Messing Around With Safeco Splits

Operator: This is 911 emergency. How may I help you?
Jeff: I NEED YOU TO ORGANIZE A SEARCH PARTY
Operator: Sir, what appears to be the problem?
Jeff: I AM FREAKING OUT
Operator: Sir, calm down. What is the nature of your emergency?
Jeff: I'M LOST
Jeff: I AM HOPELESSLY LOST
Operator: Can you inform me of your approximate location?
Jeff: BASEBALL-REFERENCE

• It took a long time for the term "park effect" to enter the public lexicon. If the old Coors Field didn't convince people, though, the new Yankee Stadium sure did. However, even now that most everyone is aware that parks affect results, nearly all the attention is paid to simple run or longball effects when, in truth, there are park effects for everything. This won't come as news to anyone who's read this article, The Hardball Times annuals, or Matthew's piece in our Mariner magazine (order now!), but some parks can do some really weird things, and since we're Seattle fans - did you realize that Safeco increases strikeouts by roughly 10%?
  
  Safeco is known as a pitcher's ballpark, and this is due in large part to the fact that no one in history has ever hit a home run to left-center ever. But that's not the only reason, and the strikeouts play a big role, too. Consider the following table of 2007-2009 strikeout data:
  

  Stat (K%)	Mariners	League
  Hit-Home	15.9%	17.1%
  Hit-Road	15.0%	18.1%
  Pit-Home	17.3%	18.1%
  Pit-Road	15.5%	17.1%

  On average, hitters strike out less often at home, and pitchers, in turn, record more strikeouts at home. Over the past three years, Mariner hitters have struck out more often at home, and their pitchers have gotten an additional boost.
  
  This is a big strikeout effect - the biggest in baseball, ~tied with Florida's. Maybe pitchers tend to work higher in the zone in Safeco, knowing that there's a giant fly ball-murdering outfield behind them. I dunno. That's a PITCHfx issue. But while I'm not sure why the effect exists, it does, and it's something one should always keep in the back of his mind. Even Ichiro's not immune, as he has a career 9.5% K% at home versus 9.0% on the road. Safeco, and all stadiums, can do some crazy things.


• Mariner fans love to talk about how the park plays way better for left-handed hitters than right-handed hitters. And that's true - while righties are basically driving the ball into Aeolus after he eats something spicy, lefties actually stand a chance of hitting a home run. However, it's important to recognize that, while lefties in Seattle can hit some dingers, it's still not so much a favorable environment as it is a less-deadly one. They're still vulnerable to the strikeout effect, and they're still vulnerable to a batting average effect as well. So, for righties, Safeco is like an angry pack of wolves, and for lefties, Safeco is like an angry pack of wolves and you have a jar of peanut butter in your pocket.


• It's interesting to me that Safeco doesn't show a bunt effect. Since 2000, and removing all road games in NL parks, the Mariners have averaged 0.258 bunts/game at home and 0.257 bunts/game on the road, while their opponents have averaged 0.216 bunts/game at Safeco and 0.218 bunts/game elsewhere. To be sure, there are fewer opportunities to bunt overall since Safeco reduces team OBP, so maybe this shows an effect after all, but I would've expected such an effect to be more pronounced. Managers know that Safeco lowers the run-scoring environment, so you'd expect them to respond by playing for one run more often. I have no idea if this bullet point was worth writing but it took me several minutes so I'm not going to delete it.


• There are a lot of numbers in the following table. Don't be frightened. You needn't concern yourself with every individual cell.
  

  	League BABIP, Hitters		M's BABIP, Hitters		M's BABIP, Pitchers
  Year	Home	Road	Home	Road	Home	Road
  2009	0.302	0.298	0.294	0.292	0.273	0.275
  2008	0.309	0.295	0.301	0.288	0.306	0.313
  2007	0.306	0.304	0.311	0.315	0.315	0.323
  2006	0.309	0.301	0.295	0.304	0.290	0.309
  2005	0.297	0.293	0.299	0.279	0.288	0.291
  2004	0.300	0.299	0.293	0.321	0.276	0.302
  2003	0.296	0.292	0.296	0.315	0.258	0.285
  2002	0.294	0.291	0.299	0.315	0.287	0.285
  2001	0.298	0.295	0.313	0.326	0.250	0.274
  2000	0.303	0.303	0.277	0.319	0.282	0.298
  TOTAL	0.301	0.297	0.298	0.307	0.283	0.296

  Look at that last line and you'll see what looks like a Safeco BABIP effect. On average, hitters post a slightly better BABIP at home than on the road, while pitchers are the opposite. Mariner hitters, however, have been hurt by Safeco, while Mariner pitchers have received a considerable boost.
  
  Now, there are some selection issues here. The M's have had some real fly ball-prone pitching staffs, for example, wittingly or unwittingly playing to the ballpark. How the Mariners have built their rosters may have had a significant impact on how those BABIP splits come out. But here's where it gets really interesting:
  
  

  	League BABIP, Hitters		M's BABIP, Hitters		M's BABIP, Pitchers
  2000-2004	0.298	0.296	0.296	0.319	0.271	0.289
  2005-2009	0.305	0.298	0.300	0.296	0.294	0.302

  (M's hitters had a .283 BABIP in Safeco in 1999, and M's pitchers came in at .279, against .308 and .352 figures in the Kingdome and .293 and .315 figures overall, respectively.)
  
  From 2000 through 2004, Safeco had a 23-point BABIP split for Mariner hitters and an 18-point BABIP split for Mariner pitchers. Since then, it's actually given Mariner hitters a ~normal home advantage while dropping its Mariner pitcher split to 8 points.
  
  Did something change?
  
  If so, what?