Get yer polls, bets, numbers & pretty graphs! Elections 2008

Here's the graph to go with that:

Wow, good to see. I like that one, thanks for updating.

Meanwhile... I HEART POLLSTER!!!

Check this out!!

http://www.pollster.com/blogs/new_charts_for_polls_polls_pol.php

Knowing someone is out there doing that, you could just take their money. Buy Obama's low and dump them high. It would be hillarious to take this guy's money while he tries to manipulate the market.

Talking of graphs, look at the developments over the past couple of days on specific sub-questions in the Diageo/Hotline poll:







Interesting that the energy question shifted massively to Obama along with the economy question. Energy's not on the agenda in this week's financial crisis that I know of, but I guess voters just associate the topic with economic issues after all.

In comparison, the indicators on national security and health care didnt budge at all - so you cant say that it must just have been a disproportionally pro-Obama sample today:

I hadn't seen that but can tell you the author hasn't much of a clue what he's talking about. This is especially clear in this paragraph:

This is bullshit. The author fails to mention that every time the score doesn't fall within that narrow gap (usually); you pay the Mirage or Bellagio $1,000 (vig) in this strategy. While it's unlikely to see that big of a gap between those two books; a 4 point difference could easily be found at local books in New York and their opponent’s home market.

This type of spread actually comes up pretty often between the opening line and the final (game-time) line (especially in college ball). Someone who wagered at the opening line has every opportunity to hedge his bet in this same fashion before game-time; but few do. Not because there is anything wrong with it… it just isn’t usually considered a favorable wager. The house vig goes a lot further than this author realizes at keeping people honest.

Further; much like candidates at Intrade, football betting will ALWAYS reflect the emotional biases of the gamblers as well as their cold hard football number crunching. Fan-favorites (like the Packers) probably show a point’s difference in the spread just because of the fan-factor.

Ultimately; the casino couldn’t care less whether the line is up or down; but they prefer stability. Their experts only exist to predict the line so they don’t get lopsided betting at the opening line and get stuck by the margin mentioned above. If they predict it ideally; gamblers will put equal amounts of money on both sides of the bet, thereby eliminating the Casino’s exposure to loss possibility. Casino’s, Bookies, etc make their money by charging a vig (typically 10% fee to losing wagers… about 5% of the total wagered). The better the expert’s opening pick; the less gambling the Casino does (by moving the line and creating such gaps). Casino’s don’t like to gamble. The line’s very movement from open to final is predicated on ensuring as equal as possible amounts of wagers on either side of the bet. The Casino also has to factor in the possibility of losing wagers that hit in the margin of the move… which is very expensive, obviously. This is why the opening line makers have to be so damned good at what they do… and why average Joe usually loses when trying to best them.

Intrade, unlike the Casino couldn’t care less when the line moves. Their fees will be collected regardless and it is NEVER their money you are wagering against. They simply provide a platform for people to bet against each other. If some wealthy moron wants to artificially make the predictor map look more to his liking; the only way he can do it is to continuously add more and more money to the candidate he’s trying to bolster. This can’t be done with moves; because any sale order would have the opposite effect. Any damn fool could sit at their Ameritrade Interface and do the same thing with pretty much any relatively obscure stock (for as long as they could afford to keep shoveling in money).

But unless there is some scheme to get others to purchase it at the artificially inflated price, there can be no profit to such a strategy. If there is such a moron, and I tend to doubt it, all he is doing is giving away his money to get an exceedingly temporary better view. By creating an artificial number; he would be simultaneously creating a vacuum to equalize it. Not terribly likely.

Much more likely: Some person or persons continues to like the presumed opportunities at Intrade and is wagering accordingly. They’ll be proven right or wrong on election day when the contracts all expire. Anyone who thinks the “line” is artificially high or low should simply take advantage of their perceived greater wisdom.

The conspiracy implications just struck me as nutty to mention, as they are wholly unsupported by the information provided.

Bill, you seem to be assuming that the market manipulation is being done for a profit; but what if it is being done for a different reason, that is, to make it seem as if McCain is enjoying more confidence from bettors then he really is? It's hard to see a profit strategy out of the pattern of manipulation, but easy to see an electoral strategy.

Well, we'll never really know, will we?

btw, the author of that last piece, Nate Silver, revolutionized the world of Baseball betting by creating models which were far more accurate then Vegas predictions. He is a millionaire, not off of betting, but off of selling his betting tips to subscribers to his service. So I think he may have a clue what he's talking about.

http://en.wikipedia.org/wiki/Nate_Silver

Cycloptichorn

Yeah- expert betting men are always worth listening to.

