My recent minor detonation in the Frankfurt marathon has me looking at my training data more objectively again. I got caught up in shorter race performances and mission-creep set in on my lead-up to Frankfurt.
Plotting out the heart rate based marathon estimate (assuming a marathon heart rate of 171 bpm and a linear relationship between speed and heart rate) you can see how well the HR data is at setting an (upper) limit to marathon performance. Covering 10 marathons over 2.5 years this is rather convincing and I shouldn’t let mission-creep and shorter distance racing sabotage my marathons in future! That said it was (for 7/10th anyway) good fun to run on feel and try to push the pace a little… quite often you’ll have to pay the piper though!
With Malaga now 3 weeks away I’ll attempt to keep the pace up a bit and the mileage in the 160-180k region at minimum. The trajectory is looking good, I just need good conditions and a cool head to execute the sub 240 plan B … and ideally Axel or Matthew to run with me 
For those that want to read more into the plot here are some details on the individual marathons with the half split difference:
(10) 2:40:50 Frankfurt +2:00 great conditions, calf pain from 32k
(9) 2:44:26 Edinburgh +2:26 gave up 7k out
(8) 2:42:08 London +0:51 great conditions, dull pain throughout
(7) 2:44:14 Malaga +0:16 cool start, hot at end, feeling ‘good’
(6) 2:45:01 Frankfurt +0:07 great conditions, right on the edge
(5) 2:49:12 Reykjavik +2:00 solo last 10 miles
(4) 2:53:04 Liverpool +3:00, windy, undulating, pulled hamstring
(3) 2:59:00 Paris -1:30 good conditions, strong finish
(2) 3:02:21 Pisa +1:30 great conditions (400m short!)
(1) 3:13:39 Amsterdam +3:39 warm, very sore from 28k