QUESTION OF THE DAY: "More on Line Movement"

This is a follow-up to the "Question of the Day" posted on Feb. 23. I'll let "DC" take it form here and close with some thoughts of my own at the end:

 

 

Nolan:

Thanks so much for a very thoughtful and prolific piece in response to my question on line movement. I appreciate the generous amount of time and effort that went into it.

You've given me some insight on why it is that we see this instantaneous move in the line happening to all these offshore books at the same time. While I knew they all surely must be using a clearing house system similar to the one you described the Las Vegas sports books using, one thing wasn't obvious to me until you brought up about RAY. That is, the true ATB late line move in response to wagered money coming in (as opposed to any "adjustment") may not necessarily be as the result of a group, or syndicate, getting down all at once at a dozen different offshore books...it's very possible it could be in response to a RAY, a couple of RAY's, or a particular group or syndicate getting down at as few as only ONE of the books.

 

In the offshore world, it would seem there would be no beards as such and so when a large wager comes in from customer ID #1234, the book can instantly recognize who they are dealing with. If it's someone that's been consistently beating them for awhile on their NBA totals or on their MLB or whatever, a little concern might start to set in. After all, the offshores don't have the alternative revenue streams that the major Las Vegas casinos have. If the sports book is down, say, 20% in revenues for the month at Caesar's, there are many other revenue centers in the operation that can smooth out the dip. But, let a single-line business like an offshore book have a 20% drop...we'll they'll certainly feel it. (My numbers are, of course, purely hypothetical and used only for illustration.)

So assuming there's concern that a) things are becoming a bit unbalanced on that game or that b) we can almost hear the book manager say, "oh my gosh...what's up with our line, is it too soft on this game? If customer ID #1234 has put down a good chunk down on it, maybe there's something we don't know. How many other players know what customer ID #1234 is playing? Gee, he could even be on a web site and be telling the world what he's playing, just like that Nolan guy. Man..we better adjust". And with that this book makes the adjustment, word goes out electronically through the clearinghouse that Book A is making a decent size line change and perhaps why...and boom, all the others play follow the leader and change their line too within a couple of minutes. Now of course you and I can't see all the adjusting that is going on as it is all taking place on the clearing house computer (and I assume individually at the websites of the various books), but once these adjustments are made the line is probably "published" at wherever Don Best and other line services get it from. So when we see this ATB move, you've given me insight to think that it's not necessarily an instantaneous reaction to a flurry of large bets coming in at a dozen books all at once, but rather that it could very well be the reaction to the action that only one or two books have experienced.

 

I want to be sure to add a very large disclaimer here by pointing out that the above is purely conjecture (and I'm sure some would argue, fantasy as well) on my part as to what would motivate an offshore book to move it's line by so much and so late. Certain particular line moves could very well be, as you suggested, generated by a computer program after a pre-set "trigger point" has been reached. After all, a well-managed book should not be too concerned with the balance of a single game (within reason) as much as whether or not the action in total is reasonably balanced. On the other hand, the world is not a perfect place and not all the offshore's sportsbook managers have the years of experience and the corporate discipline that the manager at a Caesar's or Mirage would have. Then too, I can imagine just by being in the line of work of gaming, there might be a competitive edge to some book mangers that might make them likely to forget about the balancing the whole book principal. I can envision some getting determined that, by golly they're not going to get beat up on such-and-such game because of a few sharps. But as I said, I don't know. Never having worked in the industry, I can only speculate by what I read in the trade and gambling press.

Even though conjecture rules as to the why's and how's of the ATB move, the facts are that it does happen. It's something we can tangibly observe. Last night was a good example. UTEP was a 7 point favorite over San Jose St, when BAM...the little indicators on UTEP at Don Best's site all went black at the same time as the line instantly changed from -7 to -8 at virtually every book on the screen. It then quickly went to -8.5. UTEP ended up winning the game by 20 points. Even during a "good" period such as we've had this past month or so, these moves don't ALWAYS win. Afterall, I sure even the RAY's of the world don't win 100% of the time.

 

As to your comments cautioning about whether or not what's been observed contains real patterns as opposed to random events being perceived as patterns, all I can say is possibly...possibly. I am familiar with the theory of large numbers and believe I actually read some of von Neumann's work at one time, but that's been years ago. But I give you the point only to the extent where we kept up with the observations on a casual basis. Not all our observations were done casually, however. We actually did keep a record (I shouldn't say "we" as I don't want to take credit away from the guys in the group who actually did all the hard work of keeping track). There were certain parameters set as to what constituted a true move and what didn't and records were kept. Also, to a lesser degree, what was happening to our bankrolls was also a way of keeping a record. (But even in keeping records, I do understand that von Neumann's work would probably show that human's are still predisposed to perceiving patterns.)

But with this talk of patterns I feel I may have done a poor job in conveying what I meant about getting "mixed results" and such. What I was attempting to do was contrast the strength, degree, and regularity of this correlation with the two primary golden periods I mentioned. They still have been happening since that time frame, just not to the same degree (e.g., fewer games per week or night, etc.). I think too that the "diluting" of the offshore's over the past three years with the large increase in the number of players has, as we agree, made a difference.