But "selling" isn't "betting".

In which case he is a professional liar Cyclo, think it through. If there were indeed a system more accurate than the bookies use, it would exist only until the bookies learned of it, because they would use it too. Again, the book doesn't care who wins or loses or by how much. Their incentive to accurately predict the line is to split the gamblers on either side of it. I was apparently too generous when I described him as clueless, because if he's well informed, he is a bald-faced liar. Look again:
Absolutely no risk? $1,000 is what you lose every time the score doesn't hit that narrow margin. That is substantial risk, so this statement is a bald-faced lie, especially if it is coming from an expert.

Clearly, he has an agenda of his own. What would you like to wager his true intention is to sell more of his own product? Since the professional line makers earn millions of extra dollars by being accurate; do you really think they don't apply every effective predictive technique possible? That they wouldn't immediately copy any strategy that worked better? Why do you suppose thousands of number crunchers consistently produce essentially similar opening lines?

The truth is; only fools buy predictions from the shysters who sell them. Beating the vig is difficult enough without paying a sucker-tax on top of it.

I agree entirely Bill.

OK, maybe you need to read up about the guy before making far-reaching conclusions about the professional liar he is based on the couple of snippets of information Cyclo has given you. I mean, you obviously hadnt heard of the guy, so why jump to such far-reaching conclusions on the basis of a couple of paras of argument you get about him from a fellow poster?

I know, I know - you're just going on what you see, and if you just go on what you see, your take may be warranted. But thats the point: if you apparently have never heard of the guy, maybe you shouldnt just go on what you see, if all you see is just a couple of paras of second-hand info?

FWIW, he's a polling expert, whose site is really second to only pollster.com, that I know of, in terms of the sophistication of his models and the level-headed quality of his analyses.

(I'm not so enamoured about his co-blogger on fivethirtyeight.com, but that's a different kettle of fish.)

Can't argue with his results. For a few years he was way ahead of Vegas.

And you are correct, in that they caught up quick enough. He had an edge for a while, as his regression analysis formula relied upon some 'unnamed factors' that for several years nobody could get straight. Eventually Vegas DID figure it out, and the sport of Baseball betting was revolutionized, as most bookies use a formula which is closer to his than not - though all have their own special tweaks.

I got an old friend who works in a sports book on the strip - and he told me several years ago that this guy was the real deal, and was making the odds-makers' lives tough for a few seasons in a row. That's how I found out about him in the first place. For a long time, nobody knew that the 'poblano' who was doing Regression analysis on the election this cycle was Silver, but he revealed it under questioning, and I started paying attention to his site, fivethirtyeight.com ....

Cycloptichorn

Today's daily tracking poll results:

Gallup goes from Obama +3 to Obama +5
Rasmussen goes from Obama +5 to Obama +6
Hotline goes from Obama +7 to Obama +5
Research 2000 goes from Obama +5 to Obama +6

Average of the above goes from Obama +5 to Obama +5.5

Graph:





These numbers still say vanishingly little about the impact of the debate. The overwhelming majority of the interviews for these tracking polls were done before the debate.

What they do illustrate is the impact of McCain's campaign antics this week, the "suspension" of his campaign etc. After a rare eight days straight of near-total stability, with an average Obama lead of 3.3-4% between 18-25 September, there's now been a clear shift. In two days' time, his average lead jumped from 3.3% to 5.5%, with the pollsters being in rare agreement today assigning Obama a 5-6% lead.

This confirms the impression I was talking about when I wrote in the presidential debates thread that "there were tentative signs [in the] polls that McCain was suddenly sinking and that his suspension of the campaign only made things worse." (Obama had some very good state polling too yesterday.) So that's part of how the candidates went into the debate. What that means, conversely, is that even a tie in the debate may actually have been at least relatively good news for McCain.

The CW appears to be that if the debate was a tie, that's basically good news for Obama, because he's ahead and he's not the one who needed a game-changer. But I'm thinking that in this particular context, a tie may have been good news for McCain as well. After all, it changes the impression overnight from him falling behind and treading water to him matching up evenly again.

Even if that makes his polling numbers merely stabilise, nipping some potential polling collapse in the bud, that's one step ahead, and he can go back to planning a recovery again, through their next debate for example. That's not necessarily going to be simple, not least because Palin has to face Biden first (though Palin is helped by how the expectations for her performance are by now set ridiculously low). But it's an improvement upon where he was before the debate - appearing to be suddenly sinking rapidly.

