The free but unsollicited piece of advice from the Chairman of the Midas Oracle Advisory Board to the BetFair market designers.
Chris F. Masse March 20th, 2008
Jason, I think the allegation of queue jumping made in some of the comments is based upon a misunderstanding. BetFair is just matching the orders when a match becomes available, not inserting new orders at the front of the queue.
.
The new method will deprive outside traders access to these riskless arb opportunities, so they may feel like they are being jumped by BetFair, but really it is just BetFair doing its matching job better.
.
The only issue in my view was the capture of the overround because less sophisticated traders were not aware of the implied odds available, given the existing offers on related outcomes in a multi-outcome event, leading the traders to post over-generous offers to the system. BetFair should use an approach like the ‘phantom offer’ generation that Chris Hibbert explained a while back, to show the effective prices available. That approach would leave BetFair out of the overround business and back to collecting a commission, only, on bets matched.
But his comment didn’t satisfy everyone… The polémique is still going on, unabated, as I type this…
UPDATE: See the comments under this post…
- Exchanges & Markets , Market Designs , Market Trading , Mechanism Designs
- Comments(8)
Should that not be unsolicted!
Three hundred plus Betfair traders revealed, in one way or another, that they were unhappy at the opacity displayed by Betfair in relation to the bet matching system.
Many of these, displayed a degree of sophistication, which, one suspects took even Betfair by surprise. Betfair accordingly climbed down, and accepted that in some regards its critics had a valid argument and that it was in the wrong.
If Betfair says that it is in the business of only collecting a commission on bets matched, then that is what its traders want it to do; no more, no less. That would also extend to Betfair not taking a position in its own markets.
Michael Giberson’s advice is I am sure well intentioned, but he would do better were it to come with an acknowledgement that Betfair’s plans were badly orchestrated and moreover, to all intents and purposes underhand.
The real story here is the power of Betfair traders to get the mighty monopolist to back down. Alas, you have not chosen to cover it in this way; whilst, I seem to remember reading story after story on this site concerning Tradesports and its inability to settle its markets correctly.
"Three hundred plus Betfair traders"
Where do you get this figure and what are you talking about?
-
"Betfair accordingly climbed down, and accepted that in some regards its critics had a valid argument"
True.
-
"and that it was in the wrong."
Hummm… BetFair is mainly in the right, but its new bet matching logic needed refinement.
-
"If Betfair says that it is in the business of only collecting a commission on bets matched, then that is what its traders want it to do; no more, no less. That would also extend to Betfair not taking a position in its own markets."
-
That is the "pure" model of the betting exchanges. But if you have read Mark Davies, you know that he defines a betting exchange as a bookmaker that uses trading technology to minimize risk to epsilon. They don’t have the "pure" model in mind. It baffles many people who follow this issue, but that’s the fact, now.
-
"The real story here is the power of Betfair traders to get the mighty monopolist to back down."
-
The real story here is that the BetFair people are common-sense people, and they got to understand some common-sense arguments made by their traders. BetFair will refine their bet matching logic, and re-put it up when it is perfected, and things will go well at that time.
-
"Alas, you have not chosen to cover it in this way; whilst, I seem to remember reading story after story on this site concerning Tradesports and its inability to settle its markets correctly."
-
Number one, Chris Masse is one post author among 71, here. If one post author wants to have another angle, he/she is free to publish on Midas Oracle and put it exactly as he/she wants. You don’t tell me what I write here, and I don’t tell you what to write. That’s the rule. Everybody is free, here. We are not controlled by BetFair, TradeSports, or the Gambling Commission, or the CFTC, or else.
-
Number two, when TradeSports didn’t settle the North Korea Missile prediction market correctly (or didn’t compensate the "yes" traders, however you want to put it), it lead to the facts that the traders who were in the right (who predicted accurately a real-world event) ended up losing their bet, which was a crazy situation. With this BetFair’s bet matching issue, I don’t think we are in situation where the traders ended up losing money. What we have is a company that put up on the market a not-well-refined-enough product. But could you bring evidence that some traders lose money during that trial period because of BetFair’s unfinished algorithm? If yes, then BetFair should compensate them.
I think my own experience with Betfair, behind the scenes, has been an incredible disappointment, with some astonishing decisions and justifications being reached by them. I don’t regard them as being in any way interested in policing sports or their markets.
