(Please scroll down for target betting updates for Saturday, July 31, 2010)
Pundits who oppose the entire concept of a betting strategy with long-term profitability are adamant that past and present patterns of wins and losses have almost nothing to tell us about the future.
It therefore follows that any strategy that relies on what has gone before (future data being unavailable to us!) is certain to fail in the long run.
The nay-sayers - the phrase nattering nabobs of negativity comes to mind as the sole legacy of a disgraced Vice-president - concede that the house edge is a reliable element of gambling that applies to upcoming games as much as to existing data.
But they argue that because the short-term effects of negative expectation are random and unpredictable, dependence on the repetition of prior patterns or trends is non-starter.
To put it bluntly, You Can't Win.
In fact, the only beneficiaries of the YCW conventional wisdom are bookies and casinos. And luckily for them, the vast majority of punters subscribe to the point of view that has been spoon-fed to successive generations since gambling began.
That leaves just a handful of us to challenge the status quo by first discovering, and then proving, ways to ensure that the average value of all winning bets exceeds the average value of all losing bets by a percentage sufficient to overcome to house edge.
Such proof, of course, is always dismissed as "anecdotal" because it depends on past outcomes that can never be repeated in the same pattern that resulted in a long-term win.
You just got lucky this time, say the experts, and next time or the time after that, you are sure to crash and burn.
I just wrapped up a preliminary analysis of a betting method sent my way by my old friend and collaborator #@&?#@@#$$, aka "Peter Punter," and the news is not good.
The strategy bets fixed amounts in four categories that are differing percentages of the available bankroll, and the focus is on run-line wagers with a few money-line bets tossed in.
One problem is that average odds are high (+141 for today, +148 yesterday, for example) and over time it is safe to predict that longer odds win less often than shorter ones!
The strategy's greatest weakness (as I told Pete before I started the analysis) is that in a sustained downturn, the bankroll diminishes, meaning that bet values that are tied to it will also shrink as resources evaporate.
So far this month, the system has had four winning days and 12 losing days, and has lost more than half of the $5,000 bankroll I put into it on July 5.
Bet values that get smaller with each losing day mean that even when the sustained losing streak lets up, as it invariably will, at least for a while, recovery becomes less and less likely.
Pete was not suggesting that the strategy had long-term legs: He just wanted me to take a look at it, and in passing sent me a link to an interesting news item that suggests that InvestaPick and other sports-related "funds" will soon have some powerful competition.
A European consortium is so confident that sports books can be consistently beaten that it has launched a plan to raise more than $65,000,000 in $125,000 buy-ins from investors who have reportedly already kicked in more than $20,000,000 in seed money.
Pete and I first started talking in August of 2009 about the viability of such a venture, and the idea has been the focus of my research ever since.
According to the gambling experts, remember, there's no sense in studying the past for clues to the future, even though scientists, economists, mathematicians and other very smart people are doing it somewhere in the world every second of every day.
Sports "trading" has been around for a while, but it will take the reported Euro-venture or something on a similar scale to make it respectable, probably.
Pete's request prompted me to take a fresh look at baseball statistics, broadening my attention beyond the underdogs that I have been concentrating on since last November 1 to take in run-line bets on favorites as well.
One thing I have learned (and believe me, there are many!) is that bookies know a whole lot more than anyone else about who's likely to win or lose a given game in any sport.
Duhhhhhhh...?
Not so much, because there are thousands of handicappers in business to sell "inside" information to tens of thousands of eager punters, and countless thousands more bettors who spend hours each day poring over historical data in search of their own special edge.
In my 7-dog trial, I have limited bets to a fairly wide range (+100 to +180), so I took all games from the 2009 MLB season and about half as many more from the current season, and queried the database for results in that range.
I should add that I learned just a few weeks in to the 7DT that the range was a tad wide, but I promised myself and my readers that I would not change the rules in mid-trial.
