Orbital Sciences Chief Executive David W. Thompson is not a guy I would ever want to play poker with. Discussing the company’s “go-forward” Commercial Resupply Services (CRS) cargo contract for the international space station and its Antares plans with Wall Street analysts Nov. 5 — less than a week after the smoke had cleared over Wallops Island, Virginia, from the rocket’s Oct. 28 launch failure — Thompson was confident the company would be able cover its commitment to NASA with minimal cost out of its own pocket. Clearly, Orbital continues to hold cards close to its vest as it juggles not one, not two, but three different hands. And I’m not quite sure where it may be bluffing.

The first hand is Orbital announcing it would buy one or two third-party launches for the Cygnus cargo vehicle, with a first flight as early as the second quarter of 2015. Discussions are taking place with two U.S. companies and one European company that Thompson wouldn’t name.

I presume those names are Space Exploration Technologies Corp., United Launch Alliance and Arianespace, but I am smelling a bluff already. The least likely candidate in my mind is Arianespace. Ignore integrating Cygnus on a Soyuz or International Traffic in Arms Regulations and clearing customs when moving the cargo spacecraft out of the country. Simply consider logistics plus contracts in flying Cygnus, its support equipment and technicians from Wallops Island to Europe’s Kourou, French Guiana, spaceport for a flight on a Soyuz. That’s 4,300 kilometers with no direct commercial flights between the United States and French Guiana. Baikonur Cosmodrome is a distance of 9,825 kilometers, so the logistics are even worse.

The main principle in NASA’s CRS contract is promoting U.S.-flagged and -built launch capabilities. Flying on top of a Soyuz, regardless of location, would cause political discomfort even if NASA were provided a waiver. Arianespace is exceedingly unlikely due to cost and political fallout.

SpaceX and ULA, being fine American companies with facilities in Florida, are much more likely. They’ve had lots of commercial flights, transport of Cygnus and support equipment is much shorter, and Orbital is intimately familiar with how things work down at the Cape, so there’s a good comfort level.

Trying to tell who holds the strong hand is more difficult. SpaceX has a ramping production line for Falcon 9 but also a backed-up launch manifest through 2015. Freeing up a launch slot might require either NASA delaying a satellite mission or Orbital providing financial compensation to a SpaceX customer in exchange for a later flight. Falcon 9 is the lowest-cost launch vehicle available and gets bonus points because Cygnus could be horizontally integrated onto the stack and available for cargo load in the same orientation, just like SpaceX’s Dragon capsule.

ULA would love to sneak in a Delta 2 launch, but last-minute cargo loading would be more challenging due to vertical integration of Cygnus with the launch vehicle. Cost may be the biggest barrier, with estimates of over $90 million for the last Delta 2 launch performed for NASA as compared with a list price of $54 million for a stock Falcon 9.

Orbital’s second hand of poker is with engine manufacturers. Two Russian manufacturers and ATK are on the short list, Thompson told investors. ATK appears to be a straightforward solution, a more powerful solid-fuel first stage. Orbital would gain integration efficiencies and better profit margins on Antares once the company completed its proposed merger with ATK. Going to an all-solid Antares design would require a pad overhaul at Wallops and safety rules reviews. The state of Virginia might not be keen on reworking the pads, having invested tens of millions of dollars in provisioning Pad 0A specifically to handle liquid fuel operations for Antares.

Orbital officials have often hinted how they wanted to drop the Russian-made RD-180 or another engine into the Antares first stage for a quick upgrade.

Buying from the Russians is out of fashion these days, worsening every time they drive a military convoy into Eastern Ukraine, but Orbital might be able to thread the needle with procuring a new build NPO Energomash engine directly from the factory. Given the failure of the AJ-26, offers for new build engines from Kutznetsov are a nonstarter at this point since they would require restarting the production line of the NK-33. Orbital would take some public relations heat for buying from Energomash, but a liquid first stage would appear to be the best fit to Wallops in terms of safety and cost.

The last and most private poker game is taking place between Orbital and Aerojet Rocketdyne. Mystery surrounds the May AJ-26 test stand failure at NASA’s Stennis Space Center in Mississippi. Nobody from Orbital or Aerojet shed light on exactly what happened, other than some vague hints at misconfigured hardware and Aerojet taking a contract loss of $17.5 million at the end of its fiscal third quarter and a total of $31.4 million on the program for the year to date. Nobody discussed details on the two potential root causes and Orbital asserted that both could be screened during engine inspection.

The loss of the Orb-3 Antares in such a dramatic fashion, with all preliminary evidence in telemetry and video pointing to a turbopump failure in one of the two AJ-26 engines 15 seconds after launch, leaves Aerojet with a weak hand and Orbital holding all the cards. Odds are Aerojet will quietly ante up a rather significant amount of money for providing subnominal hardware, rather than engage in a public lawsuit with the potential for compensatory damage awards to Orbital’s reputation. Orbital hasn’t been shy about using lawyers, having sued ULA for access to the RD-180 last year.

Out of these three poker games, Orbital has almost guaranteed odds of collecting compensation from Aerojet. It holds a good hand when selecting a U.S. launch provider for CRS stopgap flights in 2015. The only ambiguity is what rocket engine it will select for upgrading Antares. I’m willing to bet on a liquid-fuel engine, but not if Dave Thompson is showing up with the cards.

Doug Mohney is a technology writer who has attended five Antares launches (four successful).