Implied Tradesports Dem House Gains

Crossing Wall Street ^ | 11/3/2006 | Ed Elfenbein

[oblomov]

This is a favorite topic of mine. In investing, we can look at two markets and imply a third. That's basically how options work. Well...we can do the same for predictions markets.

At Tradesports, they offer futures contracts for how many seats they Democrats will pickup in the house. They offer contract for several different scenarios (i.e., greater than 14.5 seats, or greater than 19.5 seats). Assuming a logonormal distribution, we can find an implied mean and standard devation.

The chart above shows the mean number seats the Democrats looks to gain (black line) with plus and minus one standard deviation bands (red lines).

[Torie]

If your model is right, I am a genius, since my last public prediction was minus 21 for the House. Maybe I will just leave that one alone. :)

[WOSG]

"Perhaps all these people "investing" are people who are simply following the liberal polls. Much the same way the Clinton regeime touted the "new economy" they are embracing a future as promoted by liberal polling."

Here's a test ... DID TRADESPOTS GET THE 2002 RACES RIGHT?

In 2002, there was some GOP 'upside' surprises, as we did better than expected by the pundits/media. Wonder if the tradesports followed the polls, or were closer to actual results.

FWIW, I think we will do better than that implied loss of 21. 12-18 loss is my range, IMHO, we hang on close or we lose but not by much.

I do believe there are some polls not right.

[WOSG]

"They picked exactly what the polls told them to pick. If you want a classic example of that, take a look at tradesports charts on Bush and Kerry immediately before and after the phoney exit poll data was released on election day 2004. Within minutes Kerry soared to a 75% or so odds on favorite. If the betters really had the inside scoop, nothing would have changed as a result of the bogus polls."

That' a very good example to tell you that tradesports is not perfect. In fact once I realized the exit polls were off, I was wondering about betting on it. Three *are* ways to beat the market.

[WOSG]

"On election day 2004 bush was at about 20% at 8:00 PM EST on trade sports. Bad info in bad bets out.
"you are selecting a very narrow time frame impacted by unexpected data (exit polls) to make your point. Markets are not 'efficient' in the sense you expect. They always over-react, on the upside and downside, to news. In this case, you have a contract essentially hours from expiration and there is news in the media that a certain outcome will occur. Of course people will trade on that news."

You miss the point he is making. People are wondering if the tradesports has some inside track that polling doesnt, yet this example showed that tradesports 'investors' were trading based on (incorrect) news.

Tradesports therefore is merely a regurgitation of expected election results based on public information on races.

"The area of interest in these contracts is not in the last .1% of the contract life, but in the days and weeks prior to election."
I don't see why it matters. A campaign looks to be winning, it's above 50. It gets bad news, it goes down.

I'm concluding that tradesports is just a shorthand for getting the same info you could get at realclearpolitics.com

[Sam Spade]

Well, you know my prediction last week was -21 as well and I've been debating whether to throw it all the way up to 30 (I don't see the polling yet to push it up to antiguv territory) or maybe cut it off at 25 or so. I'll decide Monday. Maybe then I'll be in antiguv territory.

PS: I think Bill Kristol's prediction of 243 Democratic seats in the House is the closest I've seen to topping antiguv's yet. :)

[I'll be your Huckleberry]

The problem with the multi-deck shoe is the length of the curve.

In a single deck game the consistency of the deck can change very rapidly.

It is, as I am sure you understand, those times where the deck is unbalanced (rich vs thin) where the odds actually favor the better. In these situations, depending on the degree of shift, a prudent better with good money management skills can take his profit.

The casinos now deal very deep into the deck. Card counting is not an issue in most pits. Almost everyone now understands the basic relationship of face cards to small cards and for bets to rise after a run of small cards is not that unusual. However, I know of only a handful of true counters in my small town.

When the count is +5 or so, you must content yourself with a moderate win.

If you begin to hurt the casino - you are toast.

It happens this way:

You have been visiting this certain casino on your route every weekday for an extended length of time. You are disciplined and when you have had a good run, you pocket the chips (to prevent an accurate assessment, without cameras, of your win), tip the dealer, give your regards to the pit boss and leave.

But one day you sit down, move your bet into the small circle and the pit boss comes over and pushes it out of the circle and back to you. He says "Huck you are just to good for us, you need to play some other game".

The old days of being photographed and 86'd are gone, just too much competition to lose all the heavy betters.

But if you hurt them, you will be barred from the blackjack table.

So don't hurt them.

Huck

[AntiGuv]

Hmm.. I might have to up my projection since the whole point of a wave election is that it exceeds all expectations. Then again, that'd be a vicious cycle, huh? :)

[supercat]

So don't hurt them.

I would expect table manners, including how one appears to other customers, would affect things. If watching you win money convinces other people that they can win money too, and thus causes the casino to make more money off them than it otherwise would, it may be more willing to let you make some money than if your style of play causes other people to back off when the deck is unfavorable.

[I'll be your Huckleberry]

In all honesty, the deciding factor is normally how much you increase your bet in favorable situations.

If you are old school and try the standard minimum bet on negative decks and maximum deck on very positive hands then your manners are not going to matter.

On a large increase in bet size the dealer is required to notify the pit boss who will signal shuffle up.

This will both aggravate the staff and mark you as a counter.

You are much better served to disguise your play as let it rides.

Take your winnings and double your bet if the count continues positive. If you can do this successfully you might acquire the reputation as a streak better.

Try never to jump from the minimum table bet to a large bet. This will get you banned very quick.

Become a regular.

Cry like hell when you lose.

Never give them a clear understanding of how many chips you have.

Never color out from the table.

Take green chips from the table and put them in you pockets.

Understand that it is possible for the eye in the sky to calculate every dime you won. But if you are moderate in your winnings, it is almost never done.

Always tip when you leave.

Never drink alcohol.

Always have a drink in front of you, order often from the cocktail waitress. (Club soda w/ a lime, straight OJ etc.)

Play when the dealers are at their low ebb of energy. Early in the morning after shift change, and in the afternoon before shift change.

Remember the casino has two weapons:

Alcohol & Greed

And you only have one:

Discipline

Huck