The misunderstanding of "beating the closing line."

[uva3021]

the phrase "hedging guarantees profit" has just been used, and marks the beginning of the end for the "Think Tank" forum

there is scarcely any "tank" to store some posters' notion of "think", and has been reduced from a "tank" to a mere "drain"

[MonkeyF0cker]

the phrase "hedging guarantees profit" has just been used, and marks the beginning of the end for the "Think Tank" forum

there is scarcely any "tank" to store some posters' notion of "think", and has been reduced from a "tank" to a mere "drain"

This forum has steadily declined since Ganch's departure.

[trixtrix]

You were correct in "where-are-we?-you-are-on-an-air-balloon" kind of way. Correct yet pointless. Again, that you call BTCL is not what is conventionally called BTCL. Yes, you can call it this way and confuse less experienced posters for no reason as this will not teach them anything. Your terminology is unconventional and pointless because your application of BTCL has nothing to do with TCLs at all. You can buy back at ANY point in time, not necessarily taking TCLs while what you call "market efficiency" is comparing the lines the bettors got down at against TCLs specifically and finding out whether or not they BTCL.

no sir, "while what you call "market efficiency" is comparing the lines the bettors got down at against TCLs specifically and finding out whether or not they BTCL"

still being miunderstood here, what i call market efficieny is market efficiency, it just != btcl. really, is the concept that difficult. you don't have to measure market efficiency in any market in order to profit from btcl. (notice the term btcl was defined by op)

[Wrecktangle]

This has been an interesting thread.

Things are a lot simpler in my world. Rather than worrying about beating the closing line, I worry when my models don't win. When they don't, I know I have a problem.

I can control that, rather than the "arb"itrary movement of the line.

[trixtrix]

the phrase "hedging guarantees profit" has just been used, and marks the beginning of the end for the "Think Tank" forum

there is scarcely any "tank" to store some posters' notion of "think", and has been reduced from a "tank" to a mere "drain"

you're so wise, why don't you frame more quotes out of context to make you look better than us mortals

the full phrase used in context was "hedging/arbbing guarantees profit IN AN INEFFICIENT MARKET", wow seriously dude..

[thepig]

Great topic

[wrongturn]

To people that don't handicap/model games, btcl means profitable arb opportunities and they will happily take them for obvious reasons. But for people that do have a winning system, they don't hedge their plays that beat btcl, at least not totally just for profit without risk, because that is not the purpose, and it also reduces their long term profit.

[uva3021]

you're so wise, why don't you frame more quotes out of context to make you look better than us mortals

the full phrase used in context was "hedging/arbbing guarantees profit IN AN INEFFICIENT MARKET", wow seriously dude..

the context is meaningless when the phrase "hedging guarantees profit" is used in any respect

diabolical misnomer to use the term "hedging" in any context where such would "guarantees profit", i don't care if the market is a clade of dancing sifakas

I suggest you wikipedia or dictionary the term "hedge", then after your vanity is humbled, recalibrate your thoughts

[bztips]

This has been an interesting thread.

Things are a lot simpler in my world. Rather than worrying about beating the closing line, I worry when my models don't win. When they don't, I know I have a problem.

I can control that, rather than the "arb"itrary movement of the line.

Exactly my view, wreck.

But honestly, some of these guys apparently think that you just flick a switch, and voila, you BTCL. It reminds me of the old Steve Martin line:
How to become a millionaire -- first, get a million dollars.

[trixtrix]

the context is meaningless when the phrase "hedging guarantees profit" is used in any respect

diabolical misnomer to use the term "hedging" in any context where such would "guarantees profit", i don't care if the market is a clade of dancing sifakas

I suggest you wikipedia or dictionary the term "hedge", then after your vanity is humbled, recalibrate your thoughts

lol, this coming from a poster who thinks middling and arb are two distinct concepts

ps: i see you have never worked in financial markets were hedging/arbbing (by the purest definition different granted) can be used interchangeably. have you ever heard of the term "Hedge Funds"? you should send them a note telling them they have been misnamed b/c "hedging" is ldo unprofitable

in any case: using this to prove your pt that the think tank is degrading is any case invalid. this is akin to telling the world financial market makers are retarded b/c they are calling themselves "hedge funds". yea like that sooooooo proves they're stupid, dude ldo

[roasthawg]

Exactly my view, wreck.

But honestly, some of these guys apparently think that you just flick a switch, and voila, you BTCL. It reminds me of the old Steve Martin line:
How to become a millionaire -- first, get a million dollars.

I think it's more about using btcl as a measure of a handicapper's success. The btcl crowd don't care if you've won 55% of your games at -110... they want to know how you did against the closing line during that timespan. The btcl take as I understand it is that given two models that make 1000 plays each where one picks 55% winners and beats the closing line 48% of the time and the other picks 50% winners and beats the closing line 54% of the time they would much prefer the latter model to the former. I don't think anyone on either side is suggesting it's easy to either pick winners or btcl.

