Believe in something


Ok, I’ll admit, this is a slightly contrived post just to have the excuse to include this image in the blog. However, there’s been a fair bit of speculation about whether Colin Kaepernick will ever play again in the NFL after starting the trend in the NFL of taking a knee during the national anthem in protest against racism in the US. The suspicion is that he is being overlooked in favour weaker quarterbacks by club owners as retaliation for his protest, but does the evidence bear this out? Well, various informal statistical analyses, comparing Kaepernick’s most recent performances against those of other quarterbacks, definitely support this point of view.

Meanwhile, here’s another statistic: Nike sales spiked by around 31% in the days after the Kaepernick ad campaign was released. It would be nice to think that Nike would have ran the Kaepernick campaign even without doing their own market analysis based on surveys and focus groups to collate evidence that it would be viable financially as well as morally. Guess we’ll never know though.


3 thoughts on “Believe in something

  1. I wonder whether the spike in sales has continued past a few days and if it was really that significant for Nike? There seemed a lot of people on social media burning shoes / saying they’d never buy Nike again … that seems to me a potentially more permanent change in consumer behaviour. So will those losses outweigh the gains, over time? And how much difference does a campaign like that really make to a huge corporation like Nike? There are a few possible sources for information. The company’s own corporate press releases will provide some sales figures, and narrative. I looked at the latest quarterly earnings release and the transcript of the call which the companies had with investors: here: …. lots of detailed information but you had to get to the questions before anything about Colin Kaepernick came up. It’s almost as if it wasn’t that big a deal by the time they got to the earnings call. Or maybe they didn’t have enough data. The results will be more detailed over a year. If I was looking for that sort of information in future I would go to the SEC’s website to look at Nike’s 10-K. In the US companies above a certain asset size or number of shareholders must file an annual report called a 10-K with the Securities and Exchange Commission. Unlike a glossy shareholder report, the form 10-K provides very detailed information and narrative about a company’s performances and reasons for it. It’s a more warts and all picture of a company and companies are obliged to be quite detailed about what is going on for them. Look out for Nike’s 10-K.

    Is any of this useful? For me, it’s interesting how there has been a massive explosion in the amount of available information – much of which people don’t know about. I recall going to an interview about 15 years ago, at eBay. When asked about what challenges I thought eBay would faced, I quoted to the interviewer the challenges which eBay itself had already listed in its 10-K. The interviewer’s jaw dropped. He asked me how I had access to what he thought was confidential information. This HR guy had no idea what his own company was publishing for all to see. So at the very least this is a tip for future interviewees ;).

    1. Thanks Nity. Following on from the Kaepernick episode, some reports suggest that sales fell back to pre-campaign levels:

      On the other hand, suggestions that a fall in Nike stock price fell because of the campaign are likely to be unfounded:

      But I guess your main point is that it’s surprising how much information is publicly available, and your eBay anecdote is enlightening. Maybe you should have just told the HR guy that you’d successfully bid for and won the confidential information on a well-known auctioning web site. Probably wouldn’t have done any harm: you eventually ended up at Smartodds and not eBay 😉

      Anyway, from a statistical point of view, it’s interesting the way information value has changed. When I first started studying statistics seriously, there was an emphasis on techniques that were suited to the analysis of very small amounts of data. These days it’s exactly the opposite: much research is dedicated to developing techniques that can handle immense amounts of data.

Leave a Reply