite a while to make it back with just skill. At 2 points a rubber, that's 1,000 rubbers (if a rubber takes 15 minutes, and you play 8 hours every day, that's about a month).

In poker, it's always possible to avoid losing lots of money: fold. In rubber bridge, you can lose 2210 just by being dealt the wrong cards, and there's absolutely nothing you can do about it.

[barmar]

Yes, you'll lose, but you'll lose slowly. And when you get a good hand, you can bet big and make much of it back. I admit I'm not a poker player, so I realize this may not be a long-term winning strategy, it's just a way to avoid losing big.

When I used to play rubber bridge frequently, we played 4-deal Chicago, and a standard session was 3 rubbers (to get all the possible partnerships). So if someone makes a grand slam, the rest of the session is just (on average) going to get the best player back to even.

I still stand by the statement that you have to play very frequently to overcome the luck factor. Mathematicians call it the Law of Large Numbers -- it's not that hard to flip 5 heads in a row, but if you flip a million times it should be pretty close to 50%.

Duplicate bridge has similar problems -- almost anyone can win a 7-board Swiss Team match, since one big swing board can win the match. But a 64-board KO match is usually a good judge of the better team.

[andych]

Login the QA server several times but there is always only 1 person (myself) at 8-9pm EST.

Just now I see the 2nd player and tried a few hands...... Not good.

Maybe someone has mentioned before,

1. The GIB play speed.
I have the default value-medium. The initial few tricks GIB is slow but acceptable. The later tricks GIB (both of them, when I defend) is just too quick. I even couldnt see the cards. It is not human at all and offer the player a bad feeling.

2. GIB sucks
Maybe true that GIB could be reasonable/good in most hands. But it sucks at the most critical hand!!!

I am holding ♠AJxxx ♥AKxxx ♦Kxx

Bidding goes 1♠ by me, 2♣ by LHO GIB, 3♣ by GIB partner (limit raise or better), 6♠ by me......


7NT by GIB partner.....


She is holding ♠KQxxx ♥QJx ♦Axx ♣Kx

Maybe I could suit GIB style (not sure what style of her, I dont own a GIB myself) and bid 4NT and signoff in 6♠ (take the money at rubber) GIB would not bid 7NT, but it is taking away the fun.

True it could happen to my opps. But the money involved (+50/100 points) would not be enough to offset the nonsense.


Much work needed.

[Sigi_BC84]

andych, on Mar 11 2006, 03:43 AM, said:

I am holding ♠AJxxx ♥AKxxx ♦Kxx
Bidding goes 1♠ by me, 2♣ by LHO GIB, 3♣ by GIB partner (limit raise or better), 6♠ by me......
7NT by GIB partner.....
She is holding ♠KQxxx ♥QJx ♦Axx ♣Kx

LOL. Also be prepared at all times for GIB bypassing 3NT for no reason (happened to me in an offline game, uncontested auction!).

I think it's playing the cards really well, but bidding is no fun.

--Sigi

[andych]

My dream BBO 'LA Bridge CLub'

Quote (Emperor of the 2-cent game from Bob Hamman at the Table)

Quote

I was lucky that I could cut my teeth, so to speak, against some of the best.  It was an oppurtunity that is no longer available to aspiring players.  Today, only the largest cities have money bridge clubs.  For most players, unfriendly rubber bridge games for high stakes ar simply unavailable.  Believe me, there's no better way to improve your game.  If you want to learn how to play bridge, become wealthy enough to afford the losses and play with better players for stakes that can make losing an unpleasant experience.  The other schools of bridge are not nearly as good.

It is fxxking good idea to build a rubber bridge club in BBO. But much work is still needed .....

Hope no one accused me of copyright ....

[andych]

Jlall, on Mar 4 2006, 09:40 PM, said:

Something weird happened. My RHO (gib) opened 2C. I had Ax --- xx KQT9xxxxx or something, white/red. I tried 6C, mainly to see what would happen. LHO bid 6H, and my gib partner bid 7N. RHO, the 2C opener, didn't X 7N. His pard did, then he wouldn't lead and the table closed because he was too slow.

Anything wrong here?

For the sick 7NT hand I described earlier, my human opp also couldnt lead because BBO says he is too slow and the table closed. I chat with my opp after and he says ♣A is on the table. He is on lead and the ♣A is dealt with his GIB partner. I could not see any card played?

Is my GIB partner so smart to stop my human opp to lead, when she knows she commits a disaster?

[Deanrover]

I'm looking to test this out right now. While it is nice to be the only person logged into BBO again, I can't play on my own!

[Deanrover]

I am really shocked at you who suggest that lowering the rake will attract a larger client base. The demand for online $ bridge will be extremely price inelastic. The serious players will care about it, but BBO will be operating a monopoly so their choices will be limited, the casual player won't know/care about it, after all what is 10 cents when you just won a $10 pot?