-
Match after match where Nikolay Davydenko was playing in minor tournaments last year had suspicious betting patterns. In major tournaments Davydenko was fine and performed well, in minor tournaments he mysteriously kept losing to hugely inferior opponents, with huge money gambled on him to lose on a number of them, well out of line with either how matches were progressing or their relative rankings. The ‘inspired’ money was right every time.
-
Volandri’s matches had some very curious goings on, but the most memorable other match with ‘inspired’ money was an Arguello match. The absolute wall of money laying him was overwhelming, at any price, throughout the match.
-
When Arguello in one of these tournaments was scheduled in the first round to play Davydenko, the forums were rife with predictions of ’suspicious’ betting patterns. The match panned out exactly as predicted, an easy 6-2 double break victory for Davydenko in the first set, before a mysterious capitulation in the 2nd, and retirement in the 3rd. I think Davydenko was 1.10 (a 90% chance to win) when the betting opened, but by the time the match started, he was 2.30. A set up, he was STILL 2.30. Then came the collapse, and the inspired money turned out to be right.
-
Federer can win about $1.2m for winning Wimbledon, but the reported figure won by accounts based from Russia betting on Davydenko to lose that match was in excess of this.
-
Betfair’s view behind the scenes is that it is the responsibility of the ATP to police men’s tennis. However, the new internet exchange model of bookmaker is the first to allow people to bet unchallenged as much as a market can sustain against an individual player. I believe that as Betfair enjoys a rich seam of income from providing these tennis matches, and gives nothing whatsoever to the ATP, and has both an ineffective integrity team (Davydenko’s matches had had a string of irregular betting patterns) and exposure to different results in some of its markets, Betfair’s management are actually doing real damage to the ATP. Its time that a symbiotic relationship was established, where, as in horse racing, some of the money went back into the sport. Its no good just running market after market, even when there is overwhelming evidence that something is going badly wrong on a violently escalating scale with certain players, and pretending that is a valid way to make money.
-
I would say more than 90% of heavy Betfair users have very little trust or respect for the Betfair management team. The notion of them as giant-slaying wunderkind’s against monolithic bookmakers has been corpulent for a long time.
-
I am very disappointed to see Michael Giberson’s most recent post, including
.
."The new method will deprive outside traders access to these riskless arb opportunities, so they may feel like they are being jumped by BetFair, but really it is just BetFair doing its matching job better.
.
.this is so far detached from the reality, and quite sad to see. Any experience using these markets would have led to a different conclusion being reached. If that statement was put up on the BF forum, asking for people’s thoughts, it would be dismissed as very, very flimsy, and well, just inaccurate. (Please feel free someone to post it up on there and see the result).
.
The news that Betfair have scrapped this on in-play markets should indicate that there is indeed something far more sinister in this than is portrayed in Michael Giberson’s last post, which fails to realise quite why there was such a (completely justified) furore. It should never have been endorsed by Michael Giberson in the first place - in my opinion, and I am sorry to say that. If the robot had still been active last night, one mistake by one user entering a bet on the wrong selection (something we have all done a small handful of times) would have seen Betfair carve off a risk free profit of about US $40,000. That isn’t betting fairly, and Betfair has no right to mismatch orders at wildly different prices, and keep the middle ground. Users deserve protection, not the knowledge that one mistake and Betfair will keep enough money from that user to pay for a luxury car with. I would like Michael Giberson to understand why this is wrong before continuing to support the Betfair trap bet robot.
"I think my own experience with Betfair, behind the scenes, has been an incredible disappointment"
-
My own experience with BetFair, behind the scenes, has been quite good. My view of them is that they are law-abiding and ethical people. (They don’t have the prediction market approach, which I criticize them for, but that’s a minor. And their BetFair blog is run unprofessionally, but that’s a minor, too.)
-
"I am very disappointed to see Michael Giberson’s most recent post"
-
I you bring proved facts and scientific arguments, you will convince other people. If you bring conspiracy theories, you will not.
"I you bring proved facts and scientific arguments, you will convince other people. If you bring conspiracy theories, you will not."
In that case, other people should definitely be convinced. There are no conspiracy theories, just plain facts, which in my opinion, also as an economist, are unacceptable in terms of negative impact upon both betting exchange customers, and Betfair’s long term growth. Michael Giberson has a PhD in Economics, so should be able to spot that
-
(1) The new bet matching engine does not find an equilibrium price for users, it introduces an automatic deadweight loss, where the short term swelling of Betfair profit reduces the original economic rent from the transaction to both parties, and arguably, may actually reduce Betfair’s own economic rent as the supplier if the consumers affected are on the elastic part of the demand curve.
