Wednesday, September 30, 2009

There's nothing new under the sun, and casino eyes in the sky have seen it all before. Or have they?

_
I suppose I could say I feel wracked with guilt for ignoring my favorite blog for more than a month.

But since I have not been swamped with anxious e-mails from concerned readers wondering if I'd fallen off a cliff somewhere, I am going to assume that no one noticed!

In any event, I have a passing good excuse. My friend in China (I'll call him Peter Punter to protect his privacy and good name) persuaded me to take a closer look at his chances of using mathematics to make money betting on baseball, football, basketball, hockey and all those other games I don't understand and don't want to.

As I feared, the project had me hooked within a few hours, and I have spent most of the last 33 days building up a database of prior games, along with odds and other gambling-related stats, and running it all through the wringer.

I told Pete at the outset that I could not see how money could be made betting short odds and trying to overcome the the river of red ink that's bound to flow when you win more often than you lose, but lose more when you lose than you win when you win.

Target betting, as regular visitors will know by now, takes a negative expectation ranging from around 1.0% to 5.0% or more, and turns red to black by winning more when it wins than it loses when it loses.

It does very well at baccarat, so long as Banker bets are excluded, but the moment the so-called "5 percent" commission kicks in, the 95 cents returned for every $1 bet against Player results in all manner of monetary mayhem.

Pete's idea was that the "less than 50-50" return that applies to most smart bets made in a sports book can be spun into gold by heeding the advice of the best of the hundreds of tipsters who make their living predicting the outcomes of ball games.

He sent me stats that demonstrated that long-established experts (called "cappers" or "handicappers") routinely claim win rates of 55% or more - some boast well over 60% - and suggested that a long-term positive expectation must be a potential basis for consistent profit.

Pete fervently believed that it would take me less than a week to come up with a definitive answer to his Big Question: Could target betting (or Turnaround) be successfully applied to sports betting?

He admits he's not a numbers guy, and I was constantly reminded of that as I slogged through one week, then two, three, four...and realized that I had taken on something more complex than I could have predicted (and I was not optimistic about the one-week estimate to begin with!).

Pete started his own sports book experiment with real money at the end of July, and for a while was doing pretty well.

Then the wheels fell off.

I am tempted to challenge the cappers' win rate claims by testing them under precise scrutiny, but what would be the point?

I'm taken back to my halcyon days at English racecourses, where shabbily-dressed and often malodorous tipsters would walk up and down among the crowd wearing what we knew as "sandwich boards" claiming all manner of insights into the best bets of the day.

Their "tips" were mostly sure things that a novice could figure out in five minutes of reading the day's card.

And at Warwick Racecourse in particular, I did a whole lot better paying attention to Len, a flat-capped veteran of at least 70 years of racing who worked the bar in the Press tent, and relied on me and other "young 'uns" to place his bets for him while he pumped ale.

I never went too far wrong backing Len's picks, and could not help but notice that most of the chalk boards I walked past on my way to the tote disagreed with him. He won a lot more often than they did, that much I remember.

I plan to give the blog a brief respite from blackjack and other card games, and will post regular bulletins as I press on with a hopefully unbiased and realistic assessment of Pete's hypothesis.

I can tell you right now that if you consistently back favorites, as the crappers almost invariably recommend, you can win more than 55% of all your bets, and still lose money.

My guess is that Pete was undone by the fact that he relied on a slew of cappers, at least a half dozen of them, and bounced back and forth with them, sometimes betting $50, other times risking more than $1,000 to win just a few hundred bucks.

I suggested early on that he should apply my suggested adaptation of the old Turnaround rules to multiple betting lines, perhaps grouping his bets by handicapper, or categorizing them according to start times and time zones.

Pete went a different route, figuring out his bets with a cancellation "system" that I looked at and quickly dumped way back in the 1980s.

