Can someone tell me what a high contract/low split, say 90 contact/20L/20R would translate to in stats? How about the opposite, 20 contact,/90L/90R?

Depends on batting eye, speed, power, and the general randomness of HBD.

If  a guy has 20/20 splits, he probably stinks, but i've had a 100/0/20/25/100 guy with 100 speed post a career .250 avg and .400 OBP.  Conversely, splits are pretty worthless without contact/power/eye to supplement them.

Does a high split and low contact mean the batter strikes out a lot, but when he does make contact watch out?

Not really- if a player is rated something like 40/40/90/90/40, he will probably be a very weak hitter.

Really, every rating is important and most good players are rated highly in multiple, if not all categories, without any low ratings.  I.e. a LF with 80/80/80/80 contact/splts/eye can be killed by a 15 power, especially if he doesn't have 90+ speed.  The exceptions like my guy who was rated 100 in contact/eye/speed w/ 0 power and splits in the 20s are few and far between- so target guys with high ratings in multiple categories and pay attention to their performance history.  If a guy's ratings looks good but he has 2000 career AB's with weak stats, he's probably not going to turn it around with a change of scenery and I wouldn't touch him.

I would suggest that you use the player search to look at various ML players to try to pick up on patterns on which ratings yield which performances.  I would also look at the lineups of highest scoring offenses in your worlds to get a feel for how high you need to aim to build a top offense.  It's impossible to explain all of the different combinations that may or may not work, so it's really something you need to explore on your own to develop a solid grasp.

I think the thing I would look at also is this... How does the guy play defense? Because if he's not improving your defense then I wouldn't want 90-20-20 or 20-90-90.

And what deanod says about the variables is absolutely correct. Things that I am only starting to understand come into play, like push/pull tendency.

For instance, a guy with a push/pull of 5 can go yard more often with 70 power than a guy with 90 power and a push/pull of 50. Also the side of the plate they bat from is huge. Let me give you a quick example...

Contact 90 Power 20 LH 50 RH 40 Eye 75

This guy is automatically going to hit better against left handers, correct? Nope. Not at all. And there are a dozen reasons why that might not be true. If he's a speedy switch hitter he's gonna get more infield and bunt singles when he bats left handed. If he's left handed then he may actually have a better average against righties than lefties. Why? Because every time he faces a left handed pitcher he is going against that pitcher's higher split which might average 65. And when he faces right handed pitchers he is normally facing that pitcher's lower split which might average 55. Throw in push pull tendencies, bunting, speed, ballpark and you have enough data to make your head want to explode sometimes.

So like deanod said, to give you a ballpark number is nearly impossible and not at all fair. You gotta figure it out yourself.

But put either one of your guys (90-20-20 or 20-90-90) in Santa Fe and they can hit .300 with 30 BOMBS.

I'll always go with the low contact, high splits. Even if the ratings suggest they should both hit the same.

I found a correlation between ops and total batting rank (vsr+vsl+eye+contact+power). I ran this over multiple seasons with very few outliers.

I have a center fielder with a 9 contact, but all of his other ratings are over 70. And speed is 97. He is one of my most productive players. I was hesitant to play him, but it has worked out. Just FYI.

If you have a high batting eye + speed, that can make up for the low contact. Then look for the combination of splits + power. 

There's a guy I just let go but had 40 97 40 40 80 and i feel like he 'powered' through singles a lot at the plate. He was batting .270 easily over a couple of years with a ton of singles and doubles on top of homers. But with all things equal, I've found contact to correlate directly to batting average. a 50 is like .250, a 75 is like 275, a 90 wouild be .300 or better. This is all give or take the other stats. I feel as though your goal with a player is to ask if he can hit against type-A pitching in the playoffs, so your season stats I personally don't think mean much. You want a guy with good splits so he won't get run over in playoffs. That's just how I approach players.

Holy crap..as a newbie after reading this my freekin head is spinning