I like to set three time goals for a race....floor, realistic and optimistic,
where the realistic goal is what I honestly think I am capable of on race day, floor goal is the least that I would be satisfied
with, and the optimistic goal is a stretch for me. Then, I base a race plan on the realistic goal. I would like to discuss
the development of a marathon race plan with that in mind.

An optimum marathon plan for most runners is based on running small negative
splits of 1-3 minutes. That is, the last half should be 1-3 minutes faster than the first half. That requires averaging 5-10
sec/mile slower than target race pace in the first half and an equal amount faster than target race pace in the second half.
Any slower than that in the first half and you set yourself up for having to run the second half at a pace that is too fast
to be practical, assuming that your realistic goal is an accurate one. Actually, an ideal marathon is run at dead even splits
and an empty tank at the finish line, but that is extremely difficult to do. Thus, a little conservatism in small negative
splits works best for most people.

Now, let's take an example. Let's assume that a realistic goal is 4 hours and
develop a race plan, accordingly.

A four hour marathon requires an average pace of 9:09. We know that we want to
start more conservatively than that. However, if we start very conservatively and average 30 sec/mile slower than that (9:39),
we will reach the halfway point in 2:06:31. Then to reach our goal of 4-hours, we would have to run the second half in 1:53:29,
or an average pace of 8:39. Since a race calculator says that a 4:00:00 marathon is equivalent to a 1:54:19 half marathon,
we would have to run a half marathon PR effort after already running a half marathon in order to finish under four hours.
That certainly isn't likely to happen and, if it did, then our "realistic" goal was not realistic, but conservative. We would
more likely crash and finish much slower than 4-hours.

OTOH, if we average 9:15/mile, which is 6 sec/mile slower than target race pace
in the first half, that would put us there in 2:01:05. Now the second half would only have to be 1:58:55, or an average pace
of 9:07 and negative splits of 2:10. That's certainly doable.

OK. So, how do you translate this to a race plan? You could simply try to run
5-10 sec/mile slower than target race pace each and every mile of the first half and the same amount faster each and every
mile of the second half. However, that is probably too simplistic for most races and it doesn't allow latitude for adjusting
to how you are really performing on race day. I prefer to break my race down into approximate thirds, sometimes fourths, to
allow a little more flexibility to adjust my race plan as the race unfolds, based on how I'm really performing. I start a
little more conservatively in the first third and try to average 10-15 sec slower than goal pace for the first 8-10 miles.
Then, depending on how fresh I feel at the end of the first third, I try to average race pace or a little faster for the second
8-10 miles. That should bring me to about 18-20 miles needing to make up about 1-2 minutes. I then push as hard as I can for
the last third or fourth of the race.

Let's take the 4-hour example. A 9:21 average pace (12 seconds per mile slower
than race goal pace) would put us 2 minutes behind target time at 10 miles. Then, if we can average 9:09 (target pace) for
miles 11-18, we would have to run the last 8 miles at an average pace of 8:56 to hit the realistic target of 4:00 for the
race. That's very doable. Of course, if it feels as though we have to work fairly hard to run 9:21 for the first 10 miles,
we should slow down and fall back to our slower "floor" goal. On the other hand, no matter how comfortable 9:21 might feel,
we shouldn't be seduced into running any faster in the first third. We should save it for the middle and end phases of the
race. We would let ourselves run a little faster than planned in the middle and, if we are still pretty comfortable at 18,
really go for it after 20 miles and push for our "ceiling" goal.

One reason for running the first third a little slower than the rest of the race
is that the crowded conditions at the start of many races make it impractical to start out at a pace close to target pace
for the first mile or few, so we might "lose" some time. However, it's a mistake to try to make up that slow start either
in the earliest miles that we can or by waiting until the last few miles of the race. It's better to spread the "makeup" over
all of the rest of the race with the emphasis on the middle miles. I would not suggest running any mile of the first half
faster than target race pace to "recover" that "lost" time.

One caveat to all of this. A race plan and the actual mile-by-mile splits also
have to be adjusted to account for course terrain and wind conditions. The goal should be for negative split "effort", which
won't necessarily be precise negative split times each and every mile. In fact, on some race courses with a lot of downhill
in the first half, like Steamtown, a negative split effort will very likely still result in positive split times. In that
case, negative split times would mean that we ran a damn good race, but probably not an optimum one. Course terrain and wind
are two factors that greatly complicate race planning and execution and make them even more inexact.

Jim2