[andych]

Wonder why the star players not playing with their regular partnership to shoot for the money prize??
Does most of the money goes to the autioneers instead of the players?
At the same time some regular partnership could be seen also?
How much do Levin-Weinstein make this year?

[mgoetze]

Because they have to buy at least 10% of themselves, so a regular top partnership effectively gets worse odds than a non-regular partnership. They might also dislike the system restrictions.

The auctioneers don't get any money.

As for how much Levin-Weinstein made, it depends how much of themselves (and potentially other pairs) they owned.

Much of this information can be found on the Cavendish website.

[kenberg]

Larry Cohen was commenting on this while I was watching, confirming what mgoetze says. I found the argument odd, but I guess they know what they are doing.

The argument:
If the absolutely top class pair xy play together, the auction price will be huge. After buying a portion of themselves, they must finish high (LC said seventh, but no doubt it varies) in order to break even financially. So they play with a "stranger", this lowers the price and their risk.

Maybe an economist could comment. It seems odd. Presumably their price goes down because their chance of winning goes down. In a rational world these things should balance. One could imagine, for example, Zia-Rosenberg playing together but agreeing to each have three shots of scotch before each session. Presumably this would lower their price, and lower their chances. Would they see this as a good thing? I'm having trouble seeing why playing with a stranger is a good financial move, any more than agreeing to play while drunk would be.

[hrothgar]

kenberg, on May 16 2007, 03:45 PM, said:

Larry Cohen was commenting on this while I was watching, confirming what mgoetze says. I found the argument odd, but I guess they know what they are doing.

The argument:
If the absolutely top class pair xy play together, the auction price will be huge. After buying a portion of themselves, they must finish high (LC said seventh, but no doubt it varies) in order to break even financially. So they play with a "stranger", this lowers the price and their risk.

Maybe an economist could comment.

Coincidentally, when i was working on my PhD in Economics, I did a lot of work on Game Theory with a special interest in auctions...

I think that I can offer a (reasonable) explanation for this phenomena. Here's how an economist would treat this type of problem:

Lets assume that we have N people bidding on a hypothetical pair. (Each bidder has infinite resources)

Each person participating in the auction places a value on that pair. Hypothetically, I might believe that the expected value of “Zia – Hamman” is $12,000. Justin might believe that the pair is worth “$11, 500”. Matt believes that the pair is worth “11, 487”, yada, yada, yada. (For the purpose of simplicity, you normally assume that these valuations are generated randomly using some easy to manipulate probability density function (A uniform distribution or whatever)

I believe that that the Cavendish pairs auction is equivalent to a standard voice auction. These auctions have a very simple equilibrium. Whoever places the highest value on the pair ends up purchasing them. The price that they pay is epsilon more than the valuation of the second highest bidder. If we look back to the original example, I would Justin would stop bidding on Zia - Hamman when the purchase price his $11,500. I'd end up buying the pair for $11,500.01

This will (often) lead to a condition known which is known as the “Winner's Curse”. The individual who purchases a given pair is the person who is most likely to have over-estimated the value of that pair. (As always, there is a decent treatment of the winner's curse on the Wikipedia). Please note: Some economists would argue that bidders are smart enough to build a compensation for the winner's curse into their valuations and will adjust their bidding accordingly. Even so, folks can still end up overbidding at times.

The rules of the Cavendish create a perverse incentive for the players:

Players are encouraged (forced) to purchase a percentage of the winning bid. (I assume that this is designed to ensure that they have an incentive to do well). However, the nature of the Winner's curse ensures that they are required to buy a percentage of an over-valued asset. They are (essentially) forced to make a bad investment whose value was based on some liquored up high roller who might be showing off. Furthermore, the amount that they are required to invest is a function of the amount that they sell for. The greater the amount that the sell for, the large the amount that the players need to invest in this over values asset.

Accordingly, players have an incentive to depress the amount of money that they sell for. I can see how this might create an incentive to break-up established partnerships.

One could make the argument that players should be allowed to purchase their “share” at a discount. Hypothetically, Zia and Hamman should be required to purchase a 10% share at 8% of the sales price. (The actually discount would be designed to balance out the expected value of the winner's curse). Alternatively, one could eliminate this entire requirement and increase the size of the session rewards. Either of these adjustments should eliminate the incentives for pairs to deliberately handicap their performance prior to the start of the tournament.

(Cohen made an interesting observation. I had never thought about this before now)

[ArcLight]

Do any of the participants in the Cavendish pay the other to play with them?
Does a "lesser" expert in the Cavendish pay to play with Meckstroth or some other top player?

