Same Game Parlays: The Books Play Dirty

Last week we talked about parlays and how the house stacks the deck in their favor. This week we’ll build on that topic by talking about a specific brand of parlay: the notorious SGP!

The “Same Game Parlay” or SGP, is a recent innovation in sports betting that has become wildly popular with bettors and extremely profitable for sportsbooks (aka unprofitable for you).

As the Robin Hood of sports betting, taking from the house and giving to the bettors, I want to shed light on how these bets work, and tell you all how to minimize downside and boost your results.

Today we’ll review the math behind SGPs and go through “Dos and Don’ts” to get the best results. As always, we’ll make our bets of the week on some upcoming football.

Let’s get after it.

What are SGPs

Before the recent wave of legalized US betting, sportsbooks were reluctant to let you parlay two bets from the same game, like the favorite and the over, because those bets could be correlated.

If one happened, the other became more likely to also happen. Books didn’t want to offer typical parlay odds because sharp bettors could identify events that happened more frequently together, and use them to create a winning edge.

Then, in 2019, FanDuel flipped the script.

They began to model the correlation between events in the same game, going beyond spreads and totals to include a whole universe of prop bets, including individual player outcomes (e.g. — yardage, touchdowns, rebounds, assists). They then used these models to create dynamic pricing and offer the first ever Same Game Parlays.

SGPs focused on players hitting statistical milestones and were marketed like crazy. Bettors loved them.

By the fifth month after launch, adoption had ticked up to 5% of the online sports betting wagered on FanDuel.

At this point, over 90% of active FanDuel bettors have placed an SGP bet, and every major sportsbook has jumped on board to offer this product.

Beyond the popularity, sportsbooks are eager to offer SGPs because they’re profit machines. The average hold (house edge) on SGPs is 4x to 6x higher than straight bets, and 50% higher than traditional parlays.

Let’s now look at the math.

The Math (and The Extra Tax)

Last week we walked through traditional parlay math, and identified that sports books were paying +264 on a two-team parlay (with fair odds of +300), creating a 9% house edge.

SGPs are priced differently than traditional parlays in two ways.

1. First, the correlation between legs in the parlay is estimated, and the price is adjusted up or down accordingly. This allows the book to both protect themselves against overpaying on events more likely to happen together, and to offer more attractive odds on events with a negative correlation.

2. Second, the sportsbook takes advantage of the fact that their correlation estimates are not visible to the bettor, and hits the bettor with an extra tax (boosting the house’s edge), simply because they can.

Let’s tackle those one at a time:

Correlation Math

With an SGP, when outcomes are positively correlated, sportsbooks adjust your payout downward to account for that correlation.

Instead of paying +264 for a two-leg SGP, they will pay +200 or less.

Here’s an example, using this weekend’s Eagles vs. Steelers game:

• Let’s create an SGP based on the Eagles covering a first half spread of -3.5, and a full game spread of -5.5.

• In the cases where the first leg hits, the Eagles are ahead by 4 or more points at halftime.

• This makes it much more likely than 50/50 that they’ll win the full game by 6 or more, covering both legs.

• Based on that correlation, FanDuel offers us +137 on this parlay, vs the typical +264.

When outcomes are negatively correlated in an SGP, sportsbooks adjust your payout upward to reward you for predicting a comparatively rare event. Here they will pay more than +264.

Let’s use the inverse of our prior case as an example here:

• We create an SGP based on the Steelers covering a first half spread of +3.5, but the Eagles still covering a full game spread of -5.5.

• In the cases where the first leg hits, the Eagles are ahead by 3 points or fewer at halftime.

• This makes it much less than 50/50 that they’ll win the full game by 6 or more, covering both legs.

• Based on that negative correlation, FanDuel offers us +472 on this parlay, vs the typical +264.

While the actual correlation math is complex, usually requiring millions of Monte Carlo simulations for prediction, this piece of pricing SGPs is “fair” to the bettor. The next section is not.

Extra Tax

If SGP pricing was merely based on an adjustment for correlation, we’d see the same 9% house edge generated by betting all four combinations of a two-team parlay. What we’ll find, however, is an actual house edge that is often significantly worse.

Let’s try this on the spread and total of tonight’s Raptors vs. Heat game.