I suppose we ought to differentiate between 'I thought McCain did good/tied/won the debate' and 'I'm going to vote for McCain.' Apples and Oranges. I know that there has been some work done on historical trends of debates and their impact on numbers - once again Silver is involved, iirc - and it was a relatively small amount, 2 points or so.

I think it was a good debate for McCain and definitely what he needed to stay afloat, a bad performance would have hurt him terribly. But I'm not sure how many new people he would have convinced. The data from focus groups and snap polls seems to point towards Obama doing a better job of that.

I wish they would keep polling the debate question into next week, as I wonder how the long-term view will be.

Cycloptichorn

Yes, thats what I was saying, or trying to say at least. He may not have convinced any new people, but if this debate helps him to at least stop the bleeding, that's already good news for him right now, because he had started to drop precipitously.

Where he can go from stabilising his numbers to actually closing the gap again, I dunno...

This is my favourite version of the Pollster.com chart of national polls. I've used the interactivity to start the chart at 1 June, to show the more sensitive trendline, and to exclude both ARG polls and Internet polls.

The most striking part of the result to me is how stable the race has been in the last four months. The only really significant shift came during the Republican Convention, and now we're basically back to the same 3-5% lead for Obama that's been there most of the time since June.

I read (maybe a lil skimming, but mostly read) the Wiki entry Cyclo linked. It was sufficient to tell me the guy knows more than enough about sports betting to damn well know the Lie I dissected above (twice) was a bald-faced lie, even as he wrote it. How do you get around that without considering him a liar?

A lie is a lie, Nimh. Clueless was the only alternative, and clearly he isn't. No quantity of research will make that lie true.

Assuming Last Week's betting line was available 2 points higher and 2 points lower (the 4 point spread from dude's example) on every NFL game last week and you applied his RISK FREE strategy; what do you suppose the result would have been? You would have lost $16,000. ($1,000 Dollars for every game you applied his lie to.) The average difference between the line and the outcome last week was 11 points, and not one game was predicted accurately enough to score under his RISK FREE strategy.

It was a lie when he wrote it, Nimh, and he damn sure knew it. During Football season; Billions of Dollars are wagered and Millions are wasted on Shysters who claim to have a "System". These systems are championed by bookies, who like the Shysters themselves couldn’t care less who wins or loses. The Bookie’s goal is to match the money on both sides of the ball as much as possible… and he couldn’t care less if your particular system makes you win every game. His Vig is made on the other side of that bet. That is how he makes his living: not gambling against you.

I accept that you like and respect the guy. I respect you enough to simply assume that you have very good reason to do so. He might be the world's foremost expert at forecasting polls (and analyzing football for that matter). Maybe he can juggle too. That doesn't change the FACT that he knowingly told a bald faced lie, presumably to increase his own notoriety. Nor does it change the FACT that those who make their living selling gambling systems to degenerate gamblers are profiteering off of human stupidity. That doesn’t have any bearing on whether or not his predictive models or analysis is good or bad. I’ll take your word for it that he’s good, without reservation.

But, if it disturbs you to have his credibility diminished by having his lies pointed out; I suggest you drop him a line and suggest that he stops lying. My pointing it out, and proving it, isn’t the problem.

This sentence betrays your lack of understanding of how sportsbetting works. No system could possibly make the odds-makers' lives tough for a day, let alone a few seasons in a row. No one System Shyster's "report" could ever significantly alter the number of wagers on either side of a game. That's a great story for your friend, however, because promoting belief in systems increases the number of sucker's who'll believe they can buy one... and they'll use it. 50% of all systems work on any given Sunday, so the degenerate gambler is easily fooled.

I knew you would. Anyone who understands the fundamentals of gambling will. I haven’t been providing brilliant analysis; just the facts.

Well, heck. I know the guy personally, and he works in the betting industry. You, on the other hand, are a guy on the internet who asserts that he knows better than the guy in the industry. I think I'm going to go with him. I'm sure you understand.

Cycloptichorn

Do I understand that you'd rather take your friend's word over mine? Sure. Do you understand that is a fallacious argument(namely, "ad hominem")? It makes little difference who delivers the truth.

