If you’re going to work that hard, you might as well be a productive member of society.
There was one in the 90s that had to do with video poker but it was complicated and hard to pull off and was during a promotion. You had to play perfect strategy and play very quickly because a certain number of hands had to be played in an hour to max out the promotion which gave you some free plays or something and you had be a member of the casino’s player club and be playing on their card. This was another thing where Sanford Wong sent out his fax. A friend of mine did well on it.
I’d say that’s a safe bet.
For a decade or so, there were state lotteries that occasionally had odds that favored the player - but they were still pretty slim odds, so it took a significant amount of seed money to guarantee a profit, and a lot of grunt work to get it all done (just the record-keeping for tax purposes was demanding). I expect sports betting for profit to be at least as onerous, with the ever-present danger of your edge going away (when the profitability of the lottery became public knowledge to people who weren’t stats-obsessed math-types, the rules were changed - in the case of sport betting, if enough people learn to watch what you do, your edge will go away)
There are different ways for the bookies to set up the bets, to account for the fact that one team really is more likely to win than the other. One way is what you’re talking about, the spread: The bookies say that the winning team will win by, say, 7.5 points, and then everyone bets one way or the other on that at (approximately) 50-50 odds. If they win by 8 or more points, then everyone who picked them wins, and if they lose, or win by 7 points or less, then the folks who picked the other team win. I don’t think there’s any way to guarantee money on this, no matter how much the bookies disagree.
The other way it can be set up, though, is that the bookies set odds of each team winning (just looking at the win or loss, without regard to the score). If you bet the favored team and win, you profit only a fraction of your wager, but if you bet the underdog and win, then you profit multiple times your wager. On these, if two bookies have sufficiently different odds, then you can bet on one team at one bookie and the other at the other, and be guaranteed to win one and lose one, but set it up so your winnings will always be greater than your losses.
That makes sense. I have only heard of middling in a spread situation because you cash out when the score lands in the middle of the two extremes.
I’m still not convinced of this. Can you explain it using an example?
If it’s a very close matchup, one agency might be paying 2.01 on Team A, another sees it a little differently and is 2.01 on Team B. A two dollar outlay (one dollar on each) returns $2.01 or a guaranteed one cent profit.
That is correct, a middle only applies to games with spread or margin bets. You bet a middle when you find an opportunity to offer a smaller start for the favorite than the start that you receive for the underdog. Say the 49ers are available at -2 but you can get the Chiefs elsewhere (or later) at +4. You will win both bets if the 49ers win by 3.
However, because you don’t get $2 for your $1 bet on either proposition you still lose something when your middle doesn’t pay off.
Correct, but that’s not how @Chronos described it:
Of course if one team is favored by one bookie and the other team is favored by the other bookie, a profit is guaranteed, albeit small.
If BetUrLife is offering India in the World Cup final at 1-2 (i.e. a $10 bet pays $5) and KangarooWager is offering Australia at 5-2 ($10 bet pays $25) then you can arbitrage.
Bet $100 on India at BetUrLife
Bet $44 on Australia at KangarooWager
Australia wins:
You lose $100 at BetUrLife
You win $110 at KangarooWager
India wins:
You win $50 on BetUrLife
You lose $44 on KangarooWager
That’s because between these two offers the combine implied probability of the two teams winning is 2/3 + 2/7 = 20/21.
But you will find that this kind of discrepancy (actually much smaller than this even) will only occur between bookies in different countries. Punters anywhere will find it hard to place these bets without significant transaction costs in one or more of these.
Thanks. This is a great explanation.
But you are also correct in stating that this will never happen unless it’s in different countries. And that would take too much of an effort for a small payoff.
I knew a guy who spent a lot of time at the horse track. When he felt he had “a system” he’d keep track of what his system would produce had he placed bets. Some systems looked promising, but in the long run he never had a way to turn a profit over the long haul.
I wouldn’t be surprised if some of the mythos about people who do seem to turn a consistent profit might just be because they haven’t failed yet.
Everything is on a bell curve. There will be people for whom the system appears to be working, even for a very long time, and it’s just that they’ve been lucky so far.
The Gambler Who Cracked the Horse-Racing Code.
This guy had a system and didn’t say anything until his winnings were close to $1 billion.
That’s entirely possible. If you only bet a handful of games per year, the ‘long run’ could be the rest of your life.
Compare the number of bets you make on sports with other profitable games like poker and blackjack. In blackjack you play anywhere from 40-100 bets per hour. In poker, a little less. And yet, winning blackjack players have gone on multi-month losing streaks.
There is a lot of randomness in sports. You can analyze a game perfectly and have an edge over everyone else and still lose because a player slipped on the turf or a field goal attempt hit an upright and bounced out instead of in, or because a player had a fight with a spouse that morning and was off his game.
I’m sure that someone has developed an AI algorithm for sport betting, using decades of data, including all sorts of ancillary data, (weather, time of year, days, time of day, spreads, spread movement, etc. etc.) And is probably doing pretty well at it.
You want to make money sports betting? Stay in school, concentrate on math and computer science; get a job with a sports book.
It’s a certainty that lots of people are trying it. Whether any have been successful yet is hard to say.
And if AI does become profitable, they’ll eventually cause the lines to become more accurate I would think.
The data required is also very very expensive.
Expensive enough to significantly erode any winnings unless you’re betting huge, which of course means occasionally losing huge.