4 ms·
> The Data Science team didn’t go to the lengths necessary to support their models. That was their failing. However, the business is ultimately responsible for
by h2odragon 5y ago
> The Data Science team didn’t go to the lengths necessary to support their models. That was their failing. However, the business is ultimately responsible for betting the farm without demanding a real world track record to support those models.
"Well, sure, them folks were (wrong | lying), but its really your fault for listening to them! It's not like making that call was their actual job or anything, ... oh."
I'd say the lesson here is that horseshit is still horseshit even after it has been extensively processed by the finest machines and learning we can buy.
Just imagine being the poor sod there who got a model to say "Bail the fak OUT of this market" 3 months ago: think anyone would have listened?
- ianai 5y agoI think it’s long been known that if the market actors don’t want to heed results of analysis then they simply won’t.
- h2odragon 5y agosee also: https://everything2.com/title/SNAFU+Principle https://everything2.com/title/SNAFU+Principle
- JasonFruit 5y agoMaybe they would have listened if the model successfully used historical data to predict when people should have bailed in the past, in addition to now. That is, if it was a properly validated model. I am not an expert, but this seems like a reasonable way to see if a model works.
- ashtonkem 5y agoOne of the harder lessons of life is that occasionally people get wedged into ideological positions where they want comforting lies that confirm their existing beliefs. These situations often are very dangerous for companies, because it creates incredible pressure on lower level people to produce dubious models, and on middle managers to accept dubiously supported models. As far as who is culpable for those situations … it’s not clear to me.
- darkhorse13 5y agoAnd good luck raising the alarm on anything if you are a bottom tier foot soldier.