[Thremp]

I did get a good laugh skimming the last 2 pages for Data's grouping of the Ss and Ts. Was pretty lolz.

[trixtrix]

not to keep on beating a dead horse (or a bunch S's), but to illustrate why this pt is important:

people are asking how does this concept make me money when it's difficult to predict how lines will move and therefore btcl in an INEFFICIENT environment.

my response is you have the concept backwards, it will help you profit b/c WHEN you do btcl, you can better risk manage your betting portfolio. remember, wallstreet banks didn't get into trouble b/c they TOOK on risk, it's b/c they did not properly MANAGE their risk.

when deciding when or if you should buy back on the closing line when you do btcl, most bettors only consider two main factors: the size your exposure on the original bet (relative to your br of course), and how large of an arb you're getting back.

BUT there is a third factor that is rarely considered/mentioned even by some experienced bettors: that is how efficient you believe that particular market is. remember there is strong efficiency and weak efficiency, dependent on the size, type, timing, exoticness of the sportsbetting market you're wagering on. the more INEFFCIENT you believe this particular market is, the more you should be inclined to arb/buyback. (this is akin to advicing someone that the LOWER the projected total of a game, the more inclined you should be towards tease/buying pts on that particular spread)

the decision to NEVER arb is as INCORRECT as the decision to ALWAYS arb, there is no predefined 100% certainty in anything in sportsbetting (and perhaps life in general)

some will say that arbbing is easy concept to understand and these concepts are self evident. but then they indicate that they don't understand why this particular concept makes btcl is important.

btcl is NOT ONLY important when you can use it as a measure of your betting prowess when the market is efficient, but it's a tool you can use to manage/mitigate your risk when the market in INEFFICIENT. the culture of risk-managing is lacking everywhere, even (or perhaps especially so) in wallstreet. considering when to arb WHEN you DO btcl play an important part of that.

appendix:

btcl: when you beat the closing line enough so that opposite sides of opening bet and closing bet becomes arbable. ie when the sum of implied win% between the 2 bets are < 100%

market efficiency: the process of which lines move toward a risk-neutral implied fair line that is indicative of the true percentage of the actual outcome.

[Shonner]

You were correct in "where-are-we?-you-are-on-an-air-balloon" kind of way. Correct yet pointless. Again, that you call BTCL is not what is conventionally called BTCL. Yes, you can call it this way and confuse less experienced posters for no reason as this will not teach them anything. Your terminology is unconventional and pointless because your application of BTCL has nothing to do with TCLs at all. You can buy back at ANY point in time, not necessarily taking TCLs while what you call "market efficiency" is comparing the lines the bettors got down at against TCLs specifically and finding out whether or not they BTCL.

Data -- thank you for stating this, you are much better with words than me. This is what I have been trying to say for 3 days.

24 points transferred your way. Gracias.

[Justin7]

trixtrix,

I have to violently disagree with you. Most sports betting markets are ruthlessly efficient. But don't take my word for it -- do this test:
pick any major market. NFL, MLB, NBA, NCAAF, or NCAAB. Compare the closing line to the opening line. Any move reflecting a 3% move in theoretical win rate, log an imaginary bet on the "right side" of the opener. These sharp bets will typically show a return of the theoretical EV difference between the two line, plus about 1.5%. Openers will look better than they "should" if you're evaluating them to BTCL.

The markets are not completely efficient. There is juice. This effect means a 5% delta between the opening value and the theoretical value might be reflected as only a 3% line move. BTCL will over-perform your expectation, again and again.

If you are using BTCL to scalp/arb, you are pissing thousands of dollars away. You lower your variance to almost 0, and that is fine if your bankroll is small. If you have a 5k bank, and want to tie up 1k total on both sides, sure. But if you have a 100k bank, take a 3k position, and scalp out the other side at about 3k, you just pissed away a lot of value.

[Data]

Shonner, thank you for the compliment.

trixtrix, unless you can find a solution to the problem with your terminology (and, effectively, with most everything you posted in this thread) that I typed in bold in my previous post you will remain guilty of bringing an irrelevant subject to this discussion. The point you are making is understood, I doubt it is that much complicated that there is anyone who did not get what you are saying. The grief with your posts is not due to that point of yours but due to the fact that you should not be calling BTCL a method where TCLs play no role and may not be involved at all.

[mathdotcom]

Justin do you ever violently disagree with your wife?

[Data]

If you are using BTCL to scalp/arb, you are pissing thousands of dollars away. You lower your variance to almost 0, and that is fine if your bankroll is small. If you have a 5k bank, and want to tie up 1k total on both sides, sure. But if you have a 100k bank, take a 3k position, and scalp out the other side at about 3k, you just pissed away a lot of value.

