Value of Defensive Prowess

[Kimmi]

If you had 10 "trained observers" calculating dWAR for a given player all season I wonder what the variance would be in the results-since its a subjective score. If a player has a dWAR of 4 for the season, for example, I would have much more respect for this assessment if each of the 10 observers scored it as a 4 and less respect if half of the observers scored it as a 0 and half scored it as an 8 (extreme example, but you get the point).

The 'trained observers' are knowledgeable and passionate baseball people, not random Joes off the street. Many of them are former players, managers, scouts, etc. They have to pass a difficult test, which includes things like charting pitches, before they will even be considered to be a video scout. Then they go through intense training and practice. I imagine that anyone whose scoring is out of line with the others would be eliminated. Also, the 'trained observers' are rotated regularly to avoid bias.

[S5Dewey]

The 'trained observers' are knowledgeable and passionate baseball people, not random Joes off the street. Many of them are former players, managers, scouts, etc. They have to pass a difficult test, which includes things like charting pitches, before they will even be considered to be a video scout. Then they go through intense training and practice. I imagine that anyone whose scoring is out of line with the others would be eliminated. Also, the 'trained observers' are rotated regularly to avoid bias.

Ok. I get that there are qualified people ranking every play that's made and translating that into numbers for us. But here's my problem with it - and I'm saying this with a semi-straight face:
What these trained observers are doing is using statistics to tell us what to expect from a player based on their prior performance, and apparently some people think it's very close to dead-nuts on. So at what point do we not even bother to play the games? Let's just plug all the data for each team into the computer and it will tell us who wins.
For me all of this uber-information is sucking all the fun out of the game. It's a game and it's made to be played by people. Instead, as Bill James' character said in one episode of The Simpsons, "I've made baseball about as much fun as doing your taxes".

[moonslav59]

Ok. I get that there are qualified people ranking every play that's made and translating that into numbers for us. But here's my problem with it - and I'm saying this with a semi-straight face:
What these trained observers are doing is using statistics to tell us what to expect from a player based on their prior performance, and apparently some people think it's very close to dead-nuts on. So at what point do we not even bother to play the games? Let's just plug all the data for each team into the computer and it will tell us who wins.
For me all of this uber-information is sucking all the fun out of the game. It's a game and it's made to be played by people. Instead, as Bill James' character said in one episode of The Simpsons, "I've made baseball about as much fun as doing your taxes".

WAR is not a projector of what is to come: it's an evaluation of what a player has already done.

Past performance does give some indication of what to expect after factoring in age, experience and recent trends, but WAR does not try to be a tool for determining what is to come.

It is used to try and evaluate the whole player in the context of his peers.

[sk7326]

Ok. I get that there are qualified people ranking every play that's made and translating that into numbers for us. But here's my problem with it - and I'm saying this with a semi-straight face:
What these trained observers are doing is using statistics to tell us what to expect from a player based on their prior performance, and apparently some people think it's very close to dead-nuts on. So at what point do we not even bother to play the games? Let's just plug all the data for each team into the computer and it will tell us who wins.
For me all of this uber-information is sucking all the fun out of the game. It's a game and it's made to be played by people. Instead, as Bill James' character said in one episode of The Simpsons, "I've made baseball about as much fun as doing your taxes".

Here is the thing - the analytics people love baseball ... I mean you love baseball - not exactly good enough to play yourself, but want to get in the sport. One of my favorite defenses was something John Hollinger said about NBA analytics: "guys who love basketball so much they want to do math and computer programming to study it"

All the observers are doing is measuring what happened - that's it. UZR requires enormous samples (like multi-season ones) to draw any meaningful conclusions about players. But UZR is accurate about what actually happened - the same way that a .300 hitter can go for 0 for 10 ... the 0 for 10 says nothing about the dude's ability to hit, but clearly that stretch was not good.

For me - stats are output ... measuring what happened. The stuff you talk about, the watching the games stuff - that is all input ... the thing which causes players to do or to not do stuff. Personally, the information out there now are just tools to give better information about what players are doing - information which has a more realistic basis than RBIs and pitcher wins (let alone saves).