I think you should first actually define what you mean by great vs merely good.
That reminds me of a Charlie Munger quote that he would rather hire a guy with an IQ of 130 that thinks it is 125 than a guy with an IQ of 170 that thinks it is 190. The "smarter" guy in that case is a total disaster waiting to happen.
You probably don't need a great programmer anyway. Trading problems are quite simple to code if you have precisely defined what it is you want.
Last edited by Darthtrader4beta; April 19th, 2015 at 03:45 PM.
1) I feel at least from a non-financial industry, the great coders are the ones where the value $/productivity is superb. So they may be more expensive, say annualized $100-$125k, but they do the work that 3 average coders at $70k a year could do or maybe 5 off shore coders that are poorly managed. [And of course at a very minimum, they can actually do the job and problem solve for the actual guidelines, the cold isn't buggy, etc.]
2) I think the good coders and execute on a very narrow scope very effectively, but they can't design the architecture of the program being given to them, so if they pieces don't fit, they may not be very good at getting all the pieces to come together.
3) A good coder might be able to solve the problem if you define it exactly in the way he/she needs to see it, and once it's defined the way they like it, they'll solve it efficiently. But I've always gotten frustrated with them saying it can't be done, but re-framed just slightly, suddenly the same outcome is possible. Or worse, they want the tools all handed to them as opposed to finding which tools might be available to use, which pieces of code have already been written else where for basic defintions, etc.