"Best bets" articles are almost always wrong because they tell you which side to bet and almost never tell you the price at which that side stops being a bet. A pick without a price cannot be right or wrong in any useful way. The same team can be a good bet at -110 (1.91) and a bad bet at -140 (1.71). By the time most people read the article, the line has often moved, and nobody tells them whether the bet still holds.
These articles are also written to a deadline, for clicks, and every week needs its five best bets whether the market offers five good prices or none. That is a format built to sound sure, not to be correct.
None of this means the writers are bad at sports. It means the format answers the wrong question. A best bets article answers "who wins?" The question that makes money is "is this price too high or too low?", and that is the one I want to show you how to answer yourself.
What is wrong with "best bets" articles?
The problem is the format, not the people writing it. There are three parts to it.
- There is no fair price. The article says "take the favorite" or "the over is the play", but it does not say what the writer thinks the true probability is. Without that number you cannot tell whether -110 is a bargain or a mistake.
- The price is stale. Most articles are written early in the week and read days later. The number in the headline is often not the number you can get.
- The volume is fixed. A weekly column has to have picks in it. A real edge does not show up on a schedule. Most weeks, most games are priced about right, and the honest answer would be "nothing this week". No column can print that.
Put those together and you get confident opinions with no way to check them. When one wins, it looks smart. When one loses, it was "a bad beat". Neither result tells you anything about whether it was a good bet.
Why does the price matter more than the pick?
Because the same pick can be a winning bet or a losing bet depending only on the price you pay for it. Say your own number gives a team a 55% chance to win. That makes the fair price -122 (1.82). Here is what that same 55% team is worth at three different prices:
| Price you get | Decimal | Win rate you need to break even | Expected value at 55% |
|---|---|---|---|
| +105 | 2.05 | 48.78% | +12.75% |
| -110 | 1.91 | 52.38% | +5.00% |
| -140 | 1.71 | 58.33% | -5.71% |
Same team, same game, same opinion. At +105 you expect to make about 12.75 cents on every dollar you stake. At -140 you expect to lose about 5.71 cents. A best bets article that says "take this team" is giving you the team and leaving out every column in this table.
If you want to see this move in real time, the expected value lesson lets you drag the probability and the price and watch the answer change sides.
What happens to a best bet between Tuesday and Sunday?
The market keeps working on the game, and the price you read is usually not the price you get. Injury news comes out, sharp money comes in, and the line moves. If a column posts a side at -110 early in the week and it closes at -130, the readers who bet it late paid a worse price for the same team, and the writer's record still shows -110.
This is why I care more about the price at which a bet was taken than about whether it won. The test is closing line value: compare the price you got with the price when the market closed. If you keep beating the close, you are buying before the market agrees with you. On my NFL model's 2025/26 bets that number was +4.25%, measured bet by bet against Pinnacle's closing line. That is the figure that told me the bets were taken at good prices, not just that some of them won.
Almost no best bets column reports this. Many do not report a record at all. If you follow one, write down the price in the article, the price you got and the closing price, and after 50 bets you will know more about that column than its writer has told you.
Why are "best bets" parlays even worse?
Because a parlay multiplies the bookmaker's margin along with the payout. Some columns finish with a "best bets parlay" that strings three picks together. Here is the math on three coin flips priced at -110 each:
- A single coin flip at -110 (1.91) has an expected value of -4.55%.
- Three of them in a parlay pay +596 (6.96).
- The fair price for three coin flips is +700 (8.00), because the chance of hitting all three is 12.5%.
- The expected value of the parlay is -13.03%.
You go from losing about 4.5 cents per dollar to losing about 13 cents per dollar, just by joining the bets together. The parlay math lesson shows how fast this grows with each leg you add.
How do you check whether a best bet is actually a bet?
Turn the pick into a price and compare it with the market. It takes five steps.
- Write down the side and the price the article quotes.
- Check the price you can get right now. If it has moved against you, treat it as a new bet.
- Make your own probability for the game. A simple model works. So does removing the margin from a sharp market price, which I explain in devigging explained.
- Turn that probability into fair odds. The fair odds calculator does this in one step.
- Bet only if the price you can get beats your fair price by a clear gap. Otherwise pass.