But I also can speculate on several other scenarios as to why the moves are there and are positively correlated and then sometimes they are not. For instance, if (and I know it's still an "if") given the presumption that during a period of time in which there is a strong positive correlation with the ATB moves and a higher than usual percentage of winners the books are moving their line in response to the RAYs of the world, then might it not follow that there are certain periods of a season that might be deemed a better time in which to play than in others? Take baseball for example. A lot of handicappers consider the lines during the first half of the baseball season to be on the soft side and expect to find more games with positive EV's than they do in the latter half. (The same is true for other sports as well). They're soft because the linesmakers are in a position of having to make a prediction based on limited data. In other words, as you've pointed out, Vegas doesn't know it all.

 

But in the case of RAY it's a different story. He's got people who know the athletes, who know the front office, who know the trainers and he therefore stands to be in a better position to predict the outcome of a game because of it. Now, consider that as we move through the season, the line becomes a little sharper on a greater percentage of games as the linesmakers obtain more data on which to base their prediction..both in how the public behaves as well as the usual statistical data from the field. The huge edge that RAY initially had has shrunk a little. Maybe to the point, where as in the stock market, a true professional decides to take his profits and sit out for awhile, but then perhaps decides to jump in here and there when he occasionally spots a real bargain. This to me seems to be the more plausible explanation of why there are certain periods when the line moves become really hot and when they're not. But it's not the only scenario. I'm sure there are plenty of others.

 

Even having said all of the above, I'm afraid I must agree with you about your final point where you say "But, until you can qualify this with some specific parameters, the theory is bogus...". I feel you are unfortunately correct. Correct, that is, to the extent as it being a bogus theory on which to base a consistant methodology for handicapping games. I do feel, however, that it can serve as one more variable to throw into the mix when considering which games (games that have already been studied and 'capped, by the way) will make the cut as a play or not.

 

Thanks once again for the generosity of your time and thought. I've enjoyed our exchange and would hope others at MadJacks would comment as well.

 

-DC

 

NOLAN'S REPLY: Congratulations on an outstanding retort to my initial observations, DC. You have raised several excellent points which merit recognition and further discussion. I'm sure many of our readers will enjoy and learn from this exchange.

I would like to reiterate one of your key points and offer for consideration what I think might explain the "inconsistent" results of what you have observed about line movements.

The initial question was: Is there a positive correlation between against-the-spread (ATS) wins and across the board (ATB) line movements? We agree that this question cannot be answered in the theoretical sense. However, this does lead to some related issues which are very important. Consider the following: What I am about to theorize presumably would only take place at large, well-established offshore books -- with significant customer bases and histories. This would not apply to small or new offshore operations. As you suggested -- these books might keep close track of winning players and monitor their plays/results more closely than the average customer. This is in contrast to Las Vegas, where winners and losers are usually anonymous and cannot be tracked (i.e. Caesars has no idea if you walk up to the window and wager $3,000 on a game whether you are a tourist of a professional handicapper). If the book observes a phenomenon by which a very successful bettor, or groups of winning players, are betting a side or total beyond standard expectation, the book may trigger an automatic line adjustment. Off-shore books would have a decisive advantage over Las Vegas in this regard since they can see where "winners" are moving their money. Previously, we have agreed that it might be possible that the books are moving their lines (in unison) in reaction to the "RAYS," or sharp bettors with big bankrolls betting high-limits

In fact, I'd like to amend that point. After giving this subject more thought, it would not be necessary for the books to identify these players by name or subjectively try to judge their results and make adjustments. All they would need to do would be to examine the wagers of the most successful bettors and make their adjustments accordingly. Once again, Las Vegas cannot do this, which is one reason why lines at the Las Vegas books are always out of synch. This too -- might explain why the offshores are more uniform in their numbers (generally).

Let's define the upper-echelon as the top 10 percent of all (off-shore) sport bettors. Since sportsbooks can easily identify those players who are able to "beat the game" consistently, it naturally follows that they would want to minimize their losses on any single game that was bet by these savvy and sophisticated sports bettors. Agreed? What is the natural thing to do, then? Answer: MOVE THE LINE! That way, an influx of sucker money will presumably come in on the other side to balance out what could be a 40-60 losing proposition for the host sportsbook (in terms of basic probability that that winning players hits 60 percent winners).

It would be a rather simple task for a top programmer to set up a database which automatically triggers line movements (even ATB line movements, assuming the books share common information) based on a inordinate amount of wagering from the top ten percent (a hypothetical number, used merely for illustration). This could be triggered by a certain dollar amount, or a predetermined quantity of winner players betting any one side. Let's say that it's 3:00 pm EST and 80 percent of the winning players are on Kansas tonight. Or, let's say that the handle has reached more than a 2-1 proportion based on $50,000+ in action (I have no idea what the actual triggers are, I'm only speculating). It's possible that instead of tracking "Customer ID #1234" (per your suggestion) who is a known-winner, the books use a mathematical formula that simply calculates the handle and moves the line accordingly, not on the basis of ACTUAL MONEY BET, but by the winning players being given ADDITIONAL weight in the equation. Of course, this would be entirely consistent with notion that books move their lines anyway based on large action coming in on one side. It's also consistent with your theory that (at various times of the season) the sharpies know what they are doing and you can follow the lead by betting with the movements. Again, assuming there is some connection to what I am speculating (which you appear to be in agreement with) this would PROVE that a positive correlation exists between ATB line movements and ATS winners at various times. It is proven by the fact the books recognize this as fact, and are thus changing their lines so they won't get pounded with losses.

