RFan, I don't follow your logic. You rightly discern Upton's numbers as artificially inflated. But then you want him anyway?!

We don't need another Kinsler (road ineptitude). If we are going to acquire an outfielder I say we should insist on a guy who is a plus hitter on the road...and plus plus in Arlington. Otherwise why bother trading at all?!

Especially when you are talking about a guy who will not help team chemistry.

The only Upton I'm interested in is Kate.

Kinsler has posted ridiculous WAR (up until 2012) despite far worse road splits than Upton. So I'm not sold on Upton's road performance submarining his value and the team's record. I don't want to see Elvis go if possible, and Upton is not a perfect player, but the "guy who is a plus hitter on the road...and plus plus in Arlington" is probably not available for a 2-year Andrus rental. Besides, a guy named "ProfarMVP" should be more than OK with trading Andrus to fill a void in the outfield while making Profar the everyday SS.

Personally my biggest hangup with Andrus-for-Upton is that it probably precludes us from trading Profar for Stanton later on. Yes I've drank the kool-aid and I think he will really be available sometime in 2013. Unless they did both deals and somehow got a SS back in one of them. Probably a stretch.

Those interested in park effects should go back to RFan's first post of this thread. It references a very interesting BP analysis with considerable detail.

I've been thinking about all of this park effects home/away split business so suddenly popular in the forum. I'm generally very impressed with this analysis and it is generally valid. Is it possible we can improve on it though?

This is my question. If we are going to introduce all manner of home/away split information we need a better control for our observations.

Can anyone present a conglomeration of all home numbers of all hitters versus all away numbers of all hitters? In other words, to control for effects not having to do with park and other effects we need to know how hitters in general do at home as opposed to away. It seems to me a big reason why guys do worse on the road is because they don't have the comforts of home which has nothing to do with park effects. Further, it could also have to do with knowing your park better and how to play to its strengths. In other words, while it would be great to have a guy who performs as well on the road as he does at home, a lot of home/away statistical divergence may just be based upon the home field advantage which, at least in some aspects, is more intangible than park effects.

Anyways, maybe none of this is new, and the stat brains have already thought of this. So anyways, could someone give us the numbers so we can factor a control into all of this home/away statistical analysis. Of course, if the numbers I suggested show no divergence than the original observations are even more valid.

I thought of another problem. A conglomerate number may not be of tremendous help if there are a lot more hitter friendly parks than there are pitcher friendly parks or vice versa. If the respective numbers of each type are about the same the numbers would increase in validity. Or, you have to find out a way to isolate the home field advantage factor from the park effects factor, which a mathematician would be able to do by taking all of the home numbers and all of the away numbers for each park, adjusting them to account for the park effects of each park, thus creating a home field advantage number for each park, and finally averaging that number for each park to come up with a generalized home field advantage metric for each baseball season. If the number shows significant home field advantage in spite of park effects than it would behoove GM's to adjust their park effect observations by looking at the home field advantage metric if in fact it is isolatable. For example, you would look at a players home and away splits for different ball parks while taking into account generalized home advantage divergence to further legitimize park effects observations. In short, if the average player, taking park effects into account, plays better at home than on the road, then this will effect the way we look at park effect observations.

Baseball Reference has a lot of the data you are looking for. Nearly all of their tables are downloadable in to Excel (look for the csv indicator at the top of the table) and can easy be converted to Excel ready data by using the text to columns tool.

Bottom line: The Rangers have a surplus of shortstops; the Diamondbacks have a surplus of outfielders. This deal has made, does make, and will continue to make too much sense.

Park factors are park factors, but if a guy has the ability to throw up 5-6 wins in a couple of the next few years, let his road numbers be damned for all I care. Ian Kinsler has never been especially productive on the road, but it doesn't change the fact that he's an above-average-to-elite 2nd basemen when he's clicking -- with the majority of that clicking coming at home.

The Rangers' offense has been drastically worse on the road over the last several seasons, but it's not going to change that they play 81 games at RBiA every year. We have to take advantage of our home park, and an offensive alignment featuring Upton/Profar is far more potent than Elvis/Profar and the same cast we still have on roster.

"RFan, I don't follow your logic. You rightly discern Upton's numbers as artificially inflated."

I think RBIA artifically inflates Adrian Beltre's #s. I still want Adrian Beltre as a Ranger. Ideally, the Rangers might pursue someone like Ike Davis. Due to park factors, Davis has little value to the Mets. But he'd probably be a 4 to 5 WAR player in RBIA.

But that's not going to happen. If the only viable option is Andrus for Upton. That's a good deal. 3 years of Upton at a good price is worth more to the Rangers than 2 years of Andrus at a great price, particularly due to our glut of middle infielders. Upton will put up #s in RBIA similar to those that he's put up at Chase Field. And he'll be a decent hitter on the road.

"Can anyone present a conglomeration of all home numbers of all hitters versus all away numbers of all hitters?"

For all AL hitters, in 2012, home wOBA was .326. Away wOBA was .310. So for the avg. player, playing at home adds about 5% to wOBA.

There are however some players (e.g., Ian Kinsler and Justin Upton) who have extreme home/away splits. Some of this can be explained by general park effects. But wRC+ controls for general park effects. Yet Kinsler has a huge home/away split in wRC+. So general park effects don't explain why Kinsler is so much better at home.