On this glycogen burning vs fat burning - I'd like to be as sure as I can on what happens with as little really detailed science as is necessary to suggest how runners apply it in practice. Does this make sense:-
- Pfitzinger (p45) says 'if you do a great job of loading, you'll have just enough glycogen for the marathon'. He says this after outlining that 'racing' the marathon you get 75% to 90% of your energy from carbs, the rest from fat.
- I interpret this as meaning that if you get everything else right you can still 'race' the 26th mile (and .2 of the 27th) using the same ratio of glycogen to fat ie you are never running solely off fat
- he also talks very coherently (pps 120-121) about how, short of world class level where 'genetics and training put them on a higher plane', you need to 'budget' for a c 60/90 second positive split between the 1st and 2nd 1/2 to allow for running economy to falter by c 2% as you start to recruit less efficient fast twitch fibres once the slow twitch fibres are fatigued.
Collectively those things suggest to me that if you couple them with a 1/2 marathon c 3-5 weeks beforehand, the knowledge that you've done a suitable number of suitable long runs at a suitable pace/effort (more on which below) and you don't set a target that your 1/2 marathon pace suggests is 2-3 minutes just out of reach even on a perfect day, and if you pace yourself accordingly on the day, you can be confident that you can get to very far in the race before things get really hard.
Long runs - sure there are some folk who specifically use flat roads to mimic the FLM route, even using th ecourse itself, and thus, sort of, they get exactly the same stimulus/discomfort level in training as the same pace will incur in FLM. Many though do many of their long runs off road; over hilly courses; over distances that they don;t accutaely know the distance of; the day after either a hard training session or XC race; with a slight hangover; possibly starting at least partially depleted and/or dehydrated thru a combination of the prevoius factors.
So in any of those circs, and especially if several apply, a given pace, say target marathon pace + 30 secs per mile, is going to feel like flipping hard work compared to what it will be like when they are tapered, loaded etc. This, coupled with a load of egs of high achieving Brits, Kenyans, wherevers, who don't appear to execute their long runs with a hugely detailed level of pace precision, makes me wonder whether in the real world there is perhaps a limit in planning the pace of long steady runs beyond which the number of variables makes the maths of little practical benefit.
But if I'm just being intellectually lazy here, I'd welcome any comments as they'd probably be useful to other guys who have tuned in to date. Thanks