2018 Madison Mini HM Predictions
Still not quite sure where this one is going to go yet. I still feel like I don't have a totally great idea where my fitness is, hence why it's a good reason to run this race prior to Chicago to try and nail that down. But I do know that one of my favorite parts of racing is playing the prediction game. Using recent data to try and pre-predict the race outcome. My Dopey guess was probably one of my best to date with a cumulative time guess of 5:47:50 vs actual of 5:45:14. Let's see how close we can get this time around.
The EA Model
View attachment 344171
So the last few weeks I've been running my easy runs (EA and EB) almost exclusively by HR on my normal hilly route of 30 feet per gain per mile (so possibly similar hilliness to Madison Mini). Trying to average a 132 for EA and 138 for EB. As we all know, the temperature and dew point (and sun) can play games with pacing. So I focused on the HR and effort to find an appropriate pace for each day this summer.
-The blue dots on the above graph represent the relationship between my EA pace and the T+D for a given run.
-The ones with the black circle outlines are the most recent EA efforts.
-The dotted blue line represents a polynomial line of best fit.
-If my fitness was that of a 2:58 marathoner under ideal conditions, then the pink dots represent the EA pace to T+D relationship using this chart:
View attachment 344148
-If my fitness was that of a 3:02 marathoner under ideal conditions, then the green dots represent the EA pace to T+D relationship.
I think what I glean from this graph is that my "average performance" (or line of best fit) closely resembles the pink dot theoretical values (maybe just a little lower). If dots show up below this line, then that means my fitness isn't as good as perceived. If dots show up above this line, then that means my fitness is better. The 4 most recent runs are split evenly between above and below. So the conclusion I would draw from this graph is that my fitness is close to a 2:58 marathon runner (under ideal conditions) assuming my EA effort is representative of my current fitness. There is a little bit of a wiggle above and below this line on any given day suggesting more of a range than a set value.
Given the T+D is now predicted at 129 at start and 133 at finish. Then that would place the adjustment somewhere around a 2-2.5%. I'll probably run a HM (per my Garmin) in about 13.16 miles (the average HM for me over the 15 I have done thus far).
So this would suggest a tight window on 2:58 of 1:27:27 to 1:27:53. I'd venture to guess using the end T+D is probably more accurate, so a 1:27:53. To get a reasonable range, I'll also use the green dots. That would give an upper prediction of 1:30:08.
So for this model, I'll split the middle and say a 1:29:00.
Garmin Model
Garmin produces a "VO2max" value and "race prediction". These are notoriously way off (predicts better than reality) for me in terms of actual performance and I've covered reasons before. I also now believe that a better term for the Garmin "VO2max" is "VO2peak". The max represents the highest achieved under ideal conditions. Summer days are typically not ideal conditions. So a better term would be VO2peak, which is reliant on current conditions as research shows VO2peak can be influenced by ambient temperature, dew, sun, etc.
I've shown previously that whatever my Garmin says is my VO2max is usually 6 points less in a VDOT race performance. So my current VO2peak as of this morning is 56. So then, I would use a VDOT of 50 to get a guess based on historical race performance to Garmin VO2peak.
View attachment 344175
A VDOT of 50 is equivalent to a HM of 1:31:30.
Two caveats to this model:
-Garmin's VO2peak is based on current conditions. My last few runs have been a T+D of 150, 152, 135-143, 152, and 154. So since the race is at a predicted 129-133, it would seem the T+D is going to be somewhat different than my VO2peak would have been influenced by. If I adjust the 6:58 pace down by 2.5% (difference between a 5% adjustment at T+D of 150s and 2.5% adjustment at T+D of 130s), then the predicted pace would be a 6:48 min/mile. So, instead of a 1:31:30, this would predict a 1:29:13.
-The "correctness" of the VO2peak is heavily reliant on my weight being accurate. I haven't weighed myself since May-June. So, I don't know if the currently inputted weight of 164 is accurate. I'd guess I'm somewhere between 157-167. Using the calculation:
Absolute VO2max = (weight x VDOT) / 1000
We can figure out that a 164 weight and 50.0 VDOT is a 3.7 liters/min absolute VO2max. Using a weight of 157 instead with a 3.7 liters/min, you get a VDOT of 52.2. And using a weight of 167 instead, you get a VDOT of 49.1. So if we use the 1:29:13 as a value for 164, then that gives us a range of 1:33:00 (using 49.1 VDOT (or 167 pounds)) to 1:28:10 (using a 52.2 VDOT (or 157 pounds). I'm probably somewhere between 157 and 164 based on how my clothes are fitting.
So between 1:28:10 to 1:29:13 would be an adjusted guess based on the two caveats (current temp vs race temp, and possible difference in weight).
So for this model, I'll split the middle and say a 1:28:40.
Using a Recent Hard Run Model
Simple enough. Look back at the last hard run.
Wed, 8/11/18
"Flat" M Tempo - 8 + 5 miles (TDAGP- 6:58, 7:01; AP-6:58, 7:12; GAP-6:55, 7:08) *8/8 + 3/5
T+D was 130-136. So nearly the same temps. I held a 6:58 min/mile for the first 8 miles with an elevation gain around 22 ft per mile. That suggests a 6:41 min/mile for the HM Tempo. With a 13.16 HM, that would be a 1:27:53.
So this model suggests a 1:27:53.
*********
So,
EA Model = 1:29:00
Garmin Model = 1:28:40
Hard Run Model = 1:27:53
I'll go with a 1:28:50 (splitting the EA and Garmin).
Final Predictions:
1:20:59 -
@SarahDisney
1:23:45 -
@bovie
1:24:27 -
@Capang
1:24:42 -
@Bree
1:25:13 -
@KSellers88
1:25:30 -
@Chaitali
1:26:10 -
@Jules76126
1:26:35 -
@FredtheDuck
1:26:52 -
@canglim52
1:26:56 -
@ZellyB
1:27:15 -
@opusone
1:27:30 -
@MissLiss279
1:27:49 -
@steph0808
1:28:12 -
@mrsg00fy
1:28:28 -
@pixarmom
1:28:50 -
@DopeyBadger
1:28:58 -
@jennamfeo
1:29:11 -
@rteetz
1:29:26 -
@TeeterTots
1:29:58 -
@PCFriar80
1:30:03 -
@sourire
1:31:42 -
@FFigawi
1:31:47 -
@garneska
69 6 (1:39:06) - Gigi
I've won the last two predictions games (Dopey and Hot2Trot), so let's see if someone can knock me off the champ seat. Best of luck everyone!