How big should the gap be? It depends on the sport and on how good your number is. For my NBA model I used to want 5 to 7 points between my number and the line before I paid attention. NBA lines have become more efficient since then, so now I treat a gap of 3 points or more as a good signal. The point is that the rule exists before you look at the game, not after.
Are any "best bets" articles worth reading?
Yes, the ones that give you a number, not just a side. Look for three things:
- A fair price or a probability. "I make this 57%" is something you can check. "I love this spot" is not.
- A limit. Good writers say "playable to -125" or "bet it to +3". That tells you when the bet stops being a bet, which is the most useful line in any pick.
- A record graded at the posted price, ideally with closing line value next to it.
An article with all three is a second opinion you can test against your own number. An article with none of them is entertainment. Both are fine, as long as you know which one you are reading. If you want a longer checklist for judging anyone who gives picks, I wrote 20 questions to ask before you buy sports picks.
What should you do instead?
Make your own number first, then look at the price. That is the whole method, and it is what turns "who wins?" into "is this a bet?"
You do not have to start from a blank sheet. My betting models are simple, probability-based Google Sheets, one per sport, that give you a fair price for every game. If you have a bet in mind and want a fair price on it in seconds, the AI Betting Assistant does that part for you. Either way, the next best bets article you read becomes something you can check instead of something you have to trust.
Questions people ask about best bets articles
Are best bets articles reliable?
Most are not reliable as betting advice, because they give a side without a fair price and are rarely graded in public. A pick can win and still have been a bad bet, and it can lose and have been a good one, so a list of winners and losers does not tell you much. The useful ones state a probability or a "bet it to" limit and publish a record graded at the price they posted. Without those, treat the article as an opinion about who wins, not as a judgment about whether the price is worth paying.
Why do experts' best bets lose so often?
They lose often because most games are priced close to correctly, and picking sides at those prices is close to a coin flip with a fee attached. At -110 (1.91) you need to win 52.38% of your bets just to break even. A writer who is good at football but does not compare a fair price with the market price will land near 50% over time, and that loses money once the margin is taken out. Losing is not proof of bad analysis. It is what happens when the analysis answers "who wins" instead of "is this price wrong".
What does "bet it to" or "playable to" mean?
It is the worst price at which the writer still thinks the bet has value. If a pick says "Team A -2.5, playable to -3.5", it means the writer would still bet it at -3.5 but would pass at -4. It is the most useful line in a pick because it gives you a fair price in disguise. If an article has it, you can check the current line and decide in seconds. If an article does not have it, you have no way to know whether the bet still holds after the line moves.
How much does a line need to move to ruin a pick?
It depends on how big the edge was to begin with, and most edges are small. If your number makes a team 55% and you get -110 (1.91), your expected value is +5.00%. If the line moves to -140 (1.71), that same 55% team is worth -5.71%. A move that looks small on the screen can turn a good bet into a bad one. That is why you should always check the price you can get now against your own fair price, not against the price printed in the article.
Should I follow best bets from a writer with a good record?
Only if the record is long, graded at the posted price and checked against the closing line. A strong record over 30 or 40 picks can easily be luck, because sports results are noisy over short samples. What carries more weight is whether the writer's picks keep closing at better prices than they were posted at. If they do, the writer is getting ahead of the market. If they do not, a hot record is likely to cool. Even then, you get the posted price only if you bet as soon as the pick appears.
Is it better to bet early or late in the week?
It depends on whether your number is better than the market's, and for most bettors the earlier lines are where a good number pays. Opening lines carry less information, so a sound model finds more gaps early in the week. Late lines are sharper, because the market has had days to correct them. If you are following someone else's pick, early is usually better too, since the price only gets worse once other readers and sharper bettors act on the same information.
Where to go from here
If you want to stop reading picks and start pricing bets yourself, this is the place to start.
$27, one payment. It includes the +EV betting model spreadsheet, a 45-minute video training that builds your first model with you, and the four-bucket staking system. 30-day money-back guarantee.
Get The Ultimate Modern Bettor's Toolkit for $27MB built Underdog Chance around one idea: make your own number first, then bet only when the price beats it. Over the last 20+ years his simple, probability-based Google Sheets models have helped thousands of bettors stop following picks and start pricing bets for themselves.
Last updated: October 23, 2026