My advice was that, in essence, the result of a game between two groups of overpaid professional athletes is no different than the outcome of a hand of cards, a roll of the dice, or a spin of the wheel.

The only sure way to win is to bet consistently and confidently, avoiding emotional responses and betting differently on a roll than during a prolonged recovery.

Pete had other ideas, and I doubt he now has either the stomach or the cash to go back to square one in the aftermath of losses of several grand, and do things the way I wanted them done in the first place.

Fair's fair, it was his money and his call. He is still a believer in cappers, but with all but one or two of them now showing "win rates" that at under 50% should actually be called "loss rates," I'm confident that I could do better.

Matter of fact, I have, just by doing what dear old flat-hatted Len did way back when and looking at the numbers, not the names of the horses (teams) or the colors on the jockeys' backs.

Len did not like short odds, but he liked long ones even less. He aimed for a middle ground, counting on his belief that the folks who set the odds did so on the basis of science and rationale, not on a whim.

I can tell you now that whatever Pete decides to do in the long run, I will be making regular sports book bets on my own account (much smaller than his, for sure!) and will do so without feeling the need to learn more about how the actual games are played.

I am well aware that down the years people far smarter than me have studied the pro sports pastimes that obsess a significant percentage of American adults, looking for clues to trends that they can exploit for profit.

The cappers may be the cleverest of us all. They don't have to risk their own money on the games, they just have to convince other people that they have special expertise that enables them to pick winners more often than a blind man with a sharp dart and today's schedules taped to a wall.

I have my doubts. But plenty of punters (Pete, for example) believe that if you pay for advice, it has to be better than whatever you can get for free.

Thursday, October 8:

A few days ago, my friend "Peter Punter" objected to my potted description of his sports book plan, but did not really explain why.

He did, however, make it clear that he is no more inclined to trust paid advice than the free alternative, which begs the question, Why pay for tips in the first place?

Pete has shelved his project after around $8,000 in losses, but has kindly sent me info about the bets that he "woulda" made if the funds for his brave foray into betting by remote control had not run out.

The great irony is that IF he had kept going just a few days longer, and IF he had let target betting dictate the value of his wagers, he "woulda" been out of the hole by now.

And IF he had followed the principles of target (or recovery) betting from the get-go, he would not have paddled so far into and eventually drowned in red ink in the first place.

IF is, of course, the most over-used word in the gambling glossary, usually followed by coulda, woulda or shoulda, and sometimes all three wistful reveries on how things might have gone differently IFONLY...

While I was working on Pete's short-lived project, I made regular visits to my local casinos for impromptu tutorials from sports book staff and players on a subject that I knew nothing about.

And naturally, I spent some time at the blackjack tables, breaking my long-standing rule that prevents me from playing without big money in my back pocket.

My rationale was that the research fee that Pete sent me was already "gambling money" when it dropped into my bank account, so I could convert it into chips with a clear conscience.

I more than doubled it before the house edge bit down, and pulled out with about a 90% return on my original investment by applying my informal rule that the bad guys should never be allowed to get all "their" money back.

The exercise confirmed just how smart Pete is in looking for an alternative to the bumpy and frustrating ride that almost always awaits anyone attempting to profit from casino table games.

Target betting is not the problem. Playing in a fishbowl is!

And then there's the atmosphere of a casino, and the temptations and distractions created by the very nature of the environment, along with other players who stand on A,3 against an 8 (true!) and dealers who behave as if your every win is being extracted from their own wallets at gunpoint.

Pete's right: dispassionate, numbers-ruled betting in a milieu in which the sky's the limit and no one is watching seems like the perfect, lucrative alternative to hours and hours spent in smoky casinos day after day.

When a bet reaches one bookie's limit, as it must once in a great while, there is plenty of time to spread the load across as many tickets as it takes to get it covered.

And while the long delay between outcomes seems frustrating at first, especially when compared to the blinding speed of heads-up blackjack, it actually promotes the discipline and consistency that can sometimes be left behind in casino play, even by a player who drinks only bottled water served at room temperature!

