Player Factory
Some organizations get a reputation for making players better. Track every player who changes teams against what was expected, and see whose reputation holds up.
- Coach questions
- 3
- Asked on air
- 16×
- Top question
- Q-06
Coach questions it could answer
The playbook →- Do some organizations and coaches really improve the players they get?
For every player who joins a new organization, compare what he did there with a projection made from his numbers before the move. Average the gap by organization to see which ones really add value, with error bars.
- How well do international stars translate to MLB?
Apply it to players arriving from Japan and Korea, comparing their first MLB seasons with what their previous league's numbers translated to.
- Do trades and deadline moves actually add the value expected?
Score deadline acquisitions the same way, over the rest of the season, against what a projection expected, so "the trade worked" means more than one hot month.
Data it needs
The data lake →0 of 2 on hand; the rest arrives in phase 4, Front office.
Could use later: Minor leagues, Other leagues, The draft
The question
The hosts often credit a team’s coaches when a new arrival takes off. But players also take off on their own, and some regress after a good year anywhere. Which organizations improve the players they get, beyond what you’d have expected anyway?
How it would work
- For every player who changes organizations, build a projection from his seasons before the move, regressed to the mean.
- Compare what he did after the move with that projection, and average the gap by organization over several years.
- Show each organization’s lift with its uncertainty, so a single big success story doesn’t make a factory.
First version
- Hitters and pitchers who changed organizations in the last five seasons, with a simple projection from the prior three.
- An organization leaderboard with error bars and its best and worst examples.
The data
MLB Stats API and Statcast season stats and transactions. International league stats for the translation part come later.
What it needs
- A fair projection. Aging, regression and playing time all have to be in it, or every organization that acquires young players looks like a factory.
- Enough players. Some organizations acquire few players a year, so results need several seasons pooled.
Status
An idea on the launchpad. Nothing is built yet.