[mgoetze]

ArcLight, on May 16 2007, 02:37 PM, said:

Do any of the participants in the Cavendish pay the other to play with them?

If I interpreted the vugraph comments correctly, Vainkonis was probably paying Pszcola to play. I don't think Meckstroth's partner was paying him, but I wouldn't know... I'm not an expert on who's expert enough to not have to pay to play with an expert expert, you see.

Of course, there's a difference between paying the other player to play, and covering the fees (a portion of the player's pool was deducted for expenses). I suspect there were quite a few different individual agreements. I wouldn't be surprised, for instance, if Bramley had a higher stake in Bramley-Lall than Justin did (either because they agreed to split their stake disproportionately, or because Justin sold his part to Bob Hamann), but unless Justin chooses to tell us, we'll never know.

[mgoetze]

hrothgar, on May 16 2007, 02:15 PM, said:

However, the nature of the Winner's curse ensures that they are required to buy a percentage of an over-valued asset.

I don't get it. All the pairs are auctioned off, and they can't all be overvalued, can they?

I think it's more a matter of hype. For instance, German bookies will give a disadvantageous quote for a German win in an international soccer match, yet plenty of Germans will bet on Germany anyway.

The difference between the amount Meckstroth-Rodwell would sell and the amount Meckstroth-Johnson would sell for is not proportional to the difference in the expected winnings of Meckstroth-Rodwell and the expected winnings of Meckstroth-Johnson. This is due to a psychological factor which causes bidders to overestimate the expected winnings of Meckstroth-Rodwell and underestimate the expected winnings of Meckstroth-Johnson.

So, if you were Meckstroth, and could play with Rodwell, investing, say, $4,000 (half of 10% of $80,000) for expected winnings of, say, $2,500, or with Johnson, investing $1,200 for expected winnings of, say, $1,500, what would you choose?

(This is oversimplified because it doesn't take the Player's Pool into account.)

[hrothgar]

mgoetze, on May 16 2007, 06:01 PM, said:

hrothgar, on May 16 2007, 02:15 PM, said:

However, the nature of the Winner's curse ensures that they are required to buy a percentage of an over-valued asset.

I don't get it. All the pairs are auctioned off, and they can't all be overvalued, can they?

Comment 1: You can definitely have auctions in which all of the assets sell for more than they are worth. I would argue that the frequency spectrum auctions in the wireless space are an excellent example of this phenomena.

Comment 2: The case of the Cavendish is special because the price paid for the pairs feeds back into the prize pool. The more money that someone pays for any given pair, the greater the value of all the pairs out there. I haven't bothered running any numbers (and I'm not particularly inclined to do so). However, the prize pool is non-zero sum. The event organizers are taking a rake. In addition, a portion of the funds need to be allocated for operating expenses as well as the prize pool for round winners and the like. I'm pretty sure that this would permit a situation in which every single pair was overvalued by some small amount.

[jdonn]

ArcLight, on May 16 2007, 08:37 AM, said:

Do any of the participants in the Cavendish pay the other to play with them?

Many.

[kenberg]

Maybe it is starting to come through to me.

Suppose I am rich and suppose I want to bet on the Cavendish. Suppose I like picking winners. Suppose several of my clones feel the same way, and suppose Meckwell are in the game. My clones and I drive the price far above the rational. We aren't trying to make money, we have plenty of that. We all just want to buy the winner. Ego, or whatever. We are fully aware that better odds (lower probability of winning but much lower price) are available elsewhere, but we don't care.

This excess of cash is not a problem I am facing, but I guess I can imagine it. Sort of a "Winner's Curse" on steroids.

And here are poor Meckwell. I don't imagine those guys are on welfare, but probably what happens to five or six thousand dollars of their money matters, at least a little, to them. If they play together they will be forced to wager that sum. They have enough respect for their opponents to know that they would be placing a bet that, at the odds they will get, they would rather not place. So they cope.

I guess I can follow the reasoning. Still, should I be transported by Mr. Jordan to be a Cavendish contender, I would go for it with my best partner. It's only money.

[Fluffy]

mgoetze, on May 16 2007, 11:18 AM, said:

Because they have to buy at least 10% of themselves, so a regular top partnership effectively gets worse odds than a non-regular partnership. They might also dislike the system restrictions.

I didn't knew that, what if Bill gates decides to bet 100 million $ for each pair?, he would force everyone to pay higher entry, I guess organizators would be happy .

