Monday, April 13, 2015

Exactly how many runners did I pass?

In my last post where I recapped my 2015 Shamrock Shuffle race, I mentioned that I passed "tons of runners" over the last couple of miles of the race. I thought it would be cool if I could actually quantify how many runners I passed (and how many passed me) from the 5k mark to the finish line (8k mark) of the race. Luckily, the individual results show each runner's time at the 5k mark and their finish time. So, after painstakingly looking at over 300 runners who finished around my time, I have the results of my analysis:
Shamrock Shuffle 2015:



Number of runners I passed after 5k mark:

106
Number of runners who passed me after the 5k mark:

0
My place at 5k mark:

516
My place at the finish line (8k):

410

So my "tons of runners" translates to 106 runners passed in just over over 11 minutes - about 10 per minute. I went way down the list to see if anyone had passed me over the same stretch and could not find a single person. This info shows just how big a negative split I ran. However in the long run, the only time that matters is the one that displayed as I crossed the finish line. Lots of other runners who went out a lot faster than me then slowed down more than me, still beat me. Bottom line: While it's fun to pass other runners, you can still get beat by lots of faster "positive splitters" and "even pacers".

Anyway, I also saw that someone on reddit had downloaded the stats of every finisher of the 2014 Chicago Marathon into an Excel spreadsheet. So, using that data, I did the same analysis as above, except this time I calculated the number of runners I passed and was passed by after the halfway (13.1 mile) mark:
Chicago Marathon 2014:



Number of runners I passed after 13.1 mile mark:

481
Number of runners who passed me after the 13.1 mark:

117
My place at 13.1 mile mark:

2,127
My place at the finish line (26.2 miles):

1,763

The interesting thing with this is that I was down on myself for my "big" positive split during the Chicago Marathon, but I needn't really be. Of the runners who passed the half way mark at 1:32:52 or faster, I finished with a faster time than 481 of them. Only 117 runners who ran slower than me for the first half, had a faster time than me. That means that my positive split actually was decent when compared to how big other runners' positive splits were. I shouldn't get down on myself too much - I ran faster than most for the second half! Anyway, next year I hope to be one of those 117, which means starting even slower and finishing even faster!

2014 Chicago Marathon Finish
Next up: Ravenswood Run in 11 days!

19 comments:

  1. Geez. How did you do all that math? And I looked at the Marathon spreadsheet and couldn't figure anything out except that I ran even splits the whole time. 2:08 at the half and 4:17 to finish. Hmmm.

    ReplyDelete
    Replies
    1. According to my analysis of your times, you passed 4,429 runners and were passed by only 164. Nice job! Goes to show how smart you were and how few runners run even splits in the marathon!

      Delete
    2. Thank you for this! I have no idea how you figured it out, but I'm happy you did!

      You're not running the Lakefront 10 miler on Sat? I was hoping to meet you!

      Delete
    3. It involves a little database wizardry! It is a simple query once I loaded it in the database. You could do some sorting in Excel based on first half time and finish time then do a count. Unfortunately, I don't think I'll run the Lakefront 10 miler. Maybe we'll run the same race soon! :)

      Delete
  2. And our analytics wizard strikes again! Can't wait to see you at the Ravenswood 5k!

    ReplyDelete
    Replies
    1. Thanks. I hope all my analysis is correct! See you at the Ravenswood Run! :)

      Delete
  3. Holy cow that's amazing. I always wonder how many people I pass.

    ReplyDelete
    Replies
    1. Thanks. If the race reports splits, it can be done.

      Delete
  4. I love your analytics. ;)

    Impressive #s, but what about adding in the folks that started before you, but you passed during the run? By looking at time of day.. I have a feeling your number would be even higher :) but- on my phone at least, that info is not available... So maybe it is not possible.

    ReplyDelete
    Replies
    1. Thanks! Yes, it is possible to add the folks that started before me that I passed during the race. That would give me my true "kill rate". The Shamrock results include time of day for start, 5k and finish, but it is only available by selecting an individual runner (i.e. not in the "category" view). That means I would have to comb through 300+ individual records again. I should have noted those times the first time around! :)

      Delete
  5. This is the coolest analysis ever!!! Way to tear it up! You should get a shirt that says "Eat My Dust" =D

    ReplyDelete
    Replies
    1. Ha, ha. I should get a shirt like that! "Eat my Dust" would be printed on the back and I could somehow change it to "I'm Eating Your Dust" if I'm forced to slow down or walk!

      Delete
  6. So cool! You are an analytics wizard! You should make an app for this. $$$ :)

    ReplyDelete
    Replies
    1. Thanks Sara! Yes, I could definitely make an app like for this. However, would anyone pay for it? :)

      Delete
  7. One of the things I really love about your blog is the mathematical analyses you do. Love your charts and your number crunching. It's really fun to read! Thanks for taking the time to share your analyses with us, Pete!

    ReplyDelete
    Replies
    1. Thanks Beth. It's fun for me to over-analyze my races! Glad you enjoy these posts! :)

      Delete
  8. Oooh, I love some math and spreadsheets! You should be running Ragnar so you can keep track of your kills! Though I suspect you'd run out of room to mark up the van with those :)

    ReplyDelete
    Replies
    1. Thanks. Do you have to erase a "kill" if they pass you? Can you double count them if you pass them twice?

      Delete
  9. Love this! So cool you actually sat down and looked through the results to get those numbers.

    ReplyDelete