Royals vs. Blue Jays Preview — September 6, 2026

Kansas City didn't just beat Spencer Arrighetti on August 27 — they treated him like a bullpen session gone wrong, pouring 13 runs onto the Blue Jays in a 13-2 demolition. Ten days later, the same right-hander is being asked to do it again, this time at Kauffman Stadium, against the same confident Royals lineup. That rematch is the single most instructive data point for Sunday's total, and Earl's model believes the market has not fully digested it.

Earl's model projects 10.12 total runs against a posted line of 9.5 — a full run of daylight. The model's primary lean is the Over, and the expected value mathematics back it up (+0.35). For the moneyline, the model sees Toronto winning this by roughly two-thirds of a run, 5.4 to 4.7, and rates the Blue Jays' -122 price as carrying the strongest EV on the board at +3.36. The model's reasoning is direct: Kansas City's lineup is generating offense in ways that point to a lively run environment, and the market's own run expectation is leaning cautiously in the same direction.

The strongest individual piece of that puzzle is Arrighetti's current structural weakness.

The Revenge Factor Is Real

Let's be precise about what Arrighetti has done this season: 0-3 with a 6.55 ERA, and he's allowed 3 home runs in just 11 innings. His hit rate (8.18 per nine) is tolerable, but the damage behind it — the long ball and a 4.09 walks-per-nine — has been catastrophic. And the Blue Jays keep losing when he's on the mound: Toronto has dropped all three of his decisions.

Then there's the August 27 start. Arrighetti faced these exact Royals and the game spiraled to a 13-2 Toronto loss. That is not an anonymous data point; that is a scoutable, recent, and team-specific failure. Kansas City's lineup has already shown it can time his stuff, sit on his mistakes, and punish them in bunches. The psychology of a rematch cuts both ways, but when the first meeting was a blowout, the advantage sits with the hitters who have seen — and crushed — the arsenal.

There's a scheduling quirk worth flagging here: this is the third meeting between these two hurlers' teams in eleven days. On August 26, Randy Dobnak took a 3-0 loss to Toronto after Kansas City was blanked. The next day, Kansas City answered by erasing Arrighetti. Both teams enter Sunday with vivid, recent memory of exactly what the other side's starter throws. In a matchup where familiarity is this fresh, the edge historically belongs to the offense, which gets four chances to react versus a pitcher who must execute every pitch. That asymmetry is one reason Earl's model has the total clearing 10.

Dobnak's ERA Is Wearing a Costume

The counter-argument for the Under rests almost entirely on Dobnak's shiny 2.65 ERA. And it's a legitimate number — but it is also the most misleading surface stat on this board when you stack it against his profile.

Consider what's underneath: a 1.39 WHIP and 8.9 hits allowed per nine innings. Dobnak has allowed 57 hits in 57.2 innings. An ERA under 3.00 with that kind of traffic is a house of cards, and the model treats his recent form as a regression candidate, not a shutdown anchor. Pair that with 3.59 walks per nine and you have a pitcher who is constantly pitching from the stretch, constantly giving the opponent free baserunners, and constantly flirting with the big inning.

The recent starts tell that story. On September 1, Miami put a 6-3 loss on him. On August 26, Toronto held Kansas City to zero while he took the L. He did beat Oakland 6-2 on August 20, but his season has been alternating between quality and vulnerability with no reliable pattern. Against a Toronto offense that grinds at-bats and reaches base in waves, Dobnak's baserunner habit is a direct threat to the Under.

Toronto scores a modest 4.05 runs per game, but they've been better at home (37-35 home record for a team that's 72-71 overall) and they are playing with genuine playoff urgency in every inning. They're the better team on paper and the market knows it — they opened as -118 road favorites and have been bet further to -122 while Kansas City drifted from +100 to +104. The money is on Toronto, and the model agrees, but that's a secondary play. The total is where this game is won.

A Bullpen That Can't Be Trusted to Protect Anything

Here is where the premium angle separates itself from the public preview: the Royals' bullpen is currently a MASH unit, and it matters enormously for a 9.5 total.

Scan the Kansas City injury report and count the relief arms: Cole Ragans, Carlos Estévez, James McArthur, Nick Mears, Stephen Kolek, Tony Gonsolin, Alec Marsh — all on the injured list, most of them on the 60-day. That is the entire back end of a bullpen structure, gone. In a game where the model expects both starters to allow traffic, the bullpen that takes over is the real Over engine. Kansas City already allows 4.92 runs per game, and that number is inflated precisely because their relief corps cannot slam the door. If Arrighetti gets pulled early — and given his 6.55 ERA and the 13-run memory, that is a live scenario — Toronto's own thinned-out relief staff will be tested against a Royals offense the model believes is manufacturing runs effectively right now.

Toronto is not healthy in the pen either, with Jameson Taillon, Shane Bieber, José Berríos, Bowden Francis, and Yimi García all out. That means both teams are operating with patchwork pitching plans behind their starters. When two offenses face struggling starters and both bullpens are compromised, a 9.5 total simply doesn't hold. Kauffman Stadium's park factor (97 overall, and notably an 88 home-run factor) suppresses the long ball, but the model's projected 10.12 total suggests these offenses can score without needing the cheap home run — through baserunners, doubles, and the kind of manufactured damage the model specifically flags in Kansas City's current approach.

Why the Run Line Is a Trap (and the Moneyline Isn't)

Let's parse Earl's model output carefully, because the betting card here requires nuance.

The model's projected margin is -0.65 runs — Toronto by less than a run. That is why the model's run-line pick is KC +1.5; at the projected margin, the Royals cover more often than not. But the expected value on that run-line wager is negative (-6.95). That is a critical distinction for a sharp bettor: the model thinks the Royals stay close, but the price offers no compensation for that outcome. Fading the run line — refusing to lay money on +1.5 — is the correct execution even though the direction of the cover is likely correct. There is no edge at that number, and the sharp play is to skip it entirely.

The moneyline is where the value sits. Toronto at -122 with a projected margin of 0.65 runs and a positive EV (+3.36) is the model's second-most confident recommendation after the Over. The Blue Jays are the better club, they are playing meaningful September baseball while Kansas City's season is slipping away (they just lost a home series to Miami and sit 63-80), and the market's line movement toward Toronto confirms the sharps agree with the model's direction.

The Public Is Betting the Wrong Story

The casual public looks at a home dog in Kansas City, sees a 63-80 team playing out the string against a playoff contender, and assumes the Royals have nothing to play for. That's actually backwards for this market. Kansas City is 37-34 at home, which is respectable, and their lineup — the one that dropped 13 on Arrighetti — is the reason the total is high. The public also sees Dobnak's 2.65 ERA and assumes a pitcher's duel, ignoring that his 1.39 WHIP and 8.9 hits-per-nine are screaming for regression.

The sharp play is to trust the mismatch that's already been proven on the field: Arrighetti cannot handle this Royals lineup, Dobnak's profile is more fragile than his ERA suggests, and both bullpens are gutted. Earl's model sees a 10-run game in a park that suppresses homers but can't suppress traffic — and the traffic is going to be heavy.

The card: Over 9.5 (primary play). Toronto Blue Jays moneyline -122 (secondary). Pass on the KC +1.5 run line — directionally likely, but there is zero value in the price.