Because returning the samples requires vastly more money, time and complexity then landing the rover. I agree with Ozan. We currently have No rocket that can land on Mars and bring back even 1kg. We have Exactly One rocket that is currently well along to bring back tens of tons with in-situ refueling. It will be tested on the moon in 2026-7. With the addition of Hall thrusters, a small nuclear plant (already being developed by NASA), and a small fuel production plant, you have the solution. Only the in-situ fuel production plant will be novel by 2028. Given Starship's capacity to land 10s of tons, you could provide a dozen Tesla bots (hardened) to set up the equipment, drive in the rover to get the samples and return. Given Spacex's ability to mass produce Starships cheaply and rapidly (they expect to be producing at least 1/week by 2028) they can attempt multiple landings if a ship or two is lost (Musk calls that a "learning experience"). Starlink is now set to bring in $billions per quarter when they start supporting backbone traffic for Microsoft/Google/etc (they have contracts) on the E-band that they just got license from FCC on top of the subscription svcs on top of the cell services. Accelerating rapidly as Starshp starts dropping 120 Starlink V2large sats into orbit at a time. Ie - loads of money to fund the advancement of Starship/Moon, Starship/Mars well before 2028. As stated in the discussion, Spacex has both the deep pockets, the experience, the right type of industrial capacity and the motivation to do this mission with their own money once the $4billion runs out, as this is a natural precursor to sending actual people to Mars... their end game. A perfect dry run. A dual purpose mission. I think there's only really one choice. You won't need other ride shares for money, MSR will have a Much greater mass limit. A dozen Optimus bots (or similar) can be readily transported and allow alot of additional flexability. No need to pause. The tech is already being developed rapidly. Starship/lander Robots Compact Nuclear Reactors In-situ fuel production Battery tech is also improving for the rover. All this tech will be ready before 2028. Musk already expressed enthusiasm for this.

@@avgjoe5969 Nothing is stopping us from sending a rover to the surface of Mars to gather the samples and having them ready(1 rocket for that)..and having it ready to load on a very small shuttle module to be placed in orbit (another rocket for this, that's two). There's nothing stopping use from do this is there? nothing what so ever.

Probably by the success rate of Falcon 9, Falcon Heavy, and Dragon, which have been the cheapest and most reliable space vehicles ever made of their type. SpaceX has proven they can execute on fixed-price contracts, whereas other contractors have not.

SpaceX's Falcon 9 has had more successful landings than most other launch systems have had launches. I think there's Soyuz which has over 700 launches, but Falcon 9 is on track to beat that in a few years if Starship doesn't render F9 obsolete before then. As far as rocketry goes, SpaceX is doing amazing work. Starship/Super-Heavy test flight 4 next Wednesday, I'm hoping they can get the booster back for a landing and have the Starship maintain attitude control and be correctly oriented for reentry. Once they can successfully recover boosters their entire program will step up a notch in terms of cadence, and once Starship is capable of controlled reentry the next step is mastering orbital refuelling, after which not even the sky is the limit. With the current state of affairs I don't see a viable alternative to SpaceX landing astronauts on Mars to recover those sample caches by hand with or without a rover mission to collect the scattered caches and return them to a pickup location. The MSR mission is hobbled by needing a large ascent vehicle for the autonomous mission, which flows back through the rest of the design of that mission to require the largest everything ever sent to Mars, only problem is they are lacking the budget to match.

Spacex had 98 orbital launches last year, 2 test and 96 successful launches and landings. They are on track for more than 130 this year based on the first 4 months. They have a monumental ship building capacity that no one else has. (They expect to be cranking out at least 1 full Starship stack per Week by 2028 using the new Starship factory nearing completion and the 2 Raptor engine production facilities. Current rate without the factory is already 4 full Starships per year and 350 engines/year right now. They will have a fleet of supporting tankers, an orbital fuel depot (left over from the Lunar landing scheduled for 09/2026) and 4 separate Starship specific launch pads 1 completed currently, 1 near completion, 1 building, 1 planned). They are the only entity (govt or private) that has launched and landed orbital rockets. The Starship lander being designed was modified from designs that were intended to land on Mars and take off again with in-situ produced fuel. They have been working on this for maybe 5 years. They designed and built Raptor 3, the best rocket engine ever built, massively outstripping the competing engines and, as a Methane engine, ideally suited to mars as fuel can be made out of CO2 from the atmosphere and water. Just add a nuclear reactor and compact production unit to produce fuel... both of which NASA was building. If it mattered, they are also the only private concern that built a capsule to deliver cargo and personnel to the ISS... and they did this multiple times... at the lowest cost of any company or government. They are funded by the most massive Comm Satellite network ever created (Spacex built and owns more than half the satellites in earth orbit - 5500+ operational with numbers expanding quickly) (meaning they will have very deep pockets going forward as their capacity and revenue are expanding very rapidly). This makes them immune to budget cuts and congressional reversals. (They already succeeded nearly a year early on the Rural Broadband Grant Program. The FCC then stiffed them out of their $840m payment last year for political reasons, but they completed the project without the funding. Musk's affiliated companies are building human form robots (which would allow massive mission flexibility as they require little direct input - they don't need real time communications), comms satellites (that might be used for near real time comms to the landing site). The Starship launches have been very successful, Each launch revealed issues that were corrected. They are using $100m rockets (yes, they build the huge rocket that cheaply) to rapidly iterate and gain experience building starships faster and better. SLS, for example takes a year to build. Spacex already do this 4x faster for less than 1/10th the price. The fourth launch in May will demonstrate correction of the roll problem and show a simulated tower landing (at sea) with a real tower landing on flight 5 if that goes well... if flight 5 brings back the ship... then it is not only flight worthy, but now reusable and can be used over and over, dropping the price for each launch to under $20m... compare that to $1.5billion for SLS for each launch. Starship is designed to land 10s of tons of cargo (possibly as high as 100 tons) on Mars and take off with at least 100t after in-situ refueling. No one else can move that much mass to/from Mars. Importantly, Musk's purpose for Spacex is to build a self sustaining city of 1 million people on Mars (this is why he needs so many ships). The $4billion will get him started, but he can and will fund the rest to get the job done. If the mission fails, he will do it again until he succeeds. At his cost. Because this needs to be perfected before he lands actual people on Mars. Spacex is the only entity with the proper manufacturing base (in fact it is gross overkill), the money, the relevant expertise, the motivation to do this. Their lunar mission in 2026 will be the perfect dry run. ... And Musk said he was very much interested. Yes. The right call is absolutely Spacex.

Falcon is a good rocket, they didn't have to innovate much for that program. The real program is Starship, they had to engineer around Musk's unrealistic design ideas (but hey, he's not an engineer). It's not working. All those predictions about the Starship program, I want you to remember them when 2029 comes and Musk assures us the massively re-designed in-orbit refueling is on track to get done "by the end of the year".