[guardiangod]

Looks like 1 of the engines failed to re-lit during the landing.

At least this time 1 of the engines was functioning correctly all the way to the explosion. This is an improvement over SN8 where all the engines failed to re-lit properly.

[CobrastanJorji]

I've plotted a graph with these two data points and predict that it will successfully re-light all three engines in only two more tries.

[ben_bai]

The landing burn will relight 2 engines to slow down, then land on 1 engine.

SN8 was really close. Only a couple more seconds before the engine ate itself would do. :)

[_ea1k]

Uh oh, that means that in two more tries it will be relighting too many engines.

[gvb]

It is interesting and informative to compare SN9 to SN8.

SN8: https://youtu.be/ap-BkkrRg-o?t=6884

SN9: https://youtu.be/_zZ7fIkpBgs?t=704

With SN8, you can clearly see two engines lighting and burning properly until just before it (crash) lands.

With SN9, you can clearly see one engine lighting and burning properly, but the second engine did not light cleanly and did not stay lit.

[zaroth]

I don’t think that’s quite right. With SN8 they ended up with low pressure in the header tank and not enough fuel for the turbo pumps.

A so called “engine-rich” burn resulted, which is why there’s a flash of green flame.

Total thrust was much lower than spec, hence the hard landing even though both engines ignited.

[msandford]

I'm surprised they aren't trying to light all three and then shutting down whatever doesn't light as needed for the landing burn.

Both of the failed landing have looked like propellant flow problems rather than true engine problems but even so I'd love the chance to get it "right" by getting it "wrong" in that the wrong engines fired up but it didn't hit the ground too hard.