What I got from 2009 was a 44.15% dog win rate (DWR) at the 7DT range, with a 48.36% win rate for run-line bets in the same odds range on favorites (FRL) that did not clash with the dog picks for the same day. The overall 2009 DWR was 43.2%.
So far this season, the equivalent numbers are 41.02%, 41.37% and 42.2%.
Narrowing the range gets me 44.16% and 47.86% for 2010 so far, with the DWR unaffected at 42.2%, and 45.38% and 43.99% for 2009 (DWR 43.2%).
Anecdotal? Irrelevant?
To some, for sure.
But bookies are just like you and me in relying on the past to predict the future to a helpful degree.
And because their resources and knowledge are far greater than mine, I am happy to rely on bookies in turn.
The strategy that Pete introduced me to a couple of weeks back depends heavily, as I said before, on long shots that fall into the "in your dreams" category. And that just does not make good betting sense.
If the book is willing to give you +180 on a run-line bet (or a bet of any other stripe, come to that) you'd better hope for divine intervention.
Instead, look for a middle ground that recognizes that too narrow a range will not win enough to overcome the house edge, and that going too wide will greatly increase the overall number of losers, increasing the house edge and making your goal ever harder to achieve.
Assuming that you never place a bet at less than even money, just keep in mind that a DWR of 45% would require average winning payouts at (55/45)*100 = +123 at a fixed or flat betting rate to overcome negative expectation.
At the above-quoted WR of 47.86%, the flat-bet payback target would drop to +109.
And, of course, target betting is about as far from flat betting as you can get.
Negative expectation - the house advantage - is a fact of life and a pain in the ass to boot.
But it is NOT unbeatable!
Here's some more anecdotal information (I'm still playing one-thumb blackjack on my iPod, usually while sitting on the can, and I'm hoping to get back to the half-mill in funny-munny profits that I achieved before my last gizmo was stolen!):
Stats like those you see here, with a house edge of more than eight times the expectation for single-deck blackjack, make a crushing long-term loss inevitable, according to all those "experts" I refer to deferentially from time to time.
They also indicate either that I am a very bad basic strategy player with serious death-wish (I'm not), or that something is still seriously out of whack with MobilityWare's iPod blackjack app (it is).
I shoulda lost a bundle, but instead...
The BJ app allows bets from $5 to $5,000, which is a spread I can live with.
It starts a new player off with $1,000 in the bank, then allows markers of $500 each time the cash runs out.
Given the house edge you see above, there's no way in hell I could hope to "win" the amount I am ahead by busting out, and then playing 'til my thumb caught fire, waiting for 23 successive wins!!! to boost my bankroll into six figures.
Do the math, folks: +$500, +$1,000, +$2,000, +$4,000, +($5,000 x 19) is what it would take to cheat the game and achieve an overall win like the one you see above.
Then again, another way to go would be...target betting!
Here are 7DT bets for today, augmented by slim pickings to include some FRL bets as indicated:
1005 951 St. Louis +120 ML 12 (betting closed, game going very badly for my guys as I type this!)
1310 954 Philly Phillies +140 ML 8.5
1610 961 Wash Nats +115 ML 10
1605 972 Baltimore +115 ML 9
1605 971 Minnesota +125 RL (9)
1605 975 Toronto +145 RL (8.5)
1610 963 Atlanta +150 RL (8.5)
Next time, I will post anecdotal data derived from running target betting rules against the 2009 and 2010 bets.
The bets themselves, and their order, are completely out of my control, and are in essence selected by the bookies, since I apply a relatively tight odds range and could not care less which teams I'm backing or who they have to beat.
The more that emotion creeps into the business of gambling, whether it's through subjective selection of bets or that common enemy called greed, the more likely it is that the house will pocket your bankroll.
And who wants that?
Sunday, July 25 at 10:20am
I am racing against the clock, as I am most Sunday mornings, so here are dog picks for today, all of them money-line bets...
