Favorites vs Underdogs: Which Actually Wins Bettors Money

Neither favorites nor underdogs win bettors money on their own — price does. The sharp question isn't which side wins more games, it's which side wins more often than the implied probability baked into the odds. Favorites hit at a high clip but rarely at a price that justifies the risk; underdogs win less often but can be profitable when the market undervalues them. Here's how to think about it like someone who actually makes money betting.

What's the Real Difference Between Favorites and Underdogs?

A favorite is the side the market expects to win, priced with negative American odds (e.g., -180). An underdog is the side expected to lose, priced with positive odds (e.g., +150). The odds encode an implied probability — that's the percentage of the time a side needs to win for a bet to break even.

The critical concept is break-even win rate. At -180, you need to win 64.3% of your bets to break even. At +150, you need only 40%. Most bettors fixate on who wins the game. Sharp bettors fixate on whether a side wins more often than its break-even rate.

Do Favorites or Underdogs Win More Money Long-Term?

Across major sports, the historical edge leans toward underdogs — specifically, underdogs getting enough points or money to cover the gap between their true win probability and the market's implied probability. This isn't because underdogs are secretly better. It's because of two structural forces:

1. Public bias toward favorites. Casual bettors like betting the better team. Books know this and shade favorite pricing accordingly, making favorites slightly worse values than they should be.

2. The math of positive odds. When you bet a +200 underdog, you only need to win 33.3% of the time. Even modest underdog overperformance compounds. With favorites at -200, you need 66.7%, and small underperformance kills you.

The ROI Gap

Studies of closing line data across the NFL, NBA, and MLB consistently show that betting all underdogs on the moneyline produces a smaller loss (or even a slight profit in certain seasons) compared to betting all favorites, which reliably bleeds money due to vig and public shading. Neither blind strategy is a goldmine — but the underdog side is where value lives more often.

A Worked Example: Why Picking Winners Isn't Enough

Let's say the Boston Celtics are -200 at home against a weaker opponent. The implied break-even rate is 66.7%. You think the Celtics win 70% of the time — a solid edge, right?

Bet $200 to win $100. Over 100 such games at a true 70% win rate:

• You win 70 times: +$7,000
• You lose 30 times: -$6,000
• Net: +$1,000 (a 5% ROI on $20,000 wagered)

Now compare a +200 underdog in another game. Say the Florida Panthers are +200 on the road. Break-even is 33.3%. You think they win 37% of the time.

Bet $100 to win $200. Over 100 games at a true 37% win rate:

• You win 37 times: +$7,400
• You lose 63 times: -$6,300
• Net: +$1,100 (a 5.5% ROI on $10,000 wagered)

The underdog wins far less often (37% vs 70%), yet produces a comparable — even slightly better — return on investment. That's the power of price. You don't need to be right more often. You need to be right more often than the odds require, and underdogs give you more room for that gap to exist.

When Should You Actually Bet Favorites?

Underdogs aren't automatically better. There are situations where favorites are the right play:

• Short favorites in high-variance spots. A -120 favorite you believe wins 60% of the time is a strong bet. The problem is usually the price, not the side.
• Playoff or elimination games where the better team's edge stabilizes. The Kansas City Chiefs in a playoff spot as -150 can be a legitimate value if the market hasn't adjusted enough for the reduced variance of a single-elimination format.
• Situations where the market overcorrects on an underdog narrative. When the public falls in love with a trendy dog, the favorite's price can drift into value territory.

The rule: bet a favorite when the price implies a lower win probability than you believe is true. That's it. Don't bet favorites because they'll probably win. Bet them when they're mispriced.

Common Mistakes Bettors Make on Both Sides

Favorite mistakes:

• Chasing big favorites on the moneyline in baseball and hockey, where variance is high and -250 prices offer terrible risk-reward.
• Assuming a team being "clearly better" means the price is justified.

Underdog mistakes:

• Betting underdogs just because they're plus money. Plus odds don't equal value.
• Falling for the "any given Sunday" logic without checking whether the number is actually beatable.

The mistake on both sides is the same: ignoring whether the price reflects reality.

How Da Vinci Bets Approaches Favorites vs Underdogs

Our model doesn't have a built-in preference for favorites or underdogs — it has a preference for value. Here's how that works in practice:

We calculate our own win probability for every game using team-level metrics, situational factors, and market data. We then compare that probability to the implied probability from the current odds. If our model says a side wins 44% of the time and the market is pricing them at a 38% break-even rate, that's a bet — regardless of whether that side is a favorite or an underdog.

What the model tends to find is that value shows up more frequently on the underdog side, particularly in the +130 to +220 range, where public bias is strongest and books have the most incentive to shade. But the model will flag a -130 favorite just as quickly if the edge is there. The point isn't to pick a side. It's to let the math tell you when the market is wrong.

The Takeaway

If you take one thing from this: stop asking whether a team will win and start asking whether the odds give you a better deal than reality demands. Underdogs tend to offer more of those deals because of public bias and the math of positive odds — but the principle applies to every price on the board. Price beats picking winners. Every time.

Unlock all picks, signals, and deep analysis. Go Premium for full access to our AI prediction engine.