[han]

kenberg, on May 16 2007, 02:51 PM, said:

I guess I can follow the reasoning. Still, should I be transported by Mr. Jordan to be a Cavendish contender, I would go for it with my best partner. It's only money.

Maybe most important to them is to play in the world championships with their best partner, and to make some guaranteed money in the Cavendish.

[kenberg]

Hannie, on May 16 2007, 05:06 PM, said:

kenberg, on May 16 2007, 02:51 PM, said:

I guess I can follow the reasoning. Still, should I be transported by Mr. Jordan to be a Cavendish contender, I would go for it with my best partner. It's only money.

Maybe most important to them is to play in the world championships with their best partner, and to make some guaranteed money in the Cavendish.

Yes. With thought, it all makes sense.

Having World Championships to determine world champs and the Cavendish for glamor and cash is all to the good. And some darn good bridge all around.

[mycroft]

It is a well known "rumour" that Meckstroth and Rodwell don't gamble with bridge. If they play, it's because someone is paying them. They know what they're getting and what they're putting out. Whether or not that's good for bridge, the US, whatever, it's their decision, and we get to complain about it when we're as good as they are and have their choices.

It is another well known "rumour" that there are players in the Cavendish Invitational that wouldn't be there if they didn't gamble (and push up the auction pool). Not that they're *bad*, but they are one step below the level they would normally invite. Essentially, their entry fee is much higher (and they can afford it), because they won't be back next year if they haven't either bought or really tried hard to buy several pairs. If you're worth $30m, an entry fee of $100K-$150K (with a decent chance of getting a fair bit of that back) is no higher barrier to entry than $2500 per pair with a potential $12K5 buyin if nobody thinks you can win to "average bridge world-class player". Even then, these guys do want to win, so if they're sinking that much into the week, they might as well hire a good pro to have a shot at it themselves.

Note that this is good for the Cavendish (in moderation, at least), because if they weren't there, the Calcutta pool would be around 15% of what it actually is, bridge players being the notorious spendthrifts that they are. Also note that there are much better money makers for Bridge Pros than the Cavendish - I expect that many of the double-gamers could make more playing online, or even Las Vegas, poker than they expect to make from the Cavendish; and even Real Bridge Pro fees for a week probably beats fourth place. People play there primarily because it's famous and a good showing there is good for their Bridge Pro resumes; even a bad showing is good for your game, if you're new on the Pro circuit, and just knowing that you were invited (as opposed to being a buy-in) to this strong a field is good for the New Pro resumes.

Michael.

[DenisO]

Interesting discussion and generally agrees with what Larry Cohen was saying on the commentary. But for me it doesn't explain the other side i.e. why some good established pairs keep playing together in the event - in particular Levin/Weinstein who are always near the top of the auction ; look at their auction places in their winning years - 2007 (47K) top, 2002 (33K) 4th, 1999 (36.5K) 4th.

Interestingly in 1999 Lauria/Versace sold for $56K - was that a record? Yet 8 years later the top pair go for $47K - is there less money about now?

Denis

[Rain]

We may want to measure differences in prices too. I believe (unsupported, just observations) that there is less disparity in skill level between the top pairs. Information moves more freely these days, remember?

Hence top pairs are not that much better than others, enough to warrant top prices because of their certainty to win.

[JanM]

Rain, on May 19 2007, 02:07 AM, said:

We may want to measure differences in prices too. I believe (unsupported, just observations) that there is less disparity in skill level between the top pairs. Information moves more freely these days, remember?

Hence top pairs are not that much better than others, enough to warrant top prices because of their certainty to win.

It isn't that the top pairs are less good now, or that there's less disparity. It's that given the format of the event, playing better than anyone else (or than most anyone else) is far from enough to win, or even to assure being in the money. I remember when Eric Rodwell announced that he would never again play in the Cavendish with an expert partner - he and Chip had just finished out of the money after playing about the best bridge either of them could play (and that's pretty good). They then figured out that if they had played perfectly they'd have finished 6th (and their "perfectly" is a lot better than mine). That's the nature of the event - you're at the mercy of the opponents, the field, random luck. So a lot of the top players prefer not to put themselves in the position of risking their money on a "bad bet."
Possibly, the bidders in the auction have also figured out that expertise alone isn't enough to win and that's why the top price isn't as high as it used to be even though there's more total money in the pool - I don't know. I do know that I picked Levin & Weinstein to win in the Cavendish's cute "sweepstakes" but it never crossed my mind either to bid on them or to try to buy a piece of them after the auction.