1010 903 Florida Marlins +120 8.5
1035 906 Pittsburgh Pirates +140 8.5
1110 909 Washington Nats +120 9.5
1310 914 Arizona D'backs +130 8.5
1310 915 NY Mets +140 7
1005 920 Cleveland Indians +135 9
1035 924 Baltimore Orioles +110 9.5
Placing run-line bets on favorites had me breaking new ground yesterday and making a couple of dumb mistakes. I'll get the hang of it eventually, I'm sure.
Naturally, it would be odd if I did not keep touting target betting as the only sure way to recover from a prolonged downturn, but I have done exhaustive research into this challenge, and no other approach comes close.
Betting a fixed percentage of a shrinking bankroll, for example, is a guaranteed disaster, but that does not stop several online system salesmen from promoting it as the silver bullet.
More about that later!
Tuesday, July 27 at 8:35am
Today's 7DT picks:
1605 903 Atlanta Braves +120 6.5
1610 906 NY Mets +130 7
1710 909 Cincinnati Reds +115 7.5
1740 911 Pittsburgh Pirates +165 9.5
1710 926 KC Royals +115 9
1905 929 Boston Red Sox +115 8
And because I was short a 7th "dog" for today, a run-line pick...
1705 907 Chicago Cubs -1.5 +150 7
Wednesday, July 28 at 12:10pm
A couple of early games slipped by me today (the bets were placed at the crack of dawn as usual, and they should have been posted well before now).
Anyone who imagines that there's some advantage to delaying bets until an hour into a game obviously has not been paying attention!
The recent article about the new European sports book hedge fund that is in the pre-launch phase said the idea will be to bet and cash out at different times during hundreds of games in play all around the world, much as brokers in stocks and foreign exchange buy and sell continuously.
But that's all way too headache-y for me! I don't even pay attention to team names beyond what's needed to place a bet, and who's up, who's down, who's in and who's out is irrelevant, as far as I am concerned.
All I care about is numbers: If the book puts a prospect on the board at odds I can live with, I'm in. Too much personal involvement is death!
Yesterday was the first time in weeks that we managed five right 7DT picks out of seven. Hallelujah!
Today's 7DT picks:
1105 MLB 952 Houston Astros 115 8
1110 954 Milwaukee Brewers 100 8.5
1605 956 Washington Nats 115 7.5
1605 957 Florida Marlins 115 8
1610 961 St. Louis Cardinals 105 6.5
1905 966 San Diego Padres 190 6.5
1110 968 KC Royals 105 9.5
Thursday, July 29 at 10:20am
These crack-of-dawn games really are a pain!
I'm late with a couple of today's picks, but since when I checked a moment ago, the Cards were getting thrashed 3 to zip, I don't think anyone can accuse me of sleight of hand!
Yesterday gave the w-i-d-e spread column in the 7-dog trial a huge boost, bumping it back to 98% of its best win to date.
If it can set a new high between now and the last day of the month, I think I will end the experiment after a neat-and-tidy 9-month gestation.
If not...well, I will probably pull the plug anyway, and be thankful that along the way, the need for a very wide betting spread was once again resoundingly confirmed.
I started out tracking at spreads of 1 to 5 and 1 to 50, as regular readers will know, and neither of those were able to survive a very weak start to the season for MLB underdogs.
The super-wide spread of 1 to 1,000 (about a fifth of what's recommended for casino house games) soared until July 5, then wobbled dramatically for a while before yesterday's lifeline almost pulled it out of the hole.
Selections are being presented somewhat differently today:
The information is all there, as you can see, with the 7DT selections on the right.
On the left is today's betting lineup for a new tracking sheet I started when my pal "Peter Punter" alerted me to a strategy he has been trying since the beginning of July.
More detail later, but for now, know that Pete's new method has so far achieved part of target betting's primary objective, which is to ensure that the average overall value of all winning bets (AWV) exceeds the average value of all losses (ALV) by a substantial percentage.
The method's $W/$L number to date is 123%.
Unfortunately, the L/W number (117/74) since July 5 stands at 158%, resulting in the method losing more than 55% of its opening bankroll in less than a month.