Justin7, you are wrong here. Maximizing EG is not pissing money away but quite in reverse. To concentrate on +EV instead, the bettor must be risk-neutral and $100K bankroll is absolutely insufficient for this. It must be thousands if not millions times large than this.

[MonkeyF0cker]

I have to disagree with both of you. It depends on the scalp.

[MonkeyF0cker]

As well as that individual's utility...

Although, you did allude to that Data .

[Justin7]

Justin7, you are wrong here. Maximizing EG is not pissing money away but quite in reverse. To concentrate on +EV instead, the bettor must be risk-neutral and $100K bankroll is absolutely insufficient for this. It must be thousands if not millions times large than this.

Data,

hedging is not maximizing EG if your bankroll is sufficiently large.

Consider this example: you have a 100k bankroll. An an NFL line opens at +3.5. God whispers in your ear, "this line will close at +2.5" (or you have a model that is 100% accurate, and predicts CTL perfectly). Assume the "3" on this game will push at 10%.

Our imaginary bettor bets 3k on dog +3.5. Should he hedge 100% at -2.5? Or would he be better of not hedging at all (there is an optimal hedge amount between 0 and 3k. feel free to provide this amount, but it will support the concept that not hedging, and riding your bet is better than hedging in this example).

The focus of this example is... Once your bankroll is sufficiently large (and 100k meets this criteria in this example), hedging with the closing line at -110 is lowering your growth. Most pro bettors with that capital are better off embracing volatility than hedging with those commissions.

[Justin7]

I have to disagree with both of you. It depends on the scalp.

It's hard to refute that. If you can hedge with 0% commission, it makes sense to hedge almost every play if you BTCL.

[MonkeyF0cker]

Data,

hedging is not maximizing EG if your bankroll is sufficiently large.

Consider this example: you have a 100k bankroll. An an NFL line opens at +3.5. God whispers in your ear, "this line will close at +2.5" (or you have a model that is 100% accurate, and predicts CTL perfectly). Assume the "3" on this game will push at 10%.

Our imaginary bettor bets 3k on dog +3.5. Should he hedge 100% at -2.5? Or would he be better of not hedging at all (there is an optimal hedge amount between 0 and 3k. feel free to provide this amount, but it will support the concept that not hedging, and riding your bet is better than hedging in this example).

The focus of this example is... Once your bankroll is sufficiently large (and 100k meets this criteria in this example), hedging with the closing line at -110 is lowering your growth. Most pro bettors with that capital are better off embracing volatility than hedging with those commissions.

This is a bit pointless as not only could Data provide countless examples to the contrary, but you're assuming that the market is 100% efficient.

[Data]

It depends on the scalp.

Sure does, there are number of quantifiable parameters. Thanks to Ganchrow and his line_change spreadsheet for allowing most of us hedge more or less properly. I am unsure what you are disagreeing with but do share, while I am sure that the concept is right, I might be overlooking something.

[MonkeyF0cker]

Sure does, there are number of quantifiable parameters. Thanks to Ganchrow and his line_change spreadsheet for allowing most of us hedge more or less properly. I am unsure what you are disagreeing with but do share, while I am sure that the concept is right, I might be overlooking something.

Sorry, Data. I think I misread the intentions of your post. I think we're on the same page then.

[Justin7]

This is a bit pointless as not only could Data provide countless examples to the contrary, but you're assuming that the market is 100% efficient.

Monkey,

The market is not 100% efficient due to vigorish. The big sports are efficient enough.

Countless examples to the contrary? Please provide a single example where 1000 plays in a fixed time period had better openers than closers for one of the 5 major markets.

[MonkeyF0cker]

Monkey,

The market is not 100% efficient due to vigorish. The big sports are efficient enough.

Countless examples to the contrary? Please provide a single example where 1000 plays in a fixed time period had better openers than closers for one of the 5 major markets.

1 game is not 1000 plays. There's a pretty clear difference. Just because the markets are unbiased as a whole long term, doesn't mean that the market for a particular game is not. You basically defeat your own premise in this post. How can you assume 10% when, even if the market were unbiased, you couldn't possibly say FV isn't +2.5 -110?

[Data]

hedging with the closing line at -110 is lowering your growth

Justin7, as MonkeyF0cker already pointed out, you're assuming that the market is 100% efficient. I would rather say that the no-vig closing line reflects true odds. Without that assumption, it is better to arb and to have no position. That is because the value could very well be in taking -2.5.

Even assuming the above and even while paying that much juice it is frequently appropriate to buy back. No, not while laying only 3% at these insane edge. It seems much more reasonable to assume that having at his disposal the sceen full of books and exchanges the hedger can buy back at no vig line which will make appropriate buying back most of the initial stake, about 80% of it.

