I'd like to start a discussion on the phases of training for the middle-distance races. I think everybody breaks down the year in their mind to some extent. It's natural to not want to bite off the whole annual training schedule from one day in the year to another day that might be as much as 12 months away (more if you're on a multi-year plan). Aside from this psychological trick, I believe there is also a physiological advantage to training in specific phases. Our bodies respond to stress by reaching a homeostatic point where it can respond to similar stress more easily in the future. There are many types of adaptations we make to many different kinds of stress. I don't think it's a far-fetched idea to think of one stress (e.g. weight training) bringing on an adaptation (increased myofibril nuclei density) so that the next adaptation (increased motor unit recruitment) to the next stress (speedwork) can be more effective. There's still a good amount up for debate such as: What should these phases be? How many phases should we employ? How long should they last? How much should they overlap, if at all?
Given that it's now August and us mid-d guys have the whole program of training in front of us with a mind to peak for races next summer, let's take a holistic look at the training philosophies of some of the more prominent coaches in history.
Most take a prudent approach and start the framework around the first major race. This is good, I think, to some degree, but only so far as your certainty in your phase priorities. For example, if I have 3 months to prepare for my first big race, my coach (who is also me) might recommend (working backwards) an intense speed phase of 2 weeks, a pacework phase of 2 weeks, a hard interval phase of 4 weeks, a LT/hill phase of 4 weeks, and a base-building phase for whatever time is left. (This is actually a common prescription, not in the details of what training to do, but in the timing.)
Using the above example (which you might find looks something like what Lydiard uses), you could go one of two ways. One conclusion might be to say, "Ah well, it looks like 3 months really isn't enough time to train for this race coming up. I may as well move my goals back a bit." Three months may in fact not be ideal, but whenever such timelines pop up, I recall a conversation I had a few years back with the husband and coach of Constantina Tomescu-Dita (either her or Luminita Talpos--either way, they're both pretty solid Romanian distance runners). He told me that he could take a previously elite athlete and get him or her back into very close to race shape in about 2 months assuming they weren't too fat. Being in Boulder at the time where I was working a shit job and running more than I had ever run in my life for the sake of races that were still 8-9 months away, this boggled my mind. Looking back, I think h was right, if not a little drunk. (The conversation took place at my non-AC'd apartment where people frequently used the elevator as a urinal.) I saw the man's foretold phenomenon time and again throughout college with talented people coming off of injuries. Sure, in these cases phase-training might not be employed in the classical sense. That doesn't mean it doesn't work. That also doesn't mean 97 percent is the same as 100.
The other conclusion one could take to the lack-of-time problem would be a bit more inventive if not a tad less pragmatic: We mold the phasing groundwork to make it work for us. Here we are straying dangerously away from the black and white divide into the world of grey. There are many who believe, myself included, that more grey typically means less scientific foundation. That said, we make this venture in the name of variability (variability in athletes' lives, their race goals, and in the athletes themselves). In fact, I've done it myself to some degree with my own training plan. Before we start bending the rules though, let's set some rules out for bending.
Not to ignore the Scandanavians, but I'm going to start with Lydiard. This is from his book Running the Lydiard Way, and I think it provides us with a good starting point. Note the framework he employs as far as time of season is concerned.
[NOTE: I'm going to keep on the phase theme for a little while (hey, it's only August, right?), so if you've got any other famous examples, let's have them.]
800 and 1500 meters (men)
I) For as long as possible:
M-Easy fartlek running, 1 hr
T-Long aerobic running, 1.5 hr
W-Time trial 5k
R-Repeat T
F-Easy fartlek running, 3/4 hr +hills;
Sa-Time trial 10k;
Su-Aerobic running, 1.5 hr+
II) For 4 weeks:
M-Leg speed 1x120m
T-Long aerobic running, 1.5 hr
W-Hill springing, steep hills, or steps, 1 hr
R-Easy fartlek running, 1 hr
F-Repeat M
Sa-Repeat W
Su-Long aerobic running, 1.5 hr+
III) For 4 weeks:
M-Repititions 15-20x400m
T-High knees, long strides, running tall, 2x100, fast relaxed, 4x100
W-Easy fartlek running, 1 hr
R-Repititions 15-20x200m
F-Leg speed 10x100m
Sa-Time trial 3k or 5k
Su-Long aerobic running, 1.5 hr
IV) For 4 weeks:
M-12-14x100m windsprints (one every 200m*) *Book says 100 which makes no sense
T-Easy fartlek running, 1 hr
W-Time trials 200 and 600
R-Long aerobic running, 1 hr
F-Fast relaxed running, 6x100m
Sa-RACE 800m or 1500m
Su-Jogging, 1.5 hr
V) For 1 week:
M-20x45m windsprints (one every 100m)
T-Easy fartlek running, 1 hr
W-Time trials, race distance
R-Easy fartlek running, 3/4 hr
F-Relaxed striding, 6x400m
Sa-RACE 400m or 800m
Su-Jogging, 1.5 hr
VI) For 1 week:
M-16-45m windsprints (one every 100m)
T-Easy fartlek running, 3/4 hr
W-Time trials, 100m and 400m
R-Jogging, 3/4 hr
F-Jogging, .5 hr
Sa-FIRST IMPORTANT RACE
Su-Jogging, 1.5 hr
VII) Continuation of racing
M-Easy fartlek running, 1 hr
T-Relaxed striding, 6x200m
W-RACE or 12x100 windsprints (one every 200m)
R-Repeat M
F-Repeat T
Sa-RACE or Time trials
Su-Jogging, 1.5 hr
remember. It did, however have a 
