4 replies to this topic

[BSLChrisStoner]

SB Nation: The Myth Of The Myth Of the O's Run Differential
http://mlb.sbnation....stats-standings

Good balanced piece here from Jeff Sullivan.

[Mackus]

Most important part of the article, IMO, and it's a point I'm surprised nobody has really made as clearly and succinctly as is made here.

Here's the good news for the Orioles and their fans: this is no longer a 162-game season. This is, for the Orioles, a 67-game season, and all kinds of crazy things can happen over small samples. Players and teams can over-perform. The key when you're an underdog team like the Orioles is to hang around and effectively shrink the season, and that's what they've done. They don't have to be better than a bunch of other teams over a full year; they have to be as good or better over several weeks. The probability is more favorable.

We don't have to play over our heads for 162 games to have a shot at the playoffs. We have to do it for 67 games. And while some "regression to the mean" could be expected, it's a version of the gambler's fallacy to think that since we played so far over our heads in the first 95 games that we're likely to play below expectations over the final 67 games.

Without any additions, if we play to our relative talent level (70-75 wins coming into the season by some fairly optimistic expectations) that would put us at 80-82 wins. It really wouldn't be that insane for a team to play 5-6 wins better than expected over a 67 game stretch. That would get us to 85-88 wins. We could just as easily play below expectations and end up with 75-78 wins. If we add some small pieces and perhaps one large one to help, maybe that increases the expected talent level by a couple games over the remainder of the season as well.

[fan4life]

It is amazing to me how often a game with thousands of variables is constantly reduced to some specific information over a specific set of time as if that is some indication that all the variables will be the same. Really really just upsets my apple cart.


Who you play, when you play them, where you play them, who they are throwing, who you are throwing and who is umpiring, who's hot, who's not, who's injured but playing, what moves are made before the deadline, is the MiL time for some pitchers going to correct their issues in time to help or not, etc etc etc

[You Play to Win the Game]

It is amazing to me how often a game with thousands of variables is constantly reduced to some specific information over a specific set of time as if that is some indication that all the variables will be the same. Really really just upsets my apple cart.


Who you play, when you play them, where you play them, who they are throwing, who you are throwing and who is umpiring, who's hot, who's not, who's injured but playing, what moves are made before the deadline, is the MiL time for some pitchers going to correct their issues in time to help or not, etc etc etc

I think that stats are good guidelines, in general, and tend to be fairly accurate. It doesn't guarantee that something like this won't happen though. That's the great thing about baseball. The picture, as a whole can be fairly predictable, yet surprises are always possible because the games still have to be played.

[fan4life]

I think that stats are good guidelines, in general, and tend to be fairly accurate. It doesn't guarantee that something like this won't happen though. That's the great thing about baseball. The picture, as a whole can be fairly predictable, yet surprises are always possible because the games still have to be played.

I agree. I am not slamming stats. I am slamming the misuse of stats.