{"id":42112,"date":"2026-08-19T22:27:19","date_gmt":"2026-08-19T20:27:19","guid":{"rendered":"https:\/\/www.graviton.at\/letterswaplibrary\/nrcd-an-open-database-of-collegiate-running-with-unified-performance-standardization\/"},"modified":"2026-08-19T22:27:19","modified_gmt":"2026-08-19T20:27:19","slug":"nrcd-an-open-database-of-collegiate-running-with-unified-performance-standardization","status":"publish","type":"post","link":"https:\/\/www.graviton.at\/letterswaplibrary\/nrcd-an-open-database-of-collegiate-running-with-unified-performance-standardization\/","title":{"rendered":"NRCD: An Open Database Of Collegiate Running With Unified Performance Standardization"},"content":{"rendered":"<p><!-- SC_OFF --><\/p>\n<div class=\"md\">\n<p>I just saw the paper on <a href=\"https:\/\/arxiv.org\/abs\/2608.14776\">ArXiV<\/a> . It&#8217;s a dataset of US collegiate running club performances, the resulting <a href=\"https:\/\/github.com\/National-Running-Club-Database\/nrcd_xc_paper\/blob\/main\/output\/FINDINGS_EXPLANATION.md\">analysis on them<\/a>, and a software library for standardizing performances. They have several code repositories under the <a href=\"https:\/\/github.com\/National-Running-Club-Database\">National Running Club Database<\/a> which includes:<\/p>\n<ul>\n<li>a Python library, <a href=\"https:\/\/github.com\/National-Running-Club-Database\/nrcd\">nrcd<\/a>, for normalizing race results for comparison across outdoor performances with wind to performances on an indoor banked track, road performances, etc. It appears to account for course distance, elevation gain\/loss, and weather\/heat effects.<\/li>\n<li>the raw <a href=\"https:\/\/github.com\/National-Running-Club-Database\/national_running_club_database_public_dataset\">national_running_club_database_public_dataset<\/a>.<\/li>\n<li>The [code which produced their results](<a href=\"https:\/\/github.com\/National-Running-Club-Database\/nrcd_xc_paper\/tree\/main\">https:\/\/github.com\/National-Running-Club-Database\/nrcd_xc_paper\/tree\/main<\/a>) and [the findings](<a href=\"https:\/\/github.com\/National-Running-Club-Database\/nrcd_xc_paper\/blob\/main\/output\/FINDINGS_EXPLANATION.md\">https:\/\/github.com\/National-Running-Club-Database\/nrcd_xc_paper\/blob\/main\/output\/FINDINGS_EXPLANATION.md<\/a>)<\/li>\n<\/ul>\n<p>Some things I found interesting (this is just a sampling, you go read the [full doc](<a href=\"https:\/\/raw.githubusercontent.com\/National-Running-Club-Database\/nrcd_xc_paper\/refs\/heads\/main\/output\/FINDINGS_EXPLANATION.md\">https:\/\/raw.githubusercontent.com\/National-Running-Club-Database\/nrcd_xc_paper\/refs\/heads\/main\/output\/FINDINGS_EXPLANATION.md<\/a>) yourself):<\/p>\n<ul>\n<li>Teams with at least one athlete who raced 4+ times were 2-3x more likely to crack the top 15 at nationals than teams without one (23-39% success rate vs baseline). 60-80% of top 15 teams had an athlete with 4+ races.<\/li>\n<li>Men&#8217;s teams with a longer gap between their first race and nationals (i.e., started racing earlier) had significantly better finishing ranks (r = -0.283, Bonferroni-corrected p = 0.041). This didn&#8217;t hold up for women&#8217;s teams after correction.<\/li>\n<li>Single biggest predictor of an individual&#8217;s improvement rate was &#8220;experience level&#8221;. (races \u00d7 season duration) at 21%, followed by how many &#8220;bad races&#8221; (a race worse than the previous one) an athlete had, at 17%. Basically race more, race consistently.<\/li>\n<li>When testing the standardization tool, the fully weather\/terrain-adjusted &#8220;standardized&#8221; times actually predicted improvement slightly worse than just doing distance conversion alone (90.4% vs 93.1% R^2). Their theory was that conditions tend to get more favorable as the season goes on, so raw times naturally look like &#8220;improvement&#8221; partly because of the weather, and removing that weather effect (which is more the point of the tool) makes it a worse predictor of the raw number even though it&#8217;s arguably a more honest fitness signal.<\/li>\n<li>They checked their model for gender bias and found it performs comparably for both (94.5% R^2 women vs 90.4% R^2 men). This is not my paper, data, etc. I just found it on Arxiv and compiled the information.<\/li>\n<\/ul>\n<p>There&#8217;s not much easily accessible data like this on cross country running. As the authors said, it&#8217;s all on websites that don&#8217;t support bulk download. So Normally people just stick to the Riegel and Cameron formulas for comparisons, so it&#8217;s nice to see people looking into these other factors.<\/p>\n<\/div>\n<p><!-- SC_ON -->   submitted by   <a href=\"https:\/\/www.reddit.com\/user\/dpfens\"> \/u\/dpfens <\/a> <br \/> <span><a href=\"https:\/\/www.reddit.com\/r\/datasets\/comments\/1vsxtv4\/nrcd_an_open_database_of_collegiate_running_with\/\">[link]<\/a><\/span>   <span><a href=\"https:\/\/www.reddit.com\/r\/datasets\/comments\/1vsxtv4\/nrcd_an_open_database_of_collegiate_running_with\/\">[comments]<\/a><\/span><\/p><div class='watch-action'><div class='watch-position align-right'><div class='action-like'><a class='lbg-style1 like-42112 jlk' href='javascript:void(0)' data-task='like' data-post_id='42112' data-nonce='9de69db8d5' rel='nofollow'><img class='wti-pixel' src='https:\/\/www.graviton.at\/letterswaplibrary\/wp-content\/plugins\/wti-like-post\/images\/pixel.gif' title='Like' \/><span class='lc-42112 lc'>0<\/span><\/a><\/div><\/div> <div class='status-42112 status align-right'><\/div><\/div><div class='wti-clear'><\/div>","protected":false},"excerpt":{"rendered":"<p>I just saw the paper on ArXiV . It&#8217;s a dataset of US collegiate running club performances,&#8230;<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[85],"tags":[],"class_list":["post-42112","post","type-post","status-publish","format-standard","hentry","category-datatards","wpcat-85-id"],"_links":{"self":[{"href":"https:\/\/www.graviton.at\/letterswaplibrary\/wp-json\/wp\/v2\/posts\/42112","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.graviton.at\/letterswaplibrary\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.graviton.at\/letterswaplibrary\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.graviton.at\/letterswaplibrary\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.graviton.at\/letterswaplibrary\/wp-json\/wp\/v2\/comments?post=42112"}],"version-history":[{"count":0,"href":"https:\/\/www.graviton.at\/letterswaplibrary\/wp-json\/wp\/v2\/posts\/42112\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.graviton.at\/letterswaplibrary\/wp-json\/wp\/v2\/media?parent=42112"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.graviton.at\/letterswaplibrary\/wp-json\/wp\/v2\/categories?post=42112"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.graviton.at\/letterswaplibrary\/wp-json\/wp\/v2\/tags?post=42112"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}