CBB totals gravy train derailed?

[Pokerjoe]

LOL, I think anyone who works with programs must often think to themselves, when finding their mistakes, "I'm a moron."

Scoring last year was pretty static, so I never noticed until this year, when scoring has changed, that I'd never dropped a formula constant I'd put in last year while writing spreadsheet formulas. I am sometimes a moron, no doubt. Though as it turns out I would have won 2 fewer units this year, so I've been lucky. Like last night's over bet on Idaho St/Utah, 128 when I bit, closing at 125, but going over, against the grain. -FML.

Still, it's a low total day. Average totals in Dec, not including today, were about 134', and today are only 132' (early lines). Could be slate-dependent, some of it, obv, but mostly it's the market quickly adjusting, and still, possibly, to some extent, overadjusting.

[Wrecktangle]

Poker, one problem with M$Bill software is, it "likes" to insert screw-ups at random times. The roll-up sheet we use for CFB, has a spurious issue that re-occurs every week. I've called Billy-Bob's Boys and still no fix yet. And it's in Excel 2007.

Trust nothing on a spreadsheet, keep a set of copies as you move along (I never toss out any sheet for just this reason), and the bigger the sheet, expect more errors.

Can't live without it though.

[MonkeyF0cker]

Comical conversation, gentlemen. Carry on.

[Wrecktangle]

BTW, the ignore feature is a beautiful thing.

[Pokerjoe]

Comical conversation, gentlemen. Carry on.

I agree.

[Pokerjoe]

Of course, considering CBB totals today are 27-14 so far, over, it's a good kind of comical. This is a sick day, absolutely unreal. May end up being the best day of my sports betting life.

[Pokerjoe]

In fact Alabama -1 has to lose or else I'm 15-0 on the day, which is sick, it shouldn't even be allowed. If Bama comes in I'm going to end up wishing I was one of those drunks in the sportsbooks who make $5 15-team parlays for the drink tickets.

[MonkeyF0cker]

Of course, considering CBB totals today are 27-14 so far, over, it's a good kind of comical. This is a sick day, absolutely unreal. May end up being the best day of my sports betting life.

Wait. How can that be? There must be something wrong with your model.

[Pokerjoe]

Okay, the sickness continues, I'm running so good it just isn't right, all time best weekend ever, ainec: 3-0 NHL bets Friday night, 2-0 CBB bets then; 14-1 CBB yesterday; 2-0 NFL bets today, and 5-0 in Leroy's NFL contest (to move into first place in that). Which is all, yes, wayyyy off topic, but at least if I'm ever a whiney bad-beat douche you all can pull this thread up and LOL.

More to the point, it's interesting that a major market development takes place in CBB totals and no one notices, and in fact MonkeyFock thinks it's funny that I even think about such a thing (it's okay, I think it's funny that he doesn't think about it).

Even after fixing my slight model bug I was still 1.8 points tall, on average, on Saturday's CBB totals, and that just doesn't happen to me, in any sport ever, and, yes, to anyone following the sport it should be hugely interesting, that the market totals would drop so far so fast and, as it turns out, probably so wrongly. There may not be anything more important in handicapping or modeling than the underreaction vs overreaction battle, and here was a classic example with the sample size that only CBB provides, and interest in it is nil? Actually, that's a good thing. Just as is the fact that the almighty market can sometimes get truly hysterical, that's also reassuring and interesting.

Overs are 39-21 in CBB this weekend, since I first really noticed that the market's scoring estimates were in free-fall, on Friday (there are some games undecided yet today). Usually we'd all say, of results like that, well, it's just variance, no biggie. But this is a case where it isn't the results that moved away from the norm, but where the market did the moving, and that's a hugely different animal.

It's as if you looked at a slate and realized the market had collectively decided there was no longer any such thing as home court advantage. If that happened you'd probably first be suspicious as hell, and check your work, and then you'd just shake your head, wondering WTF, this has to be too good to be true, right? The market just up and decides there's no more HCA?

But in this case the market just up and decided to give 2 points to the overs for the day. Now, two points would not itself account for 39-21, so surely there's still a good bit of mere variance in that result. But then again, if you consider regression to the mean ...

Anyway, I know when I go off like this the poseur sharps generally just giggle. That's also a good thing.

See you guys after the holidays. Merry Christmas.

[MonkeyF0cker]

Poseur sharps don't have "constants in their model from last year." (LOLWTF?!?)

Obviously, a properly backtested model should be adjusted EVERY DAY against a 60 game sample rather than verifying it over the course of SEASONS.

The Book of Joe. So it is said. So it shall be done.

Incredibly inspiring, TRUE SHARP.

[suicidekings]

Okay, the sickness continues, I'm running so good it just isn't right, all time best weekend ever, ainec: 3-0 NHL bets Friday night, 2-0 CBB bets then; 14-1 CBB yesterday; 2-0 NFL bets today, and 5-0 in Leroy's NFL contest (to move into first place in that). Which is all, yes, wayyyy off topic, but at least if I'm ever a whiney bad-beat douche you all can pull this thread up and LOL.

As you say, it's hard not to attribute variance/luck for a large amount of a result that's so far above expectations, but at the same time, considering how often a failure to recognize a change leads to disaster, I would say congrats on taking advantage of the situation. Nice weekend

[MonkeyF0cker]

I think Justin is responsible for the idea of "beating the closing line." He writes about it in "Conquering Risk." He was doing some work for Pinnacle and noticed that the difference between a player's betting line and the later, closing line, more quickly predicted future success than mere win/loss records.

Justin's view was the bookie's view: when profiling players, the sharps are the threat, not the lucky. The lucky you send a limo for. Being able to see who's who is very useful for the books because they want to know who's action to respect for moving the line, booting, delaying or crippling ... and who to send the limo for.

The expression has value to the bookies. That doesn't mean it has value to us.

We don't get paid for BTCL. We get paid when we beat the spread. The point might seem trite, but the whole idea of BTCL, from the player's side, is trite. BTCL doesn't actually tell you anything. It's a gigantic "duh." If I say the way to make money in this game is to "beat the spread," would you get excited? Telling you to "just BTCL" isn't much different.

So let's look into who and why and how some players BTCL. Or maybe I should say, of what is BTCL a manifestation.

1) Line shopping helps you BTCL (I'm distinquishing line shopping from bet-timing). If the line is -7, and you, through shopping, find -7' or -6', you'll BTCL.

2) Bet-timing helps you BTCL (I'm distinquishing bet-timing and steamchasing). If you like a fave at -7, correctly anticipate that the line will rise and so bet early, you'll BTCL. I bet-time. I know I'm better off, for example, betting CFB totals early than late. This is broad bet-timing, and is simply a matter of comparing how I'd do betting at different times. Bet-timing is about getting the best price, and as such is an adjunct of line shopping.

3) Steamchasing helps you BTCL (I'm distinquishing steam chasing from syndicate membership). Steamchasing is BTCL, to some extent, at it's purest and oldest. To say it's BTCL, in fact, is redundant. Steamchasing, really, is best understood as syndicate betting without being a part of the syndicate (and by syndicate I really mean sharp action generally; RAS isn't a syndicate, but from a book's POV may as well be).

Steamchasers distinquish between sharp action and general market drift, and also distinquish between acting in time and having missed the boat. Further, they "steam anticipate," which is a great skill (related to bet-timing, I suppose, but obv I'm distinquishing it here).

Most of the guys who talk about BTCL are steamchasers, and for them it's a useful shorthand for "doing their job right." It's more useful short-term than "making money" because there's much less variance. There's still a large measure of line shopping involved, but line shopping isn't the same thing as steamchasing. Steamchasing is really about the ability to recognize what line movements mean, and even anticipate them, and that's a skill separate from, once having labelled the movement, acting on it. I don't personally steamchase (we can't all do everything), but I recognize the potential for profits with it, when done right.

A really good steamchaser (market reader) can sometimes see a line move a half point at only one book, so that the broad market line remains the same, recognize that half point indicates a sharp side, and bet, +EV, even though he won't BTCL.

But even steamchasers don't make money BTCL. They make money by being able to read market moves and react quickly, or even anticipate them. You can use the semantics of BTCL here, I admit, but "BTCL" doesn't tell you how to make the money. BTCL reflects the skill, but is not the skill. Steamschasers don't make money with BTCL. They make money making +EV bets, and where the line settles afterwards doesn't change the EV of those bets, and it's ridiculous to think it does.

4) Syndicate membership helps you BTCL. Originators (and syndicate members are essentially that, in that they are, even if mere beards, extensions of the originator), are the ones who most deeply BTCL, and are most problematic for books (though books are probably generally unable and uncaring about distinquishing between steamchasers and syndicate players).

This is where the advice "just BTCL" gets really pointless. It becomes tantamount to saying "get involved with an originator's syndicate." Other than RAS, I know of no way to just "join" a syndicate (and yes, obv, I'm labelling RAS a kind of a syndicate; it's a publicly available one).

But even originators don't make money by BTCL. They make money by handicapping. The difference is huge, so I'm going to say it again: they BTCL the most of anyone, but they don't make money from BTCL. BTCL is the effect of the betting that follows their profitable handicapping. Essentially, they BTCL the most, and potentially earn the most, but don't earn anything from BTCL. They earn from handicapping +EV bets.

You can make money line shopping. You can make money anticipating market direction. You can make money interpreting line movements. You can make money handicapping. What you can not do is make money with BTCL because it is merely a reflection of handicapping or betting skills, not the cause.

