Amy entered a triathlon on Sunday in support of Hearing Dogs for Deaf People. She did really well, especially for her first actual triathlon, and I'm really proud of her. How well did she do? Well, you could look at the results page and try to find her in the Short Triathlon 30-39 age group. Except, you wouldn't find her there. Some sort of mistake managed to put her in the 50-59 (!) age group instead.

Amy sent off a quick request to the event organisers asking to be moved to the proper category. After all, the people who pledged money to Hearing Dogs in her support might want to be able to find her in the results. But here's the response she got:

Hi Amy

 I'am really sorry for the error I have made, but unfortunately I'am now
 unable to make any changes.
 I hope this will not put you off entering for the following year!

 My sincere apologies
 Michelle

Unable to make any changes? What kind of excuse is that? Amy is requesting that they correct a factual error, not some dispute about timing or anything like that. But it appears that they're simply unwilling to consider any request to change the results.

As an aside, let's have a quick look at how the results are published. At first glance, it looks like the results are reasonably formatted in tables on the results page. But if you look closer, you'll find that those tables are actually images. JPEG images, no less! The borders of each category seem to be each slightly different, as if somebody manually clipped them from screen snapshots of a spreadsheet or something. There is further evidence of this: If you look at the results for the Short 40-49 group, you'll see an image of an insertion point overlaying the name of participant 107! Also, on the full page that particular category stands out because it's somewhat narrower than the others, and also is a lot more blurry (as if the JPEG compression level were turned up too high). Another oddity is the Long Female 50-59 group, which is blurry for the first two participants and sharp for the third. How on earth does this happen?

I thought perhaps they published the results as images in order to discourage search engines from finding people by name. However they helpfully publish a PDF version of the results that will be perfectly searchable by Google. Actually, it occurs to me that they might have taken screenshots of the PDF, saved as JPEG, and used those for the HTML version. In any case, it seems like they put far more work into it than necessary by choosing the really hard way. Given that, I can appreciate their reluctance to do it all over again, but a broken process is hardly an excuse for refusing to correct incorrect results. Not impressed.