I sure hope Beagle pulls it through, but Gary and others have made a pretty convincing case about the long odds Beagle faced.

Some suggest it is worth launching cheap but risky missions, hoping a few would succeed. You need to distinguish though, between adequately developing and testing a system that can be made cheaply enough to then use on risky mission profiles, vs simply not spending enough money build a credible system at all.

If we had a lander design that stood a good chance of working, except that it would break if it came down on a rock or encountered high winds, then sending half a dozen of them at $50 mil a pop and figuring half would get lucky might well be better than sending one $500 million lander with rock avoiding-guidance systems.

But, we don't know that Beagle would have worked, even with perfect environmental conditions. It sounds like there was not enough money or time to fully test/simulate all aspects of the EDL sequence, or to fix some clear design flaws (like the heat shield angle issue). Most importantly, we have no telemetry to determine what went wrong. Same problem with Polar Lander -- for all we know nothing went wrong and it came down on the only tall sharp rock for miles, or maybe it never even survived entry.

So, if there is no signal from Beagle, it could be because of rocks, wind gusts, or some other bad luck that would only happen 50% of the time. Or, it could be that there was a systematic problem with heat sheild separation or line tangling or whatever that would happen 100% of time -- a design flaw, not a calculated risk.

Properly designing a cheap system -- even one that works only 50% of the time as long as you *know* it works 50% of the time and then sending lots of copies of that system on risky missions makes sense. Building a bunch of one-off systems that might all have 100% failure rates without doing the testing or telemetry to find out is neither faster, better, or cheaper.