[HUY]

Say you have a model with an expected ROI of 0 when it bets at closing lines. What would its ROI be if it bet at opening lines instead?

[mebaran]

In my experience with baseball, I was getting about 25-75 cents in value (depending on the amount of games) by hitting openers. I actually wasn't even hitting closers either. I would get home from work at 5pm and hit those lines (which are far from closing). So I basically just made sure I hit the openers within 30 minutes of them being posted.

[hutennis]

@OP

Please, try to think. Please!!!!

Where that advantage you are looking for would come from?
Why would any possible advantage stay available for more than a split second
when millions are competing for picking it up first?
Who the hell are you to be able to win in that race?

@mebaran

You are not getting a wooden penny in value.
You are simply confused.
You are not paying 25 to 75 cents less for the same product.
The old product is gone and not available anymore.
Based on new information available (whatever that information may be) new product was created, properly priced, juiced up
and made available to you and anyone else with a new price tag.
There is no "open" "close" or "in between". Those are nothing more but arbitrary points in time.
Essentially, every new price is a new "open" with the same amount of "value" in it as previous "open" had.

[HUY]

@OP

Please, try to think. Please!!!!

Where that advantage you are looking for would come from?
Why would any possible advantage stay available for more than a split second
when millions are competing for picking it up first?
Who the hell are you to be able to win in that race?

@mebaran

You are not getting a wooden penny in value.
You are simply confused.
You are not paying 25 to 75 cents less for the same product.
The old product is gone and not available anymore.
Based on new information available (whatever that information may be) new product was created, properly priced, juiced up
and made available to you and anyone else with a new price tag.
There is no "open" "close" or "in between". Those are nothing more but arbitrary points in time.
Essentially, every new price is a new "open" with the same amount of "value" in it as previous "open" had.

Let's say there is no new information for the match in question. Then the line movements are because of sharp money, which means that I'm better off playing at opening odds rather than on closing odds.

[hutennis]

Let's say there is no new information for the match in question. Then the line movements are because of sharp money, which means that I'm better off playing at opening odds rather than on closing odds.

Jesus!

How can you properly assess the true nature of that move?
Has someone made an announcement?
What if there is a new information?
Who gave you a right to completely discount the possibility?
What do you know about true intentions of sharp money?
What if it is a trap?

[Justin7]

Jesus!

How can you properly assess the true nature of that move?
Has someone made an announcement?
What if there is a new information?
Who gave you a right to completely discount the possibility?
What do you know about true intentions of sharp money?
What if it is a trap?

Hutennis,

quit trolling here. He defined an assumption. Work within that assumption. If you troll anywhere in htt, you're out of here for a long time.

[Justin7]

Say you have a model with an expected ROI of 0 when it bets at closing lines. What would its ROI be if it bet at opening lines instead?

One way to get an estimate is back test your model against closers and openers, and look at the difference.

The bigger the market, the better the openers tend to be. Smaller markets tend to see bigger movements from open to close. Ergo, if you have a model that beats closers, you'll have bigger gains hitting openers in the smaller markets than the mid-size or larger ones.

[hutennis]

Hutennis,

quit trolling here. He defined an assumption. Work within that assumption. If you troll anywhere in htt, you're out of here for a long time.

Hah. Funny.

Not that I care in a least, but...

Here is a definition of trolling

In Internet slang, a troll is someone who posts inflammatory, extraneous, or off-topic messages in an online community, such as an online discussion forum, chat room, or blog, with the primary intent of provoking readers into an emotional response or of otherwise disrupting normal on-topic discussion

That's not me. The only thing I'm trying to provoke is critical thinking and using common sense.
That's probably does not go too well with your agenda so say so. Don't bring that trolling bullshit out. Be honest.

Now about assumption "there is a move but there is no information" and working within it.
The assumption is ludicrous and should not be made b/c, in a absence of confirmatory evidence you'll be right and wrong randomly.
It's like working with an assumption that the next number on a roulette wheel will be red.
Useless.

I don't see how explaining it to someone who is confused can be considered "inflammatory, extraneous, or off-topic".

[HUY]

Now about assumption "there is a move but there is no information" and working within it.
The assumption is ludicrous and should not be made b/c, in a absence of confirmatory evidence you'll be right and wrong randomly.

The assumption is also that there is a model which has an expected ROI of 0 at closing odds. I assume that you don't believe that to be ludicrous as well. So you claim that there is no advantage to bet at opening odds rather than on closing ones with that system. But it is this last position that is ludicrous, as you can take the closing lines and add random noise to them (to simulate opening odds) and you will see that if the system bets at them it will have a higher ROI (this is self-evident really, the only reason to do it is to try to get a quantitative idea of the gained edge).

One way to get an estimate is back test your model against closers and openers, and look at the difference.

That is my problem though, I have no opening odds for my system's sport, snooker. I am trying to simulate them by adding gaussian noise to the closing ones, but I was unsure what the variance of the noise should be.

[Justin7]

That is my problem though, I have no opening odds for my system's sport, snooker. I am trying to simulate them by adding gaussian noise to the closing ones, but I was unsure what the variance of the noise should be.

In a small market (snooker is small, right? I seem to remember 500 limits at Greek), I would just start tracking imaginary plays.If you record 30 plays, and 24 move your way, you probably have a winning program.

[hutennis]

You have never seen a sporting event with no critical, new information released to the public where there was a significant line move?

You don't believe in the existence of long-term, successful cappers? These are the types who can, in theory at the very least, beat opening lines.

