📊 Sports Betting Edge: The Bet Nobody Beats

There is a bet on the board every Sunday that a professional handicapper, working at the very top of his range, still loses money on. It is an NFL home team favored by anything from half a point to three. I bet it for years without knowing what it was costing me.

When I started handicapping in the early 1990s, conventional wisdom was that the NFL was the hardest league in America to beat. It drew the most attention, and it had the sharpest lines. I went straight at it anyway, and it has been the sport I have beaten most consistently ever since.

What took me far too long was learning to look at the board one bet at a time. Take that home team laying a point and a half. Now take an NFL game with a high total, and take the UNDER. Both are priced at −110. Both need me to be right a little more than half the time. For years, I treated them as the same proposition because, on the surface, they are.

They are not close. I keep a database of closing prices going back to 2003, and when I finally ran those two bets against each other, the gap between them was wider than the gap between winning and losing at this for a living.

📈 Every pick I make gets posted and scored. You can work through my full NFL archive, winners and losers, and check any of it.

🎯 What I was actually measuring

A bet like that has a name in my own work. I call it a spot: one particular kind of bet at one particular price, like a home favorite of a point and a half, or the UNDER on a high college total. Not a sport, not a league, but a single recognizable situation that comes around week after week.

“The NFL is hard to beat" is a claim about a league. Leagues do matter, and plenty of what I have found is genuinely about individual sports. A trap that shows up in both the NFL and college football does not appear in basketball at all. Small underdogs behave one way in football and the opposite way in college basketball.

But the range inside a single league is far wider than the league label suggests, and nobody told me that when I started.

“The NFL holds one of the softest bets I have ever measured and the single hardest one, on the same board, at the same price.”

Here is the part I did not appreciate until recently: I had been hunting these things by feel for thirty years without being able to see them.

I named my company as a play on the word "underdog," and in the early days I picked NFL underdogs and nothing else. That was not a philosophy. It was a soft spot, and at the time it was an enormous one, because underdogs were badly underrated and badly underbet, and the prices reflected it.

Then the market caught up. The edge the books had allowed for years was no more, and that is the part worth knowing, because it happens to every edge eventually, and I have never seen one that did not close. Adding other sports taught me the same lesson from a different direction, since what worked every Sunday in the NFL did nothing at all in basketball, and it cost me money to find that out. By the end, I had to spread out and mix bets to keep working.

Thirty years later, my own database says both of those things in plain numbers: underdogs behave differently by sport, and the blind NFL underdog edge is gone. I was right about the shape of the work and flying blind on its size.

📉 Everyone needs 52.38%. Almost nobody starts there.

At −110 you have to win 52.38% of your bets to break even. That number is fixed; it comes straight out of the price and is the same for every bet on this page.

Where you begin is another matter entirely.

Every line below is a real situation, priced at −110, and the return is what you would have made backing that side blind every single time it came up, with no handicapping involved at all. The last column is the one that matters: how many extra wins per hundred bets you would need, on top of what blind betting already hands you, just to reach break-even.

The bottom line asks you to be right about six more games in every hundred than a blind bettor. The top line asks for one more game in a thousand. Both tickets cost the same, and both pay the same.

Even the best line on that list is not free, and there is a reason. Across all seven sports in my database, sportsbooks hold a remarkably steady 3.61% on a two-sided market. That hold is not a fee bolted onto the transaction. It is the business, and it works out to about 2.38 wins per hundred that you owe before anybody has an opinion about a football game.

So the college football UNDER on a high total is a market that has quietly handed back almost the entire margin. The NFL near-pickem home favorite is the ordinary margin plus another 3.7 wins per hundred of the number being set against you.

Ask yourself who is standing at that second window. A casual bettor sees his home team laying a point and a half against somebody they should handle, and it looks like the easiest ticket on the wall. He is not doing anything unreasonable. He is doing what almost everybody does, and the price he gets reflects how many people arrive wanting that exact bet.

🧮 What a real edge looks like

None of that means much without a sense of what good handicapping is actually worth, so here is mine, and it is a judgment I will put my name on.

A strong professional handicapper beats a blind bettor by two to five wins per hundred. Against the spread, that is a 52% to 55% bettor. Not 60. Not 58. The top of the professional range is 55%, and most people who believe they are above it have not counted carefully.

Each extra win per hundred is worth 1.91% in return at −110. So a whole career of skill looks like this:

Read the top row twice. Two extra wins per hundred is real skill, the kind most bettors never develop, and on an ordinary bet it still loses money. The 3.61% the house holds is bigger than the edge, so the house takes it. Sportsbooks do not need anybody to lose badly. They need most people to be a little worse than the margin, and most people are.

🏈 The bet nobody beats

Now hold that two-to-five range against the table above.

Take that NFL home team favored by anything from half a point to three. Backing it blind wins 46.32% of the time, and you need 52.38%. That is six more wins per hundred you have to find just to get to even, and the professional range tops out at five.

