I will be honest — same game parlays nearly ruined my bankroll in my second year of NHL betting. I was building five-leg SGPs every night, convinced that combining “obvious” outcomes would produce easy returns at inflated odds. It took three months of negative ROI to understand why bookmakers love offering SGPs: the correlation penalties they bake into the pricing extract more margin than any other market type. Once I learned to work with those correlations instead of against them, SGPs became a controlled, occasional weapon rather than a nightly habit.
SGP Mechanics
A same game parlay combines multiple selections from a single game into one bet. Every leg must win for the parlay to pay. UK bookmakers calculate SGP odds differently from traditional accumulators because the outcomes within a single game are correlated — they are not independent events.
Here is the crucial point most bettors miss: the bookmaker adjusts your odds downward when your legs correlate positively and upward when they correlate negatively. Backing a team to win and the total to go over sounds logical — a winning team probably scored goals, which pushes the total up. But the bookmaker recognises that correlation and reduces the combined odds below what you would get by multiplying the individual prices. You are paying a correlation tax, and it is often steeper than you realise. The most common total line in the NHL sits at 6.5 goals, and combining the over with a moneyline favourite is one of the most popular SGP structures — which means it is also one of the most efficiently priced.
The best SGP value comes from legs that the market treats as independent but which your analysis reveals as correlated. That requires understanding which outcomes are genuinely linked and which the bookmaker’s algorithm mislabels. The league-wide save percentage dropping to 89.6% this season means more goals, more shots, and more counting-stat opportunities for individual players — all of which affect how prop legs inside an SGP should be priced.
Correlation Traps
The single biggest trap in NHL SGPs is stacking legs that all depend on the same underlying event. Backing a team to win, the total to go over, and a star player on that team to score is a three-leg parlay where all three outcomes are positively correlated. The bookmaker prices that correlation in, and you end up with combined odds that barely exceed what you would get from a standard two-leg accumulator. The extra leg adds risk without proportional reward.
A subtler trap involves goaltender saves and the total. If you back the under on goals and the over on a goaltender’s saves, you are creating a contradiction. A low-scoring game typically means fewer shots, which means fewer saves. The bookmaker’s algorithm might not catch this perfectly, but more often than not, it will — and you will be offered odds that reflect the inconsistency, making the bet almost impossible to win at the price given.
The correlations I target are ones the market underprices. A player on a trailing team taking more shots in the third period is correlated with the other team winning, but that player’s shots-on-goal prop is priced off his season average, not his behaviour in a specific game state. Combining the opposing team’s moneyline with my target player’s shots over is a positive-correlation play that the algorithm adjusts for less aggressively than the more obvious stacks.
Sample SGP Builds
Let me walk through two structures that have produced consistent value across my tracking.
Structure one: underdog puck line +1.5 combined with the game total under. Underdogs covered the +1.5 puck line roughly 60% of the time in the 2024-25 season, and that coverage rate improves in low-scoring games where margins stay tight. The correlation here is positive — tight games favour the underdog spread — but the bookmaker’s SGP algorithm does not always penalise it as heavily as the moneyline-plus-over combination because it is a less common build. I run this structure when two defensive teams meet and the total is set at 6.0 or below.
Structure two: a high-volume shooter’s shots over combined with the opposing goaltender’s saves over. These legs are positively correlated — more shots from the shooter means more work for the goalie — but the bookmaker prices them off different baselines (player stats versus goalie stats), and the correlation adjustment is often lighter than it should be. This structure works best when the shooter’s team is an underdog, because underdogs tend to generate shot volume through desperation and the opposing goalie faces a heavy workload. I pair this with a simple moneyline on the favourite for a three-leg SGP where all three legs quietly reinforce each other.
Sizing and Frequency
SGPs should occupy a small, clearly defined corner of your bankroll. I allocate no more than 5% of my weekly NHL budget to same game parlays, and I rarely build more than two per week. The variance is extreme — even a well-constructed SGP with a 25% estimated hit rate will produce long losing streaks that feel personal.
Stake size matters. I use a flat 0.5-unit stake for all SGPs, half my standard unit size for single bets. The reasoning is mathematical: a bet that wins 25% of the time at combined odds of 5.00 produces the same expected value per unit as a single bet at 2.00 that wins 55% of the time, but the ride is far rougher. Keeping the stake small protects me from the emotional damage of a ten-bet losing streak, which is statistically inevitable across a full season.
The biggest discipline challenge with SGPs is resisting the urge to add legs. Every additional leg feels like it adds value — “if those three hit, why not add a fourth for an even bigger payout?” Because that fourth leg drops your win probability by another 20-30% while the odds only increase by 40-50%. The maths punishes greed. Two or three legs maximum, every time. If you want to build bigger accumulators, do it across multiple games where the outcomes are genuinely independent, not within a single contest where everything is interlinked.