The baseball season ended last weekend, and the database I started for Pete now has just 737 contests in it, proof of nothing, perhaps, but containing more bets than most sportsbook customers will make in an entire year.

Analysis of the games played from August 12 through October 4 naturally confirms that at best, a 56% win rate for bets on favorites is only enough to break even.

On average, favorites paid $0.86 to $1, which dooms flat betting to long-term losses but can be profitable if basic target betting rules are applied, adapted to allow for multiple lines (or series) of wagers set up to allow multiple bets to be placed in a single daily trip to a sports book.

Pete's in China and has to place his bets online, coping not only with a 15-hour time difference but the fact that Internet bookies offer odds that are almost always less generous than those found in my neighborhood.

So he has to risk more to win less.

I used a mix of online and land-based odds in my database, keying in real-world numbers whenever I had a sportsbook printout to work from, and relying on Internet sources when I didn't.

The average "dog" (underdog) return from my sample of 737 baseball games was $1.24 to $1, which made it theoretically possible to sidestep a long-term loss with a pick accuracy rate (or win rate) of 47.6%.

I have not yet reached the point where I can confidently advocate a "dogs only" approach to sports betting, but when I applied a cautious set of selection criteria to my baseball sample, I ended up with a 9.85% return on action in spite of a win rate of 49.3% (or a loss rate of 50.7%!).

That's encouraging, to say the least.

Making the right "dog" choices brought back to my mind's eye a picture of old Len squinting through blue-tinted spectacles at the day's race card, looking for numbers that hit a middle ground between too much risk for too little return and a snowball-in-hell's chance of a win.

Len, of course, was able to add inside knowledge to the mix, and what I know about baseball, you could "stick up your nose and still have room for your finger," as he might have said.

Moving in to the hockey, basketball and pro-football seasons, my plan is to make up to five funny-money bets on each sport every day that the lineup permits, posting my picks right here in advance of game time.

I will split the selections into separate lines to determine wager values, and publish a progress report at least once a week, detailing the win or loss to date for each series.

I have not yet figured out how individual bets will slot into the various lines, but their value and placement will be posted here ahead of time so that readers can evaluate the process with untainted data.

From the work I have done so far, which has extended way beyond the few short hours that Pete commissioned, I don't see how a combination of target betting and sports outcomes can fail in the long run.

What I don't yet know is how huge bets will sometimes need to be to ensure EOS (the table games acronym for end of series).

Wherever sports betting is discussed online, the phrase "money management" is invariably part of the picture, and that makes absolute sense to me.

Imagine the sheer joy of making steady money at the expense of the casino industry without having to endure hours in badly-lit, smoky conditions in the company of bad-tempered dealers and ignorant players possessed of a death wish!

I should add right now that most dealers I know are the personification of charm, wit and friendliness, just as most players I encounter are not a hazard to my bottom line.

But the horrendous exceptions, however tiny a percentage they may represent, tend to stick in the mind!

For the record, I sent Pete a breakdown of all his bets, with target betting applied to one a day, simulating a circumstance in which he would have waited for an outcome, then used it to determine the value of his next bet.

Monday morning quarter-backing is always a whole lot easier than picking the right moves ahead of the game, but I believe there is real relevance to an unbiased analysis of a representative data set. It works in politics, so why not in sports betting.

If my pal Pete had bet systematically, one pick a day, and had refused to put his hard-earned money on any proposition paying less than even money, he would have done a whole lot better than falling eight grand or so into a deep, red hole.

That's a fact.

As it was, his cappers ended up with nine more right picks than wrong ones in the two months in which Pete bravely soldiered on.

Because of the crappy odds (and because Pete placed bets ranging from $20 to over $3,000 while ignoring target betting and relying on a selection process that was essentially random) more wins than losses did not save him from the slide.