Do I understand why you'd deliberately not apply your own annylitical mind? Not at all. Unlike your friend; I didn't whisper the "inside scoop" to you. I relayed easily verified factual information about how the industry operates. A working understanding of same is all that's required to render your friend's statement Bullshit.

FACT: No Bookie ever feared a "system" because his income isn't derived from winning gambles.
This is how the system works:
Bookie tries very hard to predict "the line." The line in question is only incidentally predicated on what he believes the most likely outcome of the game to be. The actual purpose of his line is to set it where he believes he will get an equal number of wagers on either side of the line. This is how he limits his exposure to loss potential, and common sense should tell you that this is the only responsible way for him to do his business.

Example: Here in Packerland; more people will want to bet on the Packers than against them based on emotion. This tends to drive the local line further in the Packer's favor than true emotionless analysis would suggest. The line maker will factor this in because he doesn't want to gamble against his clients, even if he believes they are wrong. Why? Because his income comes from the 10% vig attached to every losing bet. By placing the line closer to the equalibrium of gambler's bets; he guarantees his profit, rather than gambling himself. He does not care who wins and loses, because that's not where he makes his money. To the extent that he is never able to perfectly split his clientel's wagers; he carries the risk of the difference. The line, by design, is set to limit this exposure as much as possible. This is a fundemental FACT, Cyclo, not an opinion that can be defeated with an ad hominem argument.

So now we understand that a line is set to equalize the number of bets on either side of it; NOT as the Bookie's side of the gamble (No responsible bookie gambles any more than he has to... and in FACT, most will lay off any substantially lopsided bet on a larger bookie to all but eliminate the risk potential.)

In short; Bookies derive their profit from the vig (the fee attached to every losing wager.) This simple matter of FACT renders your friend's statement Bullshit.

How does a bookie increase his profits? Quite simply; he increases the total number of wagers. He earns approximately 5% of the total wagered (10% of the losing bets.) This profit is COMPLETELY INDEPENDENT of who wins and loses. Belief in the "Holy Grail" of systems among degenerate gamblers is a bookies best friend. Why? Because that very belief encourages more wagering, and his profits are derrived as a rough percentage of the total wagered.

Common sense should tell you no system could possibly work, or the profit potential for the gambler would be virtually unlimited. The Holy Grail doesn't exist. Do some systems work better than others? Of course. Does this scare the bookie (as your friend suggests)? HELL NO! The bookie LOVES it when any system appears to be working... and the longer the better! Why? Because there is absolutely nothing better than a winning streak to encourage a gambler to gamble... and the more wagers the bookie takes; the more money he makes.

The idea that a bookie would be worried about a system working is just flat out ridiculous if you understand these simple fundamentals. The bookie doesn't lose when you win.

More fundamentals:

The only reason an odds-maker is concerned about where to set the opening line; is because his profits are reduced by any wagers that fit in the narrow window of a line's movement.

Example: Packers are favored over the Bears by 3 points. If too many more gamblers take the Packers instead of the Bears, the line will be adjusted upwards to encourage betting on the other side. If too many take the Bears, the line will adjusted downward for the same reason. The float in-between allows for the possibility that the final score will indeed strike between the two, allowing payouts up and down at the same time as described in Silver’s supposedly RISK FREE example. Obviously, this cuts into the bookies profits, so it is in his best interest to accurately predict the line (equilibrium). So, the bookie has to weigh the risk of getting caught in this middle (created by moving the line) against the risk of losing the lopsided portion of wagers if he doesn’t move the line. This is the reason predicting the opening line is important to the bookie.

Now if there was a super “system” that became so widely used that it started to affect the bookie’s float: Said bookie would simply buy it and factor it in to his line in the first place. And he would LOVE IT! Why? Because his lines would be adjusted to return the equilibrium (equalizing wagers on both sides of the wagers), AND he would see a tremendous increase in total wager volume (because winner will bet more! ), which is where he makes his approximately 5%. This, not any imaginary fear of loss potential (that doesn't really even exist), is the reason a bookie encourages belief in systems.

You are smart enough to understand everything I’ve written above, Cyclo. You are also smart enough research it for yourself and satisfy any doubts you may have about any aspect of it. Don’t come at me with some ad hominem nonsense, when the truth is easily verified and has nothing to do with who provides it.

Like I said to Spendy; this isn’t a superior opinion of mine that you have to take my word for. This isn’t really opinion at all. It is simply the only logical conclusion for anyone who understands the basic fundamentals of the business. My apologies if the arrogance in my delivery causes you to resist the simple truth.