[...] awareness of the effects of biases has done little to improve the quality of business decisions at either the individual or the organizational level.
An insidious feature of cognitive failures is that we have no way of knowing that they’re happening: We almost never catch ourselves in the act of making intuitive errors. Experience doesn’t help us recognize them.
This inability to sense that we’ve made a mistake is the key to understanding why we generally accept our intuitive, effortless thinking at face value. It also explains why, even when we become aware of the existence of biases, we’re not excited about eliminating them in ourselves.
I stumbled upon this paper, and even though it's two years old and oriented towards business decisionmakers, I think the above sections have some interesting implications for hockey analysis, especially for casual viewers who are willing to rely solely on "watch the game" as a foundation for their analysis. It's also a foundation for why scouts who rely on "watch the game" can be an extremely useful source of information, so long as there are steps in place to combat cognitive biases, but that's not really what were interested in. Just keep those in mind as we go forwards.
Cam Charron wrote this piece on "anchoring bias," which is when an observer sets their expectations early - sometimes unrealistically - and is reluctant to revise those expectations. Tyler Bozak is a phenomenal example of anchoring bias, because people assume he's an incredibly talented goal scorer after watching his first NHL goal. This can also happen in contract negotiations - by asking for $2M, Mark Fraser is attempting to make a 1.275 deal look like a bargain, when it may have been fairly close to his real expectations.
Joe Haggerty wrote the following on Charron's article:
Cam Charron attempts to write something about "anchoring bias" – an interesting topic that actually contains little about hockey and hockey players. He lost me with his elementary usage of former B’s winger Benoit Pouliot as an example, a player that’s on his fifth NHL team because of bad penalties, issues with several of his head coaches during his underwhelming career and extraordinary bad timing with mental mistakes at key points in games. Example: it was Pouliot’s failure to get the puck deep on a dump during a line change in OT of Game 7 against the Capitals two years ago, and that sealed his fate in Boston. It has literally nothing to do with the potential associated with his formerly high draft place in the first round. Those are the kinds of things that Charron can’t find digging through advanced stats(which absolutely have a place at the table for evaluating players), and the reason why he completely misses the boat on his point about Pouliot. If you can’t marry the stats with what you see with your eyes, or with your insight into particular player’s personalities, habits and flaws, then the statistics are just as useless as valuing players based solely on anecdotal evidence.
Here's a good example of why we should be on the lookout for anchoring bias: Haggerty believes the information that made the biggest impression on him is also the most significant. It's not that Pouliot's performance isn't good enough to overcome his penalties, it's that his penalties in Boston were so bad, he shouldn't be picked up despite of his strong possession statistics. I'm not a mind reader, and I don't know how he came to this conclusion, but I think there's a good chance that Haggerty's overvaluing the earliest information he had.
Haggerty also writes about the dangers of relying solely on anecdotal evidence, but relies on anecdotes of "mental mistakes" and key/clutch situations, like "penalties in the third in a one-goal game" (Only one of his penalties turned into a goal against in a loss. Unfortunately for Pouliot, that one was Troy Brouwer's OT game winner). The aggregate data also shows that Pouliot drew more penalties than he took as a Bruin during the regular season, but Haggerty says he noticed that, and a team with a bad PP should devalue drawn penalties, which both ignores that the Bruins had the 15th-ranked power play (basically the definition of "not bad," and the thought process doesn't seem to apply to weighting penalties taken less on a team with the 11th-ranked PK for the season).
This isn't really an attack on Haggerty, although he definitely shouldn't be so confident in assertion or subsequent tweets (I'd lean towards "sign Pouliot," given the two arguments). I think that, especially in the context of the earlier quotes, it's just an interesting case where anchoring bias may (or may not - I'm not a mind reader) have been at play, even though Haggerty was just reading about anchoring bias. You can't really trust your brain/memory to do good, large-scale information processing on its own, even when you're aware of cognitive biases. We're hard-wired to take short cuts, like disproportionately valuing early or impactful information. I think the biggest advantage that scouts (either pro or hobbyist) have over casual/off-the-cuff "watch the game" analysis is the ability to make structured notes and do a little more structured analysis. I guess I'm just trying to say that even though stats should marry qualitative analysis, it doesn't mean qualitative analysis is suddenly free from cognitive biases. We're still human, after all.
Anyways, here are some links:
An interview with Toronto Maple Leafs Director of Player Development Jim Hughes | Toronto Maple Leafs
I thought the comments on Blacker were interesting, and probably a little concerning.
Border Wars: NHL 2012-13 Penalty Minutes Per Game By Country - The Copper & Blue
Cutting edge analysis of international politics.
Stick Gretzky used to score 1,000th lifetime goal fetches $38,838 in auction - The Hockey News
Follow-up to last week's link.
Kukla's Korner Looking for Bloggers
Just an FYI if anyone here wants a good writing outlet.
Boston Bruins, National Hockey League of Nations
Puck Daddy Series Continues.