In the 65-day trial period, permitting only one bet a day, there were only 21 picks that offered odds of even money or better.

Backing them "my way" (cue for a tune!) would have required an average bet value of $885, action of $22,100, and a risk/exposure of $3,890.

Pete's way involved $66,000 of action, an average bet of $297, and a final loss of $7,432 (-11.16%).

The target betting final win, betting only picks offering 1-to-1 or better, was $3,280 (+14.84%).

Applying a slightly more complicated selection process to the 737 baseball games between August 12 and October 4 "woulda" given me a flat-bet win of $2,380 from 215 "dog" bets, or 9.85% of total action of $24,190.

As I showed earlier, my "dog" WR was 49.3%, giving the bookies an edge of 1.4% that coincidentally is in the same ballpark (hah!) as the NE for craps, baccarat and blackjack.

There is, of course, no sense in putting money on teams that the experts (the real handicappers, not the flashy tipsters Pete relies upon) suggest have almost no chance of winning.

On the other hand, if you bet $3,000 on a "sure thing" that will pay the equivalent of 1-to-5 if it wins, you had better have a strong stomach and stay away from the TV while the game is on.

In spite of his protests to the contrary, Pete is not much interested in numbers, and he has that in common with most gamblers.

It's not a sin or a crime, and perhaps not even stupid, to think of betting as a thrill...an opportunity to beat the odds and be rewarded for it, or to think of losses as a fair price for a unique moment of excitement.

I have heard many times that my laborious reliance on gigantic data sets and the probabilities they reveal takes the "fun" out of gambling, and I cannot argue with that.

Pete's sports betting idea appealed to me from the start because I know how vulnerable I am to the pressures and distractions that upset and sometimes derail my fun-free process when I take on a table game in a casino.

I know a lot more about sports today than I did a couple of months ago, although even now what I know could probably be matched or exceeded by the dumbest of the regular punters I have met and chatted with in recent weeks.

I plan to wring some more dry, un-fun numbers from the baseball database, and eventually expand it.

I also intend to collect representative samples of data from NBA, NHL and NFL archives, to see if the little I have learned from nine weeks of baseball can be applied at least in part to other sports.

I am leery of college football, because at first glance it seems so chaotic and overwhelming, but I daresay I will take a closer look at it one day.

Watch this space!

Today's NHL lineup has a dozen matches, and "the numbers" recommend the following picks:

NEW YORK ISLANDERS/Ottawa Senators
PITTSBURGH PENGUINS/Philadelphia Flyers
CHICAGO BLACKHAWKS/Detroit Red Wings
ATLANTA THRASHERS/St. Louis Blues
COLORADO AVALANCHE/Nashville Predators

Sunday's (Oct 11) NFL lineup includes odds so short they're almost out of sight (1 to 12 for the Philly Eagles and the NY Giants against their separate opponents!) but a few more palatable options are suggested by the numbers...

WASHINGTON NATIONALS/Carolina Panthers
ATLANTA FALCONS/S.F. 49ers
DENVER BRONCOS/New England Patriots
HOUSTON TEXANS/Arizona Cardinals
TENNESSEE TITANS/Indiana Colts

It's not cool to back underdogs, probably, but two out of five wins from these two sets of uncool picks would come close to recovering losses from the other three, something that a selection of five favorites with three gone south could not hope to do.

In the MLB sample in my files, 106 winners out of 215 "cherry-picked" dogs delivered a 9.85% profit (win/action) even though less than half of the selections actually won; 45 winners out of 83 cherry favorites represented a 0.11% profit.

Those numbers may be a onetime fluke that has nothing to tell us about upcoming baseball games, or any other sport.

But maybe, just maybe, Peter Punter is onto something good...

An important reminder: The only person likely to make money out of this blog is you, Dear Reader. There's nothing to buy, ever, and your soul is safe (from me, at least). Test my ideas and use them or don't. It's up to you.
_

No comments: