Projection Models and Baselines
Every player prop starts with a projection. The sportsbook estimates how many strikeouts a pitcher will record, how many hits a batter will accumulate, how many total bases a slugger will tally. These projections form the baseline from which lines are set.
Components of a Projection
Modern projection systems combine multiple data sources:
- Historical performance: A player's career and recent-season statistics, weighted by recency and sample size.
- Matchup adjustments: How does this hitter perform against left-handed pitching? How does this pitcher fare against high-contact lineups?
- Contextual factors: Park factors, weather conditions, umpire tendencies, lineup position.
- Usage expectations: How many plate appearances will this hitter get? How deep will this pitcher work?
The best sportsbooks use proprietary models that synthesize these factors into a single expected value for each prop. Smaller books may license projections from third parties or rely on simpler heuristics.
The Quality of Projections
Not all projections are equal. Major sportsbooks invest heavily in modeling talent, employing quantitative analysts who refine projections continuously. Smaller books may use cruder models or simply copy lines from market leaders.
This quality gap creates opportunity. When a smaller book sets a line based on an inferior projection, sharper books and bettors move the line. If you can identify the inferior projection before the line moves, you capture the edge.
Where Projections Fail
Even sophisticated projections have blind spots:
- Unusual matchups: A pitcher with a unique pitch mix facing a lineup with a specific vulnerability may not be fully captured by generic adjustments.
- Recent information: Lineup changes, minor injuries, or workload concerns announced close to game time may not be incorporated.
- Small samples: Players with limited history against specific opponents or in specific conditions present modeling challenges.
- Non-public information: Clubhouse factors, personal issues, or undisclosed physical concerns that affect performance are invisible to models.
The gap to exploit: Your edge comes from having better information or better interpretation than the sportsbook's projection model. If you identify a matchup-specific factor the model underweights, or react faster to late-breaking information, you can beat the line.
Outcome Distributions and Line Setting
Once a sportsbook has a projection, it must convert that projection into a line. This requires understanding not just the expected outcome, but the distribution of possible outcomes around that expectation.
Discrete Outcomes in Baseball
Most hitter props involve discrete outcomes. A player can record 0, 1, 2, 3, or more hits, but not 1.7 hits. The sportsbook models the probability of each discrete outcome to determine where to set the line.
Projection: 1.1 expected hits
Probability distribution:
0 hits: 33%
1 hit: 37%
2 hits: 20%
3+ hits: 10%
Under 1.5 hits: 70% | Over 1.5 hits: 30%
With this distribution, the sportsbook sets the line at 1.5 hits. The under is heavily favored at 70%, so the book offers something like Under -200 / Over +160 to balance action and extract margin.
Choosing the Line Number
The sportsbook chooses a line number that creates actionable odds on both sides. Setting the line at 0.5 might make the over too expensive to generate action. Setting it at 2.5 might make the under so cheap that nobody bets it. The line number is chosen to balance these considerations.
For most hitters, lines cluster around 0.5, 1.5, or occasionally 2.5 for hits. For total bases, the range extends from 0.5 to 2.5 or higher for power hitters. The book picks the number that produces the most balanced and liquid market.
The Poisson Distribution
Many counting stat distributions follow something close to a Poisson distribution, where outcomes cluster near the expected value with a long tail of higher outcomes. Strikeouts, hits, and runs all exhibit this pattern.
Understanding the shape of these distributions helps you evaluate whether a line is fair. If the distribution is highly skewed (most outcomes are 0 or 1, with rare spikes to 3+), lines set at 1.5 or 2.5 may offer value on the side with lower but more reliable probability mass.
Juice, Hold, and the Vig
Sportsbooks are businesses. They make money by charging more for bets than the fair value of those bets. The mechanism for this is the juice (also called vig or vigorish), the margin built into every line.
How Juice Works
Consider a fair coin flip. True odds are +100 on each side (2.00 decimal), reflecting 50% probability each. A sportsbook offering this market might price both sides at -110. This means you must bet $110 to win $100. The implied probability of -110 is 52.4%, not 50%.
With both sides at -110, the implied probabilities sum to 104.8% instead of 100%. That extra 4.8% is the juice. No matter which side wins, the sportsbook collects more in losing bets than it pays out in winning bets (on average), creating profit.
Prop Market Juice
Player props typically carry higher juice than sides and totals. While game spreads might be -110/-110 (4.5% hold), props often run -115/-105, -120/-100, or even -130/+100. This higher juice reflects the lower limits and higher volatility of prop markets.
| Market Type | Typical Juice Structure | Implied Hold |
|---|---|---|
| Game spread | -110 / -110 | 4.5% |
| Game total | -110 / -110 | 4.5% |
| Player prop (standard) | -115 / -105 | 5.5% |
| Player prop (sharp market) | -120 / +100 | 6.5% |
| Alt lines / long shots | Varies widely | 8-15% |
Overcoming the Vig
The juice is your hurdle rate. If you are paying 5.5% juice on average, you need a 52.75% win rate to break even on -110 bets. Any edge you identify must exceed the juice to be profitable.
This means marginal edges are not worth betting. If you think you have a 2% edge on a prop with 6% juice, you are still losing expected value. Bet only when your edge clearly exceeds the juice, and consider shopping multiple books to find the best available price.
The juice reality: Higher juice on props means you need larger edges to profit. A 5% edge on a -110 side bet is worth more than a 5% edge on a -130 prop bet. Factor juice into every bet decision, not just expected win probability.
Market Shading and Public Tendencies
Sportsbooks do not always set lines to reflect true probability. They shade lines toward outcomes they expect recreational bettors to favor, extracting additional margin from predictable public behavior.
How Shading Works
Imagine a star player whose true strikeout probability is 45% over and 55% under. A fair line might be Over +100 / Under -120. But the book knows recreational bettors love betting on stars to perform well. They expect 70% of money on the over.
Rather than face imbalanced liability, the book shades the line: Over -120 / Under +100. The over is now worse value, but recreational bettors still bet it because they want action on the star. The book profits from this behavioral tendency.
Predictable Public Preferences
Sportsbooks have learned what recreational bettors like:
- Overs on star players: People bet on Aaron Judge to hit a home run, Shohei Ohtani to record strikeouts, not against them.
- Popular team players: Yankees, Dodgers, Cubs props see more over action than Royals or A's props.
- Recent hot streaks: A player who went 5-for-12 last series gets hammered on overs, regardless of underlying quality.
- Prime time games: National TV matchups attract casual bettors who bet overs on recognizable names.
Exploiting Shaded Lines
Shading creates value on the contrarian side. If a star's under is priced at +100 when fair value is -120, you are getting a premium to bet against public perception. This does not mean you should blindly bet every under on popular players, but you should recognize when the line has been shaded and factor that into your analysis.
The key question: Is the line reflecting true probability, or is it reflecting expected betting patterns? When those diverge, the probability side offers value.
The contrarian angle: Some of the best prop value comes from betting against popular players when the line has been shaded to accommodate public money. Stars going through rough patches are particularly valuable because the public keeps betting the over while the line remains inflated.
Why Alt Lines Exist
Beyond standard prop lines, most sportsbooks offer alternate lines at different numbers with adjusted odds. Understanding why these exist and how they are priced reveals additional opportunities and pitfalls.
The Purpose of Alt Lines
Alt lines serve multiple purposes for sportsbooks:
- Customer choice: Bettors have different risk preferences. Some want long-shot payouts; others want higher-probability plays at lower odds.
- Parlay integration: Alt lines let bettors create same-game parlays with multiple player props at different thresholds.
- Market coverage: Alt lines capture bettors who do not like the main line but would bet a different number.
How Alt Lines Are Priced
Alt lines are derived from the same underlying distribution used for the main line. If the book models a player's hit distribution, it knows the probability of 0, 1, 2, 3+ hits. From that distribution, it can price lines at 0.5, 1.5, 2.5, or any other number.
Distribution: 0 hits (33%), 1 hit (37%), 2 hits (20%), 3+ hits (10%)
Over 0.5 hits: 67% true probability → Approx -220
Over 1.5 hits: 30% true probability → Approx +230
Over 2.5 hits: 10% true probability → Approx +800
Alt Line Pitfalls
Sportsbooks often price alt lines with higher margins than main lines. The further from the main line, the higher the juice. A +800 payout on over 2.5 hits might reflect 10% true probability, meaning fair odds would be +900. That 100-point gap represents significant extra margin.
Long-shot alt lines are particularly juicy. Sportsbooks know recreational bettors love lottery-ticket payouts, so they charge a premium. Before betting alt lines, calculate the implied probability and compare it to your estimated true probability. The edge must exceed the inflated juice.
Why Prop Limits Are Lower
Sportsbooks set betting limits, the maximum amount they will accept on a single wager, based on market characteristics. Prop limits are consistently lower than limits on sides and totals. Understanding why reveals important truths about prop markets.
Liability and Variance
Props are more volatile than game outcomes. A game spread is influenced by 50 players and hundreds of game events. A player prop depends on one player's performance in a small sample of opportunities. This concentration creates higher variance and harder-to-hedge liability.
If a sportsbook takes $50,000 on a game spread, it can hedge across multiple positions or let variance average out over many similar bets. If it takes $50,000 on a single player's strikeout prop, one player's performance determines the entire outcome. Books limit exposure to concentrated risk.
Model Uncertainty
Sportsbooks have less confidence in their prop models than their game models. Game lines are sharpened by high volume and sophisticated bettors who correct mistakes quickly. Prop lines see less action, less sharp correction, and more potential for mispricing.
Lower limits protect the book when its model is wrong. If a sharp bettor identifies a mispriced prop, limiting the bet to $500 or $1,000 caps the book's exposure. On a game spread, the book might accept $10,000 or more because it has more confidence in the line.
Implications for Bettors
Lower limits mean you cannot bet as much on props even when you find genuine edge. This changes optimal strategy:
- Volume over size: Finding many small edges is more valuable than finding one large edge you cannot fully exploit.
- Line shopping matters more: Getting the best available line increases your expected value when you cannot bet large amounts.
- Steaming is limited: Sportsbooks will limit or ban you faster for winning prop bets than for winning game bets. Sustainable edge requires discretion.
The practical reality: You might identify a 10% edge on a prop, but if you can only bet $250, your expected profit is $25. Props reward consistent grinding more than occasional large bets. Structure your approach accordingly.
Where Edges Come From
Understanding how sportsbooks price props reveals where edges emerge. Edges are not random; they cluster in predictable areas where sportsbook processes have blind spots.
Edge Sources
- Matchup-specific information: Sportsbooks use generic adjustments for platoon splits, pitch-type matchups, and similar factors. If you identify a specific matchup that warrants an extreme adjustment, you may find value the generic model missed.
- Late-breaking information: Lineup changes, weather shifts, umpire assignments, and minor injury news emerge close to game time. Sportsbooks may not adjust lines immediately. React faster than the market to capture edges.
- Market shading overcorrection: When public money pushes a line beyond fair value, contrarian bets have positive expected value. Identify situations where shading has created a gap between the line and true probability.
- Model blind spots: Unusual pitchers (openers, piggyback starters), non-standard lineups, or exceptional circumstances may not be well-modeled. Human analysis can identify value that algorithms miss.
- Juice inefficiencies: Not all books charge the same juice. Shopping for the best available line reduces your hurdle rate, converting marginal edges into profitable bets.
What Is Not Edge
Many things that feel like edge are actually well-priced by the market:
- Obvious matchup advantages: Everyone knows a high-strikeout pitcher facing a high-strikeout lineup should fan more batters. The line reflects this.
- Hot streaks and cold streaks: Sportsbooks adjust for recent performance. The line already incorporates what you see in the last week's results.
- Public narratives: If a storyline is on ESPN, it is priced into the line. The market is efficient at incorporating widely-known information.
Sustainable Edge
Edge in prop betting comes from doing more work than the market. Sportsbooks use good models but cannot perfectly capture every relevant factor. Bettors who systematically identify underweighted factors, react faster to new information, or find shaded lines can develop sustainable edge.
The key is discipline. Edge is not about gut feel or favorite players. It is about having a repeatable analytical process that identifies mispriced lines more often than it generates false signals. Over time, correct process plus edge identification compounds into profit.
The bottom line: Sportsbooks are good at pricing props, but not perfect. Edges exist in matchup-specific information, fast reaction to news, contrarian plays against shaded lines, and model blind spots. Your job is to systematically identify these edges while avoiding the many traps that masquerade as opportunity.