Opening lines accuracy

[pedro803]

always too much bickering in the think tank

Now, when you say "always" do you mean in every thread, or do you mean that somewhere amongst the threads there is always an instance of too much bickering?

That is, are you using the term "always" in the traditional sense of a time bound qualifier? If so, I'm not sure if it makes sense because who is to say if there is always someone reading in the think tank, if a tree falls in the forest and there is no one to hear it does it make a sound?

Even more pointedly, it is not useful to say always in the context of an internet forum because once there is one instance of anything posted in the think tank it can always be found in the think tank (barring the extenuating circumstance that it is deleted by the poster or an admin). Therefore, the first thread that featured "too much bickering" has forevermore (or until deleted whichever comes first) rendered the think tank to "always" have too much bickering. So always doesn't differentiate much since once anything is posted here it is always here (I will at this point dispense with the discussion of deletion)

Or, are you using the term "always" to refer to the cyber real estate out in the ether known as the think tank? Something more akin to the traditional term "everywhere?" So if this is the case your post would read:

"everywhere too much bickering in the think tank"

this usage would be problematic also because not every single thread in the think tank features bickering, there are almost certainly some single post threads that do not indicate opposing views of any sort! So there is definitely not too much bickering "everywhere" in the think tank!

perhaps it is the case that our language has not adapted fully to the technology of internet message boards but pray tell, exactly what do you mean by "always"?

Do specify kind sir!

[pedro803]

Yes, closers will tend to be 50-50, but that is for all closers. If you truly randomly (throw a dart) select a sub-set of closers, that sub-set should also be close to 50-50. The problem comes in when your sub-set is not selected truly at random. That sub-set may or may not have the properties of the entire data set. The only way that one can "know" if a wager is + or - EV is to compare the results of a large enough sample of wagers selected by the same criteria.

I like this, somewhere along the line in that long thread started by PokerJoe about BTCL I remember having a thought close to this -- and our plays are by design just such a non-random sub-set.

I can see how Justin7's original eureka moment was really valid about how to pick out winners when you haven't seen many of their plays, but that doesn't mean that it is not possible to win without BTCL from a modeler's point of view (maybe quite unlikely yes, but not impossible)

[CFA]

Theoretically speaking the answer is yes the number should always close on its true market value and the opening number would have no effect on that. considering that a moneyline is just an expression of the probability of a team winning the line should almost always close withinin the accurate range of each teams implied win percentage.

False. The market implied win percentage will not always be captured by the zero vig line.

[pedro803]

Now, when you say "always" do you mean in every thread, or do you mean that somewhere amongst the threads there is always an instance of too much bickering?

That is, are you using the term "always" in the traditional sense of a time bound qualifier? If so, I'm not sure if it makes sense because who is to say if there is always someone reading in the think tank, if a tree falls in the forest and there is no one to hear it does it make a sound?

Even more pointedly, it is not useful to say always in the context of an internet forum because once there is one instance of anything posted in the think tank it can always be found in the think tank (barring the extenuating circumstance that it is deleted by the poster or an admin). Therefore, the first thread that featured "too much bickering" has forevermore (or until deleted whichever comes first) rendered the think tank to "always" have too much bickering. So always doesn't differentiate much since once anything is posted here it is always here (I will at this point dispense with the discussion of deletion)

Or, are you using the term "always" to refer to the cyber real estate out in the ether known as the think tank? Something more akin to the traditional term "everywhere?" So if this is the case your post would read:

"everywhere too much bickering in the think tank"

this usage would be problematic also because not every single thread in the think tank features bickering, there are almost certainly some single post threads that do not indicate opposing views of any sort! So there is definitely not too much bickering "everywhere" in the think tank!

perhaps it is the case that our language has not adapted fully to the technology of internet message boards but pray tell, exactly what do you mean by "always"?

Do specify kind sir!

What? nobody liked my joke? lol And I thought it was so funny and I worked hard on it too. You gotta admit we do split a few hairs in the old think tank

oh well guess I better not quit my day job -- come to think of it, its time to go there now

everyone, I am putting my popcorn away there is nothing to see here

[subs]

False. The market implied win percentage will not always be captured by the zero vig line.

sorry CFA if you have a minute could you explain this please. i thought that the no vig line was the implied market %. are we basically splitting hairs or is there some important lesson here that my 89.9 degree angled mind can not grasp by itself? are you saying the market % will be somewhere between the line and the no vig line? much appreciated.

Pedro not too bad, you brought a little lightness which was needed. but i'd say half the laugh is in the timing.

[CFA]

Unless the book that you believe best represents the true market moves on air from MB (or if you think MB is the true market), the vig will prevent the no-vig line representing the market implied win probability in games where the markets implied win probability is different from the opening line.

Suppose book A is the most accurate representation of the marketplace. There are 4 games in which the markets implied win probability is exactly 50% on both teams. Book A opens the following 4 Games as follows. Ask yourself what will be the closing line on all these games, and what will be the no-vig line on all these games if the market is acting in 100% efficiency.
Opening lines
Game 1:
Team A +104
Team B -112
Game 2
Team C -102
Team D -106
Game 3
Team E -104
Team F -104
Game 4
Team G -108
Team H +100

[statictheory]

Wow... Is this post intended to just confuse the issue more?

(1) There's no paradox here. There's the true odds of an event happening (defined by numerous factors) that handicappers are attempting to calculate, and then there's the implied probability of an event happening (defined by the market activity on the ML). They're not the same thing. The market price is set to balance betting so it may not always accurately reflect the true odds of the matchup. However, in general, you could say that a line is much more likely to move in the direction of it's true value over time, with the understanding that it will not necessarily close at the optimal value (or always move in the expected direction) due to external influencing factors that do not directly affect the match (like public opinion).

Couple of questions. Wouldnt the first change away from the original line be the smart money(regardless of how the line moves after) since it would be implied they are hitting a bad number? Second, how can any model or individual accurately predict in the NFL when it has been shown that turnovers have such a high correlation to winning, but cant be predicted with any real accuracy.
So if the number 1 factor in most game outcome is turnovers (example: san diego/ Raiders), how can you expect to succeed?

And again, market effeciency relates to information, but is information the book balancing the betting action?

[likelike0611]

Good one dude..keep doing nice work

[George7904]

Couldn't one take an entire year's worth of MLB (or many years), for instance, and determine whether the opening or closing line was more accurate by looking at the actual results and comparing the 2 lines? If all +100 opening lines end up winning 50% of the time, then it is more accurate than if all closing lines at -115 win at 50%. Do this for all opening and closing numbers and see which is more linked to the actual results.

For example, one could record all opening/closing numbers and see at what percentage they win. It seems like this would mathematically solve the argument once and for all.

However, I can see that if the opening number is more accurate; it wouldn't be for long because everyone would bet closing numbers that were different and make a fortune.

[subs]

CFA - thank you sir!! sent you a few points.