I don't understand where some of this talks about 'fairness' come from. BBO are offering a service that isn't available anywhere else, they don't force you to use it. If they want to charge a fee of 200% of the pot then that is perfectly fair, if you don't like it then don't play.

I suspect BBO will charging a small % fee of the pot, perhaps capped at x$ per hand, where x <=5. They will want to charge a fee which allows the good players to maintain a healthy enough edge to keep coming back, and which doesn't depleat the fish's resources too rapidly - this will likely be the profit maximising strategy.

As for whether to have a higher rake or to charge transaction fees - from a purely economic standpoint the latter is almost certainly preferable, but economics tends not to take into consideration psychological factors which would suggest pursuing the former strategy. If BBO feel like being nice perhaps they will offer 2 options: i) Just a rake, ii) The client pays transaction fees but receives a % reduction in rake. The casual player will almost certainly prefer the first, but the regulars will go for the latter.

[Deanrover]

To answer which has more luck - poker or bridge, is an extremely complex question, but I'll give it a shot.

Note: I am comparing total points bridge with 6 max limit hold em. Using other types of bridge scoring and other types of poker WILL produce different results.

My hypothesis is thus:

If a poker player makes $x/ hour, and a bridge player of the same ability plays at a limit which also enables him to earn $x/hour, then whichever player's earn rate has the higher standard deviation is playing in the more 'lucky' game. Restated, If a winning poker player has a standard deviation of $x/hour, and a bridge player of the same ability plays at a limit which also provides him with a standard deviation of $x/hour, then whichever player has the lower win rate is playing in the more 'lucky' game.

A fast 6 max limit holdem table will get 100 hands in per table. A very good player will earn 2 bb/100. His standard deviation will be 18 bb/100. Both these figures are approximate but come from databases on 100s of 1000s of hands. So a poker player who earns $100 per hour per table will have a s.d. of $900 per hour per table.

Uday mentioned earlier that the average standard deviation per hand was 300. I will assume that the s.d. is the same for every player. So a player who wins on average 50 points per board will have a s.d. of 300, as will one who loses. This likely approximates reality. Assuming 9 hands per hour (quite possible when playing with GIBs), this gives an hourly standard deviation of 900.

So: a bridge player playing for $1 a point will have roughly the same hourly variance as a poker player playing $25/50 6 max. I've already said that a very good (but not world class) poker player will make $100/hr/table at these limits, so:

if a very good (but not world class) bridge player can average more than 11.11 total points per board edge, over the long run, then total points bridge is more skillful than 6 max limit hold 'em.

I have no idea what a good player's edge is like in bridge, but I would expect it to be considerably more than 11 points per board.


*Disclaimer* I haven't done any math of this type for 21 months so it might be horribly wrong, in which case my whole post is nonsense, sorry.

[Deanrover]

I notice that GIB stars playing his cards VERY quickly (i.e. instantly) towards the end of the hand. Is it possible that timing inferences can be drawn? If so, you should consider implementing a (randomised) minimum time that GIB takes for its plays. A couple of seconds would be sufficient.

[fred]

Deanrover, on Mar 11 2006, 03:22 PM, said:

I notice that GIB stars playing his cards VERY quickly (i.e. instantly) towards the end of the hand. Is it possible that timing inferences can be drawn? If so, you should consider implementing a (randomised) minimum time that GIB takes for its plays. A couple of seconds would be sufficient.

In the next version (coming soon) there will be a slight delay imposed each time GIB plays a card and a slightly longer delay imposed at the end of each trick.

This has not been done for the reason you suggest (because people might be able to drawn inferences from GIB's speed). It has been done because when GIB plays super-fast it is hard to follow the play.

Fred Gitelman
Bridge Base Inc.
www.bridgebase.com

[Deanrover]

Is there any time limit ? I don't see any displayed prominently.

You mention that for disconnects you will use some automated system to find out the result. Are you going to implement this to remove any incentive to disconnect? I can envisage a scenario where a declarer has only a few tricks left but has lost track of which previous cards have been played - he knows the contract is 100% cold but can't remember who discarded what, so he disconnects and lets Gib play knowing full well the bot wont have made such an error.

[uday]

Hands that are incomplete are queued for (not implemented yet) settlement.
I'll implement this before we go into real money. Probably settlement will involve 4 gibs finishing off the hand. If there are very few settlements maybe I'll have a human eyeball the board instead. Haven't decided yet...

[Deanrover]

Uday,

I guarantee you will have to deal with angry emails on the issue of disconnect protect.

Edit: You actually wording your post very subtlely (and well). If your proposed solution involves having 4 GIBs finish the hand BUT the disconnector gets penalised in addition, that might be the least imperfect solution. You should still keep open the option of adjusting the score if you judge a player to have potentially unfairly gained from disconnecting.