The book looks at how clubs value players in sport.   The characteristics that are valued by clubs often did not reflect the characteristics that lead to results.  Thus characteristics that were valuable (such as on base percentage) were undervalued.

Clubs often overvalued things that actually had impact on performance like whether or not a player appeared confident or had a good looking playing style.  The statistical data showed that these measures were flawed.   Leading some players undervalued.   Such as having emotional problems, being physically small or being overweight.   But if the data shows their actual performances these perceived weakness are irrelevant the actual performance is what counts.

Instead the Oakland As looked at statistical data.   This eliminated the inaccurate and flawed judgements of individuals.   It also eliminated biases such as the Availability Heuristic which lead scouts to make judgements based on recent performances.   Or from limited viewing in a small number of games.   Where as the long term statistical data is far more revealing of what is actually going on.  This also tackled the fact that people tend to over value their own experience as if it is the most valid.  Taking emotion and bias out of the equation with statistics eliminates these biases.

This insight lets us understand that perhaps the way we look at people is biased and may not lead us to getting the best options based on our own biases.   We might unfairly discriminate against an older worker, someone who is quiet, or someone who seems overweight when in fact this bears no relation to their performance.  We may therefore unwittingly miss out on the best options in favour of options that appear good on the surface.   We don’t have statistical data like baseball clubs do but we could perhaps place a greater weight on past performance as an indicator of future performance.   If we want an innovator for example we should probably look for someone who on close inspection has implemented lots of innovations in their career history.  Not someone who talks about innovation at interview and sounds good at interview but can not back it up with a proven history of action.   We might also consider that people who are not sociable or do not interview well may in fact be good employees but this fact is disguised by the process we choose to use to select them. 

(Summary of the book [above] and my thoughts on it).

It is the process that matters not the outcome.   People tend to overvalue outcomes and disregard process.   But if a player gets the desired result by doing the wrong thing this is actually something that will hurt you in the long term. (eg running when they shouldn’t but getting way with it)   But a player who fails when doing the right  thing will befit you in the long term (getting caught out when doing a statistically safe move).   The fact that luck and factors out of your control means that something that does not work out should not necessarily lead to a change of course.   Instead you need to focus on the correct process and understand that luck means that even the correct process won’t work all the time.  But sticking to the correct process will lead to greater success in the long term.   P 146.  (paraphrase and own notes on idea)

It is very hard to teach certain skills.   Even when you try to drill it into people.   For instance, it is very hard to make someone disciplined who is not with in a work environment.   It may be simply better to hire for these skills if they are important to you.   Acknowledging that other skills may be more easily taught such as product knowledge so someone who lacks them than basic personal skills.  P148 (idea developed by me)

Bill Beane did not watch the games of the Oakland As : “All they provide me with is subjective emotion and that can be counterproductive” (P245 – quote) Instead he focused on non-subjective data (statistics).