Use of splits by media

[SirKodiak]

Many times on a TV baseball broadcast (and probably in news articles), I've heard/seen statements like:

• (Player X) is the key to (Team A) because when he hits (much better than normal) when they win, but (much lower than normal) when they lose; and
• (Pitcher Y) needs to be more consistent because he has (great pitching stats) in his wins and (terrible pitching stats) when he loses.

In games they won in 2007, MLB batters as a group:

• hit
  
  • .304/.375/.494 (.869 OPS) w/ .335 BABIP,
  • compared to .231/.294 /.349 (.643 OPS) w/ .269 BABIP when they lost
  • while ML overall was .268/.336/.423 (.758 OPS) w/ .303 BABIP
• hit 47% more doubles, 74% more triples, 77% more home runs
• scored 219% more runs
• struck out 15% less
• stole 49% more bases at a slightly higher rate (3%)
• grounded into 11% fewer double plays
• reach base on error 41% more
• had nearly double the number of SH and SF
• were three times more likely to be intentionally walked

For example, Jimmy Rollins

• in Wins, .346/.388/.610 (.998 OPS) .348 BABIP
• in Losses, .232/.288/.428 (.715 OPS) .234 BABIP
• overall, .296/.344/.531 (.875 OPS) .300 BABIP

and Shane Victorino

• in Wins, .318/.373/.487 (.859 OPS) .332 BABIP
• in Losses, .231/.312/.338 (.650 OPS) .261 BABIP
• overall, .281/.347/.423 (.770 OPS) .302 BABIP

"(Player X) is the key to (Team A) because when he hits (much better than normal) when they win, but (much lower than normal) when they lose" could be said of either player. (Rollins is more extreme, but it still holds true)

Reverse the wins with losses above and you will see how the pitchers did. Pitchers who got the Win had a 2.27 ERA and pitchers who got the Loss had a 8.38 ERA, and those that did not get a decision had a 4.18 ERA (ML ERA last year was 4.47).

So when they give splits and analysis like that, it is probably true, and probably not that useful.

[ohms_law]

Nice work here, It's nice to see this in concrete numbers. I always thought those splits were awfully banal, personally. I'm betting that half the reason their so overused is because their one of the big things that Elias provides.

[jcbarr]

I never really paid much attention to them because I always just thought;

"Of course they pitch better when they win."

[ohms_law]

Yea, exactly. I was going to say the exact same thing.

[Swampdog]

The media provides these dramatic splits for all sports, as if there is some real telling information here. It's as though they have the "cause and effect" wrong. When players perform well, they tend to win. It would be like saying, "When Bob runs fast, he wins, but in the races he has lost, he's run much more slowly." Well, duh.

[HoustonGM]

I think, worse than this, is how they display the batter vs. pitcher splits for nearly every batter, and actually draw conclusions from ridiculously small sample size, such as saying "Kevin Millar really sees the ball well off Phil Hughes. He's 3 for 5 lifetime."

Or, when they display a batter's splits vs. a specific team. Ignoring that a lot of times, these players played against the team many years ago when the team was made up of an entirely different group of players...they still draw silly conclusions. I was watching a Yankees/Orioles game a couple weeks ago, and they displayed Jeter's stats vs. the Orioles in like 154 games and were like "Jeter does really well against the O's in his career" and his stats were something like .310/15 HR/80 RBI. Jeez, it seems like he does exactly as well as he does overall. So, he doesn't do "really well against the O's", he does "himself" against the O's.

[HoustonGM]

Which is really well!

True, but it's utterly pointless to say.

[kenny1234]

A similar statement is often made around playoff time when they talk about how few teams have come back from 3-0 down or 3-1 down. The last time I heard it talked about was in hockey and when they gave the stats it was almost exactly 12% of the time a team will come back when down 3-1. But that is exactly the probability (1/8) that you'd expect if there were three consecutive coin flips. Announcers always talk about it like there is some mystery as to why it is hard to come back in a series.

[boomboom]

Check out Jeff Weavers 2007 stats with the M's

When he wins and loses, when he wins, he was something like a Twin of Pedro, when He loses he turned into Bobby Ayala.

[RedsoxRockies]

I wonder if anyobody has a better era/ Batting average in losses? It is plausible

[boomboom]

yes it is. In Baseball ANything is possible

[ohms_law]

humm... now that's an interesting question. There's gotta be someone.

[CatKnight]

Many times on a TV baseball broadcast (and probably in news articles), I've heard/seen statements like:

• (Player X) is the key to (Team A) because when he hits (much better than normal) when they win, but (much lower than normal) when they lose; and
• (Pitcher Y) needs to be more consistent because he has (great pitching stats) in his wins and (terrible pitching stats) when he loses.

I'm just waiting for the media guy who says:

"This team wins more when they win, and loses more when they lose."

[gRYFYN1]

I'm just waiting for the media guy who says:

"This team wins more when they win, and loses more when they lose."

Sounds like a Joe Morgan Quote . . . .

[gRYFYN1]

I wonder if anyobody has a better era/ Batting average in losses? It is plausible

Proabably several batters, i doubt there would be very many pitchers.