So put the best handicapper that range describes, the 55% bettor, into that exact bet. He wins 51.32%, and he returns −2.03%.

He is not grinding out a thin profit. He is losing money.

That does not mean the bet cannot be beaten. It means the price sits far enough against the bettor that ordinary professional skill will not get you there, and anybody genuinely beating it is doing something well past what two to five wins per hundred describes.

There is an obvious next thought here, and you have probably already had it. If that side is the worst bet on the board, take the other one. Back small road underdogs every week and print money.

The arithmetic behind that instinct is sound. In a two-sided market, somebody has to be on the good end of a bad price, and the road side of those games does come back better. It is not clean enough to bet. The road numbers in my data span a range wide enough to include losing, which means I can honestly say that home teams are systematically overpriced and the road side is roughly fair. Road is where the bleeding stops. It is not where the profit starts, and anyone selling you the second version is reading one number and ignoring the interval around it. I am running the full home-versus-road picture across five sports in a piece next month, and it is the strongest thing in this entire body of work.

🏀 College basketball and the NBA are not the same market

The same problem arises between two leagues that share the same sport.

In college basketball, backing a double-digit underdog costs you almost nothing, and you need 1.1 extra wins per hundred to break even. In the NBA, backing a double-digit favorite costs 4.6. Those are opposite ends of the price board in two leagues people file under one word, and that word does not tell you which one you are standing in.

College basketball is the largest sport in my database at roughly 83,750 games, which is why I trust that number more than most. It is also where a lot of my own work goes, and the soft end of that range is exactly where my college basketball picks live.

💵 The moneyline is a steeper hill entirely

Everything above is priced at −110, where a return converts cleanly into a win rate. Moneyline prices do not work that way, so I keep them separate and never mix the two.

Backing heavy moneyline favorites at −300 or shorter returned −1.58% blind across 59,323 examples. Backing moneyline longshots at +300 or longer returned −19.73% across 48,001 examples. A longshot has to produce roughly twenty times the edge a heavy favorite does just to climb back to even.

College basketball has the widest version of that gradient anywhere in my data. Its heavy moneyline chalk returned −0.79% blind across 38,175 examples, the softest moneyline price I have measured in any sport. Its moneyline longshots returned −25.04%. One sport, and the distance between its cheapest and its highest price is wider than the distance between most sports.

People want to risk a little to win a lot. That is the whole appeal of a longshot, and it is why far more money walks into that window than into the chalk window, and why the price there keeps getting worse while the chalk price stays close to fair. I measured the prices. The crowd standing at each window is what I have watched for thirty years.

It is also why my Parlay Club builds its tickets around heavy favorites rather than dogs. A parlay is not free money, and stacking legs multiplies the house margin, so nobody should read that as a shortcut. But every leg you add starts somewhere on that price scale, and the difference between starting at −0.79% and starting at −25.04% compounds across a ticket in a way most people never stop to look at.

One number in my moneyline data deserves a warning, and it is a personal one. NFL small underdogs priced at +110 to +150 returned +1.37%, which looks like a blind profit sitting out in the open. That is the window I built a company on, and it is closed. There are not enough of those games in my database to separate that losing 4.06% bet from the 6.52% win, and the truth lies somewhere in that range. When the range covers both losing and winning, you do not have a finding. You have a number that has not settled, and I will not put a bet behind it until it does.

🎯 Where this leaves you

Finding weak spots is the job. It was the job when I was picking nothing but NFL underdogs, and it is the job now, and the only thing that changed is that I can finally see the board I was feeling my way across.

So do not read that table as an instruction to abandon what you know and go bet college football games UNDER.

Where your interest lies and what you are actually good at still matter enormously. Half a win per hundred of headwind in a market you do not understand is a worse place to be than six wins per hundred in one you have studied since the early nineties, because in the second one you might genuinely have the six. That is precisely what I did with the NFL, and I would not undo it.

What the map changes is what you do inside the sport you already know. Every league on this list has cheap situations and expensive ones sitting side by side on the same board, and working out which is which costs you nothing.

It is also work that never finishes. Every number on that table is a snapshot of a market that is still moving, and the soft spot I built a company on took years to disappear without ever announcing it was going away.

If you are not going to do any of that, there is still a free version. Stop laying a point and a half in near-pickem NFL games, which is the bet this whole article is named after and the one no professional beats. Stop backing double-digit NBA favorites. Look harder at the UNDER on high college totals. That alone moves you from six wins per hundred underwater to under one, with no skill acquired and nothing new learned.

I have documented more than 89,000 official picks since 2001, winners and losers, every one posted where anybody can check it. You can get today's free picks, and if you want my season, the NFL packages are here.

Every figure here comes from my own database of closing prices across 405,822 team-games in seven sports, 2003 through 2026. Each return quoted is what a blind bet in that situation produced, not what any pick of mine produced.