I think you're taking EMH to an extreme with your arguments. Your use of terms like "ludicrous" and "common sense/critical thinking" are totally misguided. Do you seriously think that every profitable opportunity is sports or even financial markets is pounced on in a matter of seconds? From my own experience, this is totally inaccurate. The markets are not *that* efficient.

Looks like we need to talk about EMH a bit, do we? I hope it will not be considered as trolling for a change.

EMH comes in 3 separate forms.

1. Weak form.
Current market price incorporates all past publicly available information. Past market statistics provide no information that can be used by anyone to predict future prices.
Countless studies have been done and this claim has been empirically confirmed.

2. Semi strong form.
Market reflects all publicly-available information. New information is being priced in too fast for most of the participants to be able to gain an edge.
So, those with vastly superior information gathering and processing methods can have an edge.
No one argues with that either.

3. Strong form.
Market is always right. No one can have any information or do anything to gain an edge over market. Period.
Now, this is the one that failed miserably. Market is venerable to fraud, manipulations, using inside information, deception, collusion etc.

Even a biggest proponent of "Market is always right" theory, MFing Alan Greenspan had to finally admit: "I was wrong".
He did, but not before strong form of EMH became a intellectual foundation for a skillfully orchestrated campaign of deregulations (who needs regulations when market is always right no matter what anyone does) that enabled bunch of financial oligarchs to perpetrate
a massive fraud, corrupt capitalist system to a huge extend, create unimaginable problems around the world and make out in a process like all other bandits in a history of the world combined.

Back to sportsbetting and this forum and your quote though.

I don't take EMH to extreme. I simply don't apply failed strong form of it to subjects discussed here.
Why? Because it don't belong. No one (99.9%?) here or anywhere else in SB world are in a position to take advantage of fraud, manipulation, inside information etc. That's given. Moreover, there is also no way to benefit from a simple knowledge that someone out there can. We just don't belong to a club.

So when I post my "market efficiency trolls" I look at things from form 1 and form 2 perspective.
And from that perspective markets are "that" efficient. Based on scientific evidences the gotta be.
I mean whether you believe it or not thats your issue. Some people believe in a tooth fairy. Some in burning bush.
Some people even believe Earth is less than 10000 years old. People do believe weird things, you know.

But based on what we know about form 1 it's very safe to say that building models based on historical data trying to predict future results
is nothing more but a stupid waste of time and should be done only if you have no better things to do.

And form 2 applied to SB simply means that by the time you finished listening to Mike and Mike or done analyzing internet wright ups
its already way too late. Every last bit of useful information has been already priced in. So if you don't have a sister who is starting pitcher's GF there is not much you can do as far as getting an edge.

Well, as Forest Gump so eloquently put it, that's all I have to say about that.

[jgilmartin]

That is my problem though, I have no opening odds for my system's sport, snooker. I am trying to simulate them by adding gaussian noise to the closing ones, but I was unsure what the variance of the noise should be.

http://www.oddsportal.com/snooker/results/ has opening and closing odds going back a few years in some cases (depends on the league/competition). You would have to scrape it if you wanted to do the backtest automatically, or you could manually go through each match if you have the patience.

[HUY]

http://www.oddsportal.com/snooker/results/ has opening and closing odds going back a few years in some cases (depends on the league/competition). You would have to scrape it if you wanted to do the backtest automatically, or you could manually go through each match if you have the patience.

It looks like the closing odds include live betting for some bookmakers, is it really so?

[big0mar]

What kind of action can you get down on snooker???

[MonkeyF0cker]

What kind of action can you get down on snooker???

The guy talking about betting Croatian soccer in aisle 5 doesn't know the limits on snooker?

[HUY]

What kind of action can you get down on snooker???

If the match is televised you can get in about 20 thousand most of the time.

[jgilmartin]

It looks like the closing odds include live betting for some bookmakers, is it really so?

Not sure if they are accurate or not, sorry. The archived full game odds have been accurate from my experience (I don't know about snooker in particular, I am speaking in general). The main errors I've seen have been in really small markets where the opener will be up at Pinny for such a short time before someone makes a limit bet on them, that Odds Portal hasn't even recognized that Pinny had a number up yet, and will record the number AFTER they move the line as the opener, when in fact there was a better number available (albeit for a very short period of time, and for a very low limit).

[HUY]

Not sure if they are accurate or not, sorry. The archived full game odds have been accurate from my experience (I don't know about snooker in particular, I am speaking in general). The main errors I've seen have been in really small markets where the opener will be up at Pinny for such a short time before someone makes a limit bet on them, that Odds Portal hasn't even recognized that Pinny had a number up yet, and will record the number AFTER they move the line as the opener, when in fact there was a better number available (albeit for a very short period of time, and for a very low limit).

I'm not referring to the accuracy per se, but rather to whether or not live betting can be excluded from the odds. For example, examine this:

http://www.oddsportal.com/snooker/wo...8/#home-away;2

You see that the real closing odds were about 1.5 but for the bookmakers that played the game live the data include the live odds which closed at about 1.02 (or at least that the last odds the site recorded).

[jgilmartin]

I'm not referring to the accuracy per se, but rather to whether or not live betting can be excluded from the odds. For example, examine this:

http://www.oddsportal.com/snooker/wo...8/#home-away;2

You see that the real closing odds were about 1.5 but for the bookmakers that played the game live the data include the live odds which closed at about 1.02 (or at least that the last odds the site recorded).

Ahh...yeah, I see what you are saying. I've never seen that before on the site (although I haven't looked at snooker odds much at all). Don't know anything else about it, I'm afraid. Sorry