Negative splits are the time proven way
to best run a marathon. All "experts" and most experienced marathoners that I
am aware of recommend targeting negative splits of 2-3%.
The problems with a race plan that is
based on positive splits, even small ones, are: (1) faster consumption of glycogen, which means that your body becomes more
dependent on fat for fuel earlier in the race; (2) you reach your AT/LT earlier in the race, so a greater portion of the race
is spent running anaerobically, i.e., the wall arrives earlier and harder; and (3) it allows for less margin of error in case
you miscalculated your ability on race day or any of many other variables bite you in the butt. You wind up running a greater percentage of the race in the less efficient mode, which exacerbates your
late race "decline" and can lead to a hard crash.
The ideal race of any distance is to
run dead even splits and finish with an empty tank, which means that you would not be able to continue to maintain the pace. It isn't too difficult to learn to do that in short races, like 5-10k. However, the longer the race, the more difficult it is to do because more variables come into play. It
is extremely difficult to do in a race as long as a marathon.
Since it's so difficult to run even splits
in a marathon, it's usually better to err slightly to the conservative side than to be a little overly aggressive. Not by much....like an average of 5-10 sec/mile over the first half....which is 1-2%
for most runners.
I've tried it both ways. A little conservatism (5-10 sec/mile) in the first half worked better for me than the same amount of aggressiveness.
We are almost splitting hairs. If you run a 3:30 marathon with the first half 2-3% faster or slower than the second half, the splits would
be 1:44 and 1:46. That's about as close as most people will ever come to even
splits. However, coming that close to even splits does require a very accurate
assessment of your ability and target time.
Here are some references from the experts:
In their book, “Better Training
for Distance Runners”, Martin and Coe say, “....deviating from the mean race pace by more than + or - 2% is
metabolically more costly than remaining within this window.” This recognizes
that pacing that close is essentially even splitting. They also say, “Ideally
you should run the first portion of the race a little slower than your expected mean pace, thereby conserving carbohydrate
supplies. Later on....adequate supplies will be available to provide energy for
pace increases as well as a strong finish. This is the concept of “negative-splitting
In “The Lore of Running”,
Tim Noakes says, “The golden rule is that the effort of both halves of the race must be as equal as possible. Never listen to those who advocate running the first half of any race faster so that
you will have spare time to cushion your reduced pace in the second half. In
fact, the fast early pace is the very reason for the fade in the second half.....My personal preference, gained from many
unhappy experiences, is to aim to run the second half of any race slightly faster than the first. I have found that it is always preferable to speed up in the second half when others are slowing down. It gives you the impression that you are running faster than you really are, and the
mental lift of passing others is great.”
In “The Competitive Runner’s
Handbook”, Glover says, “What’s a good starting pace? Statistical
studies show that runners starting at more than 2 percent (about 10 seconds a mile) faster than their average pace slow significantly
over the last 6 miles compared to those who run with even or negative splits......Some marathoners benefit from starting slow
(up to 10 seconds a mile) for the first 2 or 3 miles and then picking it up. Others
like myself prefer to start slightly faster (no more than 5 to 10 seconds) than goal pace in order to run more by equal effort. Some prefer running the first half a minute or two slower than they will run the second
half (negative splits). If you go out a bit slow, you have adequate time to make
up for it. So it’s better to err on the side of caution.”
Their rationale differs somewhat. But, they all generally conclude that slight negative splits result in a better
performance than a slightly aggressive race most of the time for most runners. It
really does come down to how accurately you have assessed your true ability and set a realistic race goal. If you accurately pegged your race goal, then it doesn't really matter if your splits are positive or negative
as long as they are no more than 2-3% apart.....that close is essentially even splits.
However, if you set an ambitious race goal, then an aggressive race plan, even one of just 2%, will extract a big cost
compared to a 2% conservative one. OTOH, if you set a conservative goal, then
a slightly aggressive race plan will give you a better chance to beat it more, since it will be closer to your true ability.
Then, there is the effect of course terrain
on pace to consider.....this stuff ain't an exact science no matter how much we theorize it.