Here are some more points:

If your focus is on BTCL, you aren't going to come up with anything new, because if it's new, it won't be in the CL. I BTCL big time in CBB totals. But that's coincidental (well, maybe not, because I don't really know how everyone else is approaching things; but it isn't deliberate, anyway, in that I don't know how everyone else is approaching things). IOW, if you focus on BTCL, you'll never be the originators who are the biggest cause of BTCL, because they cause BTCL by handicapping. (and using BTCL to reverse engineer originator's action still isn't the same thing as making money with BTCL; it's making money with reverse engineering).

Some of the real BTCL nuts will say that no one wins handicapping. The illogic of that is mindboggling, considering that, if no one is outhandicapping the openers, how can the CL be more efficient than the opener? They misunderstand their own results. They'll cite stats about how they lose when they don't BTCL and win when they do, failing to understand that, no, what's really happening is they're losing when handicapping poorly and winning when handicapping well, and BTCL is somewhat correlated with that.

If your focus is on BTCL, you'll fall if the cause of BTCL falls. Suppose you were on Dr Bob after his heyday. If you got his picks early and bet well, you'd BTCL big time. And get slaughtered at some point. You'd have been BTCL and slaughtered. BTCL is absolutely not in and of itself blindly a predictor of profits. Yes, the market is long run more efficient at close than at start, GENERALLY, but that isn't the same thing as being more efficient for any given game SPECIFICALLY. If you've been in this game a while you've seen a lot of hot hands come and go. Things change. I've known two steamchasers who won for years and then got wiped. I kind of think it's always a matter of time, blustering posters here notwithstanding. Think "Black Swan."

If your focus is on BTCL, you'll miss all the +EV bets that come about at the CL itself. This truth alone reveals the illogic of BTCL. If a line is moving the right way, I wait until it settles. Sometimes, then, I'm better off, not BTCL, but tying it. If you agree that's possible, then you are admitting to a logical flaw in the theory of BTCL.

If the Lakers are home and fave of -8, and I like the dog, I know there's a good chance it'll creep up to -8' by game time. So I'll wait. I'm deliberately not trying to BTCL. Does that mean I can't possibly have value? I think it's ridiculous to live in fear of opposing line movements. If I like the dog when the Lakers are faves, if it creeps up a half point close to game time, that gain of a half point is pure value, imo. IOW, I'm saying the CL was less efficient than a previous line.

If your focus is on BTCL, you'll become what poker players call "results oriented." That's a terribly misleading phrase, in that it means, actually, that you should not be results oriented (short-term results), you should be EV oriented. Here's my point, re BTCL: if your focus in on BTCL, if you bet and the line moves such that you should have bet later, you'll think you failed. And if it moves your way, you'll feel like a winner. But that's being results oriented. Inasmuch as line movements can be, from your POV, random, what they do after you bet can not intelligently be seen as the measure of your bet's wisdom. It's like judging the wisdom of your flush draws by whether they hit, not whether they were priced right.

Let's say, by my metric, I like a CBB total <140. Let's say I know RAS is going to release a play on that total, but I don't know which way he'll go. Then, his release, which will move the line, is just a random variable, from my POV, and unrelated to my bet's wisdom. It might be correlated to my bet's outcome, but that isn't the same thing as saying I was wrong to make the bet with the information at the time of my wager.

If you make a football bet, your bet should not be judged stupid if, later, your QB is announced out with an injury, nor judged brilliant if it's the opposing team's QB announced out. These are post-bet randomizations, and not at all related to your bet's wisdom.


Okay, enough. I did this a little quickly. I'm not getting paid to write it. It's also a little tl;dr. That's less okay, but, WTF. I not getting paid to edit it, either.

Cliffs: from the books POV, BTCL has value; from ours, no. Don't tell yourself to BTCL, tell yourself to line shop, bet-time, market-read and handicap.

Gee, Joe. A few conflicting points here. Which is it, Mr. Sharpie?

[MonkeyF0cker]

Okay, the sickness continues, I'm running so good it just isn't right, all time best weekend ever, ainec: 3-0 NHL bets Friday night, 2-0 CBB bets then; 14-1 CBB yesterday; 2-0 NFL bets today, and 5-0 in Leroy's NFL contest (to move into first place in that). Which is all, yes, wayyyy off topic, but at least if I'm ever a whiney bad-beat douche you all can pull this thread up and LOL.

This part was (by far!) my favorite though.

The "Look at me! I'm going to tell you why I'm sharp, and he's not." preface.

Brilliant.

[RickySteve]

Joseph, can you please tell me where you play poker?

[Wrecktangle]

Happy Holidays, Poker.

Glad you found the problem.

[subs]

Gratz Joe

interesting thread, thanks.


lol @ some of u guys... just can't help it, can u?

*shakes head and smiles*

[benjy]

I find it interesting too. Nice thread.

Fantastic run PJ, congats.