And the question is about my pace. I ran a half on Sunday and then added 7 miles for my 20 miler. Granted I probably ran that half a bit faster than I should have for a training run, my overall pace for the 20 was around 9:15 (8:45 average for the half, and then I forced myself to slow down for the extra 7 to around 10min miles). I'm thinking I should aim for 9:15-9:30min/mi on the marathon? (I may also beg and plead at number pick-up to see if I can be moved to a POT corral bc I hadn't done a race before the cutoff and now I have 3 POTs. It was really, really nice not to jockey for space at the half on Sunday)
My advice, wait until at least 3 days prior to the event to solidify pace goals. The weather at a race, but even more so Disney's possible swing in temps from 20s to 80s, will dictate actual race day pace to a significant degree. So coming up with a solid pace plan this early could lead to unrealistic expectations come race day. Because you'll start imagining what could be now based on current data, and then not appropriately adjust those goals when adverse weather conditions come up.
With that being said, I would be very careful trying to evaluate training runs into race paces. Most runners are notorious for "racing" their training runs. So they get a false sense of possibilities.
For example, let us reverse engineer the 20 mile training run. The final average was 9:15 mm for 20 miles. According to my calculations, a 9:15mm LR pace is associated with someone who can run a 23:12 5k, 48:23 10k, and 1:46:45 HM. So do any of those predicted race times come close to the POTs you've recently run? Or do you have any reasonable belief that if you "A" raced right now that you could hit those times? If no, then it suggests to me that your 9:15mm long run was closer to a race effort than you'd want in training. There could be some worry that you blew your peak on a training run instead of saving it for race day. For someone who had those times above (23:12 etc) I would put their equivalent marathon effort somewhere around 3:48-3:56 based on training.
Now let's take the training run from the other angle and assume the 20 miles at 9:15 was a max race like effort (even though it very well may not have been). What we're doing is creating a bracket. The 3:48-3:56 is best case scenario based on actually being a 9:15mm LR fitness, and the opposite will tell us the low end assumption. That looks to be something like a 25:39 5k, 53:29 10k, and 1:58:00 HM. For that profile and based on the training, the marathon would be around a 4:13-4:22. So the bracket is 3:48-4:22. A pretty wide range based on a singular data point.
If you provide me the following, I can probably give you a more accurate range.
-Your three recent race results.
-The three highest training duration weeks and mileage during them as well (so I want time and mile). And when they occurred.
-Your three longest training runs in terms of time and miles.
-When looking at the time spent training on the three longest runs compare it as a % to the total time spent training that week. For instance, someone who did 2 hrs for the longest run and trained for 6 hrs that week has a 33% LR.
-How many weeks in total did this training plan encompass?
-Lastly, what do you want for your Disney marathon experience? Are you hoping for a maximal effort? How much risk are you willing to incur? In a theoretical world, if I said you had a 10% chance of 3:58 and 90% chance of 4:20 based on an aggressive pace strategy, vs a 40% chance of 4:08 and 60% chance of 4:12 based on a conservative pace strategy, which would you choose?