...and to make them well enough that you'll prefer them and if someone else makes them, they'll prefer them.
I've gone rounds with chisel formats, first making them and just heating and hammering on a bolster, then coming up with a way with limited tooling to forge weld the bolster on so that it can stand on its own with a ferrule-less handle.
And then ultimately at this point, just forging the entire chisel out of round bar. These are not all that neat - they can be finished further later, they're just for me, anyway. 
The first ones are W1 steel, which is a steel that nobody is going to work with industrially, and it's almost like old cast steel, but not quite. it's very easy sharpening and capable of high hardness.
The bottom chisel is a test chisel in 52100.
These can both be had in flat bar and round bar, but I need round bar so that the bolster is integral - as in the whole thing is one piece forged out of one section of rod. 52100 has to be well into the upper hardness range or a chisel is no good from it. It's characteristics are high toughness when it's not hard, but that's bad for a chisel. A deflected knife edge can be straightened, but a deflected chisel edge enormously increases the cross section that has to go through wood and the result is not tolerable.
the heat treat process here allows managing hardness, but especially, pushing alloys to the top or slightly above the industrial ranges without increasing grain size. Which in this case, allows 52100 to behave like a steel that is in between where it will favor chipping or rolling easily. In this case, the 400F tempered hardness is around 63, and hardness out of the quench is 68/69.
it's also a steel that will not show up in commercially made chisels, and if it's specified as 59/60 hardness, it may make a good timberframing chisel in wet wood, but the result is a disaster in hardwoods.
W1 is easier to keep even in terms of its characteristics - if it's a point or two soft, it doesn't increase in toughness much. the objective for almost all woodworking tools is to bias failure by minor chipping if it's going to occur. Those types of failures leave the edge and what is left behind does not inhibit cutting. W1 requires a fast quench, though, very fast, to get to full hardness, and it's lack of toughness is enough that I just can't settle on it as the final winner in making chisels out of round bar. It makes a good chisel, but I don't know that it's any better than commercially available chisels. I don't think there is a commercially available chisel other than japanese chisels that will match the 52100 chisel shown above, though.
Can't say as much about knives- chipping in knives is a pain, so I can see the bias toward rolling (vs. chipping), especially in consumer knives where chipping can also be a sign of a knife easier to break if misused.