Pete did me a big favor, though: a prolonged slump for underdogs has made the 2010 MLB season tough going for target betting, and run-line wagers on favorites are looking more and more like a promising means of cushioning the fall until "the math" inevitably bounces back in the next few weeks.
Friday, July 30 at 3:30pm
The latest wave of visiting family headed home early Thursday, and I have been busy since then collecting yet more data about the 2009 and 2010 baseball seasons for presentation here in what I hope will be a comprehensive and comprehensible format.
These things always take forever, and it will probably be mid-week before I'm ready to rock the world with verifiable proof that the odds in the book's favor can be vanquished consistently, safely and effectively.
What I don't want is a set of numbers that look anything like the 2009 MLB summary offered as "proof" of the strategy that "Peter Punter" asked me to take a look at.
It shows breathtaking profits for last season, then wimps out on proving that the same method does as well or better for the current half-season.
It is a critical axiom for all system "debunkers" that odds-busting formulas are always developed by tweaking various tricks against existing data until an imaginary big win is seen, with no hope of repeating the performance against future random numbers.
Verifiable baseball results from more than 4,000 games in 10+ months have proved yet again that target betting is a viable, unique way to consistently beat odds that are far worse than anything found at casino table games.
The reason that the strategy stays so successfully in the game in the long run is that while wins are harder to come by, paybacks are almost invariably much better than even money.
It works just fine with money-line bets, but the run-line option, always backing favorites as long as the price is right, adds an interesting new dimension that I will talk about at length when I finally get all my ducks in a row.
Mix the two non-conflicting approaches by opting for RL bets only when a given game does not offer an attractive ML payback, and many of the slumps that occur when underdogs suddenly can't get arrested are made dramatically less costly.
Below are today's picks, in a summary that shows how the 7DT's fortunes have been greatly improved this week, followed by sneak peaks at ML+RL data that I will explain later.
The way target betting is applied in a sports-book setting is much as I predicted it should be when Peter Punter and I first talked about the change of focus almost a year ago.
Multiple bets are placed every day, and their values have no relationship to one another.
Instead, bets are tracked in series, so that the result of bet #1 on the first day determines the value of bet #1 next day, #2 is linked to #2 day after day, and so on.
The MLB data sets for both the 2009 and (part) 2010 seasons require as may as 16 bets a day, and I admit that a month ago I would have said that that much skin in that many games was madness.
I have had to eat some crow over that, although as much in fear of unmanageable paperwork as anything else, I have run tests that reduce the workload by imposing variable odds ranges.
In the 7-dog trial, my range was from +100 to +180.
On the one hand, I was frustrated by the need to stick to seven bets a day, which required me to apply a certain amount of subjective judgment to the task of making selections.
On the other, I got a big kick out of seeing wins that paid as much as $1.80 on the dollar - good news in the era of 6-5 payouts at blackjack!
I have had to build a whole new spreadsheet template to handle up to 16 bets a day in real time, and I decided to do it in the newest version of Excel rather than sticking with my old faithful (QuattroPro in a version so old that it might even pre-date the advent of computers!).
Excel is a marvel, for sure. I have used it to a limited extent for years, but now my demands on it are such that I am having to learn a whole new language.
Getting everything right will take me a while.
One interesting development: target betting is (just!) ahead in the ML+RL trial after six betting days, while Pete's "bookie-busting" method (the salesman's claim, not Pete's) is in the middle of a long slide.
Slumps happen to even the smartest of bettors, but I cannot see how a strategy that values bets at fixed percentages of a diminishing bankroll can ever get out of the hole.
Saturday, July 31 at 7:15am
I have plenty to say about the updates below, but no time to say it (today's a play day for me and the missus!).
So, here are today's picks, in a new format, plus the last 7DT wrap-up:
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. One more piece of friendly advice: If you are inclined to use target betting with real money against online "casinos" such as Bodog, spend a few minutes and save a lot of money by reading this._
No comments:
Post a Comment