[Justin7]

1 game is not 1000 plays. There's a pretty clear difference. Just because the markets are unbiased as a whole long term, doesn't mean that the market for a particular game is not.

Huh? The question is: how efficient is the market, not a single game. I put a hypothesis out: on any set of 1000+ plays in a major market over a time-based sample, the closing prices are more efficient than closer. I spelled out a test to verify this.

If you're worried about a single game, follow whatever whim you want. What do you suggest as a long-term strategy, over the next 1000 games, if you have a model that BTCL?

[trixtrix]

trixtrix,

I have to violently disagree with you. Most sports betting markets are ruthlessly efficient. But don't take my word for it -- do this test:
pick any major market. NFL, MLB, NBA, NCAAF, or NCAAB. Compare the closing line to the opening line. Any move reflecting a 3% move in theoretical win rate, log an imaginary bet on the "right side" of the opener. These sharp bets will typically show a return of the theoretical EV difference between the two line, plus about 1.5%. Openers will look better than they "should" if you're evaluating them to BTCL.

The markets are not completely efficient. There is juice. This effect means a 5% delta between the opening value and the theoretical value might be reflected as only a 3% line move. BTCL will over-perform your expectation, again and again.

If you are using BTCL to scalp/arb, you are pissing thousands of dollars away. You lower your variance to almost 0, and that is fine if your bankroll is small. If you have a 5k bank, and want to tie up 1k total on both sides, sure. But if you have a 100k bank, take a 3k position, and scalp out the other side at about 3k, you just pissed away a lot of value.

okay so what if you btcl in mma, boxing, wnba, or japanese baseball? should you hedge? what factors do you use to determine whether to arb or not?

sure to most big bettors w/ br big enough that never run into eg issues when betting small markets, yes you can probably never justify hedging.

simple question: a small bettor w/ 10k br that bet 500 on japanese baseball that moved 50c, he also bet 500 on a game in NFL that moved 50c. would his decision whether to hedge (partially) or not for both games be the same?

[uva3021]

Our imaginary bettor bets 3k on dog +3.5. Should he hedge 100% at -2.5? Or would he be better of not hedging at all

of course he should hedge because, as we all know "hedging guarantees profit"

[MonkeyF0cker]

Huh? The question is: how efficient is the market, not a single game. I put a hypothesis out: on any set of 1000+ plays in a major market over a time-based sample, the closing prices are more efficient than closer. I spelled out a test to verify this.

If you're worried about a single game, follow whatever whim you want. What do you suggest as a long-term strategy, over the next 1000 games, if you have a model that BTCL?

If you trust the market over your model, why not just steam chase then, Justin? It would seem to me you're just doing a lot of work for nothing if you want to put everything on how the market reacts.

There are plenty of games that have closing value.

[Justin7]

okay so what if you btcl in mma, boxing, wnba, or japanese baseball? should you hedge? what factors do you use to determine whether to arb or not?

sure to most big bettors w/ br big enough that never run into eg issues when betting small markets, yes you can probably never justify hedging. but what do you tell bettors w/ 10k br that bet 500 on japanese baseball that moved 50c, vs. nfl that moved 50c? are your decisions to arb or not the same?

There are 3 things the bettor must know:
1. How large is his BR relative to the market he is betting? If he has a 10k br and can bet 2k on beavers, it makes more sense to overbet and hedge. If he has a 50k br, BTCL by 5%, and bets 2k, it might not make sense to hedge at -110.
2. How efficient is that market? Smaller markets are often less efficient. This weighs against hedging if your model is good. The opener is +5, fair line is +3. Market closes at +4. Your hedge does better against an efficient closer.
3. How badly are you getting juiced in these smaller markets? MB won't help you as much in these. If you want to hedge, you might get stuck laying -110 (or worse).

As with any problem, there is an exact mathematical solution if you have all the facts.

[Justin7]

If you trust the market over your model, why not just steam chase then, Justin? It would seem to me you're just doing a lot of work for nothing if you want to put everything on how the market reacts.

There are plenty of games that have closing value.

Monkey,

I'm trying very hard to take you seriously. Is this a serious question?

[trixtrix]

There are 3 things the bettor must know:

2. How efficient is that market? Smaller markets are often less efficient. This weighs against hedging if your model is good. The opener is +5, fair line is +3. Market closes at +4. Your hedge does better against an efficient closer.
.

disagree entirely w/ this post, in less inefficient markets, MAYBE your model is good, MAYBE it's not , you can't tell b/c you can't trust the closing line, the weighs for hedging.

in your example, the fair line could be +3 or it could be -3 (how can you be so certain) unless you have a wallop of a model