-
(2) Raising income to Betfair from the consumers with the least skill is also categorically a foolish move. The charge for a product should be proportional to the inverse elasticity of demand of the user. The ones with the highest e.o.d. are the less-skilled users, whereas the real money to be made is to be extracted from the winners who use skill, and the winners who use more morally hazardous advantages. Michael Giberson would recognise that sound business would be to charge the users who are willing to pay more the most for a product, and I can see why he hasn’t recognised that this bet mismatching will disproportionately impact upon the less-skilled users, but it shouldn’t be that difficult to understand that.
-
(3) Extracting more money from each market also has implications for the long term sustainability of the markets. There is a tragedy of the commons type effect where charge after charge after charge is introduced to the markets. Betfair used to have just commission charges, now there are data charges, and then the introduction of the arbing robot. Can the market sustain that? We don’t know. Will profit from the original source (commission) fall? Almost definitely - the more that is extracted from each market, the higher the rake, the more quickly losers lose, and the less winners win. That is obvious.
-
(4) Is Betfair in a monopolistic position in the marketplace where they can restrict supply and charge artificially high prices should they so choose? Yes. Is this in the interests of consumers - probably not (though by having one dominant market player with only one set of fixed costs to recoup, rather than multiple companies all trying to recoup heavy outlay into building the technology needed, a monopoly could potentially be preferable). I cannot see why a new bet matching logic which results in a sizeable extraction from each market, unannounced, risk free, is worth approving.
-
(5) In running markets run completely differently to the way presented by Dr Giberson
‘"The new method will deprive outside traders access to these riskless arb opportunities, so they may feel like they are being jumped by BetFair, but really it is just BetFair doing its matching job better."
The fact is there probably aren’t any riskless arb opportunities in running. The markets are very, very fast moving, and it is almost impossible to do it, if not actually impossible. It is certainly possible before the off, but the risks on a fast moving event of trying to take two overbroke prices, with a 5 second delay, exceed the benefit from trying to do it. You are very unlikely to be matched on the value side of the price.
-
(6) People could be better matched if every single bet was matched irrespective of completely different prices could be asked for. Sure if someone wants to lay at a short price, and someone else wants to back at a big price, Betfair could just declare them matched, and not pay out a deadweight loss in the middle to whoever the victor was. Betfair would be, as Dr Giberson states "doing its matching job better", but the long term health of the market depends on people getting a run for the money, and trades being struck at somewhere close to an equilibrium price, not building more market distortions into the mechanism to drive the wedge between backers and layers to broader spreads.
-
(7) Dr Giberson should also look at the example of the Hantuchova v Sharapova match. Had that robot been on, Betfair would have taken a risk free profit of $30,000 to $40,000, profiting from the mistake of a user, who submitted a bet on the wrong selection. It is not ethical, in my opinion, to introduce a robot that punishes users much more severely than previously. The user submitted a much larger amount than was matched, and managed to cancel some of it. He/she would not have been able to had the robot been in place, which would have arbed the money instantly against available liquidity on the other side. I agree with what I think Dr Giberson is proposing, offering "ghost" reciprocal odds on the other selection that are the inverse of the odds being backed/laid on the opposite selection (I’m not certain that is what he is proposing, but I think that is what he is saying), and if best execution had been in place with the ghost odds available, then the user would also not have been able to cancel. However, Betfair should not be profiting to the tune of US $30,000 to $40,000 from a genuine user error. They should have announced that if someone was the victim of an error like this - which are reasonably common on the exchange - that they would refund their ’slice’ of rake of $40,000 to the user, who has already lost a huge sum before Betfair took the $40,000. This bet matching logic was designed with the at best flaw (and at worst the cynical decision to take advantage of) of profiting hugely from user errors. It isn’t "betting fairly", and I don’t think it should have been endorsed for this reason either.
-
(8) As for "conspiracy theories", everything I have said can be put in a question to ask Betfair, which they will either confirm is correct or ignore. They won’t say anything I have said isn’t true. There is a procedure in place to check these claims, which are all correct, and very much not ‘conspiracy theories’, and probably should not be described as such.
Ed Murray, thanks for your comment, which I read carefully. We will see what the others are convinced by your arguments…
[...] - The free but unsollicited piece of advice from the Chairman of the Midas Oracle Advisory Board to th…. [...]
[...] I have concluded that their market designers are not competent enough, and that their P.R. managers don’t compute the prediction market [...]