Here are the lines on FanDuel:

Reviewing each combination’s generates the following prices:

1. Heat -10 and Over 224.5 produces parlay odds of +250

2. Heat -10 and Under 224.5 produces parlay odds of +220

3. Raptors +10 and Over 224.5 produces parlay odds of +233

4. Raptors +10 and Under 224.5 produces parlay odds of +263

Not a single one of those combinations reaches our standard two-team parlay pricing of +264 — ouch!

If we look at a weighted average of those parlay combinations, this SGP will pay +241, significantly lower than a standard parlay’s +264, and creates a house edge of 14.8%.

In this instance, the book has added an extra 5.8% to their already large two-team parlay house edge.

This hidden tax is why sportsbooks love SGPs — they’re stacking extra juice into every leg without bettors noticing.

Dos and Don’ts

For reasons discussed above, I see some big risks with SGPs, but they’re a lot of fun to make and can produce winning opportunities. I have four key guidelines I follow to bet these well.

Here are my rules of thumb:

1) DO consider the distribution of stats in an average game, and bet within these bounds.

Every game only produces a finite number of yards, touchdowns, and other countable stats.

It's like a pie — there's only so much to go around, and if one player gets a bigger slice, someone else is getting less.

To provide some heuristics:

• NFL teams average about 2.64 TDs per game. If you're betting three players from the same team to score anytime touchdowns, you're essentially trying to predict every TD that team scores.

• There's a cap on receiving yards too usually between 240 - 300 per game. If you're betting three receivers on one team to each go over 70 yards, you're again trying for too much of the pie.

• High-scoring games still often cluster around one outlier player.  It’s easy to fool yourself into thinking that if an offense goes off, everyone will have a big game. To contradict that notion: Seahawks receivers Tyler Lockett and D.K. Metcalf have surpassed 100 receiving yards in the same game only twice in more than 90 career games.

The best SGP bettors recognize a finite pie of stats, and avoid trying to predict too much of any one category, instead betting on a smaller number of outliers stacked together.

2) DON’T forget about Unders.

SGPs are designed to induce overs and yes props. You see a ton of promotion of “anytime touchdown” bets, but almost never a mention of betting a player will not score. Many books don’t even offer these bets.

As we discussed our piece on totals, unders win more than 50% of the time. So grit your teeth for the experience of watching a whole game in dread of George Kittle busting a long TD, and incorporate some under legs into your SGP.

3) DO compare pricing at different books for your SGP, because their correlation estimates may vary significantly and produce very different odds.

I use Juice Reel as a starting point to identify the best line for my most critical SGP leg, but there’s really no substitute for spending the time to manually compare the same SGP amongst different books.

To demonstrate this, I compared pricing for the following SGP in tonight’s NFL game between all major NY sportsbooks.

• San Francisco -3

• Brock Purdy 250+ Passing Yards

• Cooper Kupp 80+ Receiving Yards

Check out the crazy spread of prices, ranked from best to worst:

You can almost always find an SGP boost promo of 30% or more to stack onto the odds you see above, so spend the time to maximize returns for the times your long shot bet hits.

4) DON’T stack a lot of (seemingly) easy wins together to produce an SGP.

I see “betting experts” on social media promoting many-leg SGPs that combine a lot of highly probable events.

Here’s an example where every leg is at least -1000.

Do not do this! 

The house edge applied to alternate lines of highly likely events tends to be much higher than outcomes that are closer to 50/50. I’d estimate the house edge on the SGP above at 50% or more.

The above approach takes a bad betting premise and makes it much worse. No thank you.

Bet of the Week $$

Last week we went 1-2, losing 0.6 units. We won big on Ohio in the MAC title game Saturday morning, but lost close ones with Penn State falling to Oregon and the total in the AZ/SEA game drifting past 45.

Even so, since starting the newsletter, bets given out in this section are ahead 4.0 units, at a positive 19% ROI. We’ll update this regularly.

Based on my research, I am making the following bets this week:

• Carolina Panthers ML @-145 on BetMGM for 1 unit

• Seattle +8.5 (teased from +2.5) & Minnesota -1 (teased from -7) @-120 on DraftKings for 1 unit

The Panthers have covered the spread five weeks in a row, but haven’t gotten in the win column during this positive stretch. I like them to finally do so.

The Wong teaser of Seattle/Minnesota crosses key numbers and is supported by the early betting data I see on Juice Reel’s SharpMush.

Next week we’ll expand our betting horizons with a different approach to winning.

Please fill out the survey about today’s newsletter and let me know your thoughts!