Oh, I gotcha. I want to say that it is more stable but that's based on some vague memory of a BPro article. I don't have a subscription anymore so I can't help you with that one.

Yeah, I'm struggling to come up with a true hypothetical example, but my point is that when we say that someone's value depends on his BA too much, it's with the underlying assumption that BA can swing wildly from season to season, and that unless he draw walks and gets on base in other ways, he's less valuable. Why?

Sure, we'd all like to see BA-dependent guys draw more walks and get on base in other ways. For that reason, when two guys hit .310, we'd rather have the guy with a .400 OBP than the guy with the .350 OBP. That's not really my point.

What I'm wondering is why, when two guys have a .350 OBP, we tend to like the guy that did it by drawing 60 walks and hitting .280, rather than the guy who drew 40 walks and hit .300?

I mean, let's look at three hypothetical guys, playing the same position equally well and having the same age, past years, plate appearances, ballpark, era, etc.:

Player A -- .310/.360/.400
Player B -- .280/.360/.400

Why do we assume that Player B is nearly as valuable as Player A, if not more so?

Well, excuse the snark, but that's two players, not three. Sure your two players are probably roughly equal in value because they have identical OBP and SLG - although of course a more precise breakdown of their lines could reveal significant differences in value. Now, then you raise the question as to which is less "variable." That raises two questions. What do you mean by variable?

Are you saying Player A and Player B just put up those seasonal lines, or are you saying those lines are their "true talent level"? Because if it's the first, then all kinds of factors will go into projecting them forwards - for instance, their respective strikeout rates. If it's the second, then variability doesn't really matter.

What I'm wondering is why, when two guys have a .350 OBP, we tend to like the guy that did it by drawing 60 walks and hitting .280, rather than the guy who drew 40 walks and hit .300?

I think it's an overreaction to mainstream love of batting average. But if all things were equal (which they rarely are) I prefer the .300 hitter, because they tend to do better with runners on.

But here's the thing about Pierre; he's a veteran. We know what his game is, what he's capable of - we have much more than 1 season's data. He had difficulties adapting to a new park - witness the strangely high strikeout totals for the first two months, for example, and the excessive number of fly balls. But now he's made the adjustment and he's doing his normal thing. Even if you insist on including the first two months, he's hitting .276/.323/.373, with excellent speed and baserunning. That's not what you want - the SLG is 50 points too low - but it's no disaster. And just because he has a weak arm I think he's very underrated in the field. But even BPro has him league average. If I had to guess at his "true talent level" in Wrigley I'd say .310/.350/.390, with 40+ steals. I don't know what you want at the top of your order, but I'll take that.

But even BPro has him league average.

Which puts him at -31 over the last four years. The website differs from the annual for some reason. Strange.

If I had to guess at his "true talent level" in Wrigley I'd say .310/.350/.390, with 40+ steals.

Wrigley is famous for the tall infield grass being death on grounders.

If I had to guess at his "true talent level" in Wrigley I'd say .310/.350/.390, with 40+ steals. I don't know what you want at the top of your order, but I'll take that.

I think anyone would take that. The question is, is that really what you can expect from him? And would you be willing to bet a lot of money over three years on that?

If I had to guess at his "true talent level" in Wrigley I'd say .310/.350/.390, with 40+ steals.

I suppose that's not too far off his unadjusted career line of .301/.351/.375, which has been logged in a hitter's paradise (Coors), a hitter's hell (Pro Player), and a hitter's, um, purgatory (Wrigley). But a weighted average of his trailing three years (including his career year of 2004) -- a sample of nearly 2000 PAs -- yields a significantly lower line of ~0.285/.330/.370, with ~30 net steals. That's still not a disaster, but it's not nearly as good as you make it out to be. Moreover, his performance is trending downward -- Pierre's going to be 29 next year