No it doesn't, it just makes it less likely for their candidate to win. Also when it's split 50/50, all that means is that half of the population of that district loses, and is going to be unhappy, and basically unrepresented. Which is why it's better to have a 75/25 district or 90/10.
Actually no Kara, when you get such a lopsided demographic ratio, there is almost no point to have a general election then.
----------------
Districting is very simple, but also very hard. Geographic maps don't do justice to what is going on. Most districts will look somewhat weird. A primer
1) The perfect district is circular if population is distributed evenly. It isn't. 2
2) Geography and state boundaries get in the way.
So, basically, say 650K people per Congressional district. You have 11 such districts to make. Now, how do you do that in say, Ohio, while keeping to the 1 person 1 vote requirement in the constitution which means that you need these districts to be very close to the same size in pop.
There are a good many constraints. It is not an easy problem, and if you think it is you simply don't understand the problem of filling in a matrix where the row and column totals are known but none of the internals. That is a problem statisticians at Harvard struggle with.
In general, in the past, politically led state redistrictings have resulted in non-competitive seats (bad for democracy), deals between groups (as in GA, give the black dems 3 seats and they'll let the reps drawn favorable districts for the others) to Tom Delay's district in Texas (ugly).
Also, keep in mind its not very nice to split recognizable communities up in say, congressional disticts. Doesn't make a whole lot of sense, but guess what, political districting do this all the time to split up a minority demographic group
I cannot say much else, other than to say that I've floated some ideas about national districting solutions for congressional districts that relies on census data and a somewhat advanced nodal model. But states, not the fed, control their districting processes.