THE PROBLEM: The challenge then, is to determine the conditions where this is actually occurring versus the bogus random line movements that are a natural by-product or market forces (see point to come). Obviously, not every line move indicates a winner. Observers would have to decide how much line movement to look for, which books should be monitored, and ascribe some parameters for different sports and perhaps even times of the season (early season, mid-season, late-season, playoffs). There are other factors to consider, as well. Obviously, the work you and your team have done is an excellent start and perhaps may lead to the uncovering of what we all seek -- a inherent formula for picking winners. But there is still a long way to go.

WHY I'M STILL SKEPTICAL: My response to your theories would not be complete without an added dose of skepticism. Keep in mind the employees that work in these sports betting operations are "people." There are programmers, ticket-writers (by phone), accountants, and a host of other staff working at the sport books. Sure, they are loyal to the companies they work for. But, they want to make a buck as much as the rest of us. I presume that because of the nature of off-shore operations, this is a rather close-knit sub-culture, where internal operations and trade secrets do not leak out easily. Accordingly, since it is easy to just track the smart money and then place your bets accordingly (as an outsider), imagine the advantages the insiders have. If ATB line movements were an indication of above-average winners, wouldn't this have already been identified and practiced by the individuals working in offshore sports betting operations? And, if so -- might this actually be the "monkey wrench" in your system?

Let me explain further: Let's presume that you have identified patterns between ATB movements and ATS winners. My contention: If YOU have identified this correlation, so too have OTHER people. Perhaps these coat-tailers are already hijacking the odds by pounding a side heavily when they see large wagers being made, which then creates a FALSE INDICATION of the actual strength of the play. Example: A winning bettor makes a big play on Kansas. Programmer at XYZ Sportsbook sees the big play on his own screen and then calls his "associate" who then wagers big on Kansas at ABC Sportsbook. Then, people at ABC Sportsbook see that two large wagers have been placed in the last hour and they call two friends and so on -- and so on. The line moves a point, maybe two. Then, let's say three winning bettors come in on the other side later in the day. Then, other employees at other sportsbooks jump on the other side based on that so-called "insider" knowledge. The end result is a sham. It's damn near a Ponzi scheme. I suspect that some of this is actually occurring, making identification of these real ATB line moves very, very difficult to track. This only adds to the complexity of what we are dealing with and illustrates just how challenging your task would be to distinguish REAL smart money from line moves triggered solely by volume.

MY THEORY AS TO WHY YOU ARE GETTING INCONSISTENT RESULTS: This phenomenon (above) IS occurring right now in my opinion. Once the insiders start jumping on the bandwagon however, the results begin to diminish since what were thought to be wise-money moves are suddenly not always so. The coat-tailing effect begins to taper off and the percentages drop as the reliable indicators are now fuzzy. This is due to the insiders diluting the money pool and giving many false alarms to line watchers who then begin to see their own percentages drop. You and I, who are watching these movements from the States are completely at the end of the food chain, and are left with the scraps. By the end of the run, no one knows which lines are REAL smart money moves, as the correlation between ATB line moves and ATB winners seems to have vanished. Then, the coat-tailers take a break when they see the results taper-off. Now, for the real kicker: Once the insiders and observers see their percentages drop and start to lose interest, the trend actually starts to become stronger again! Since there are now less coat-tailers toying with the system and fouling up the process with secondary money. The effect ultimately, would be a "wave" -- that is, temporary success, then a gradual decline and ultimate failure. Then, it happens all over again. THIS WOULD EXPLAIN THE PHENOMENON YOU EXPERIENCED. You said, the correlation worked for awhile, but then tapered-off and was no longer valid after a certain period. Then, you observed stronger correllations again later. Might this be the answer we have been seeking? I'll let you decide.

I'll try to crystallize our points of agreement, as follows:

1. A wager of significant size placed by a known-sports betting aficionado does appear to affect early line "corrections" and perhaps even initiates later line movements.

2. Many off-shore books share betting information between themselves (this point is obvious, judging by correlating ATB movements) and perhaps are even tied into a database that triggers line movements based on predetermined conditions.

3. As the number of offshore accounts has increased exponentially, tracking these "smart money" wagers have become increasingly more difficult. The larger the number of customers a sportsbook has, the more "diluted" a large wager becomes -- hence making the task of tracking this phenomenon more of a challenge.

As to my theory about the seeming validity, but inconsistency between ATB line movement and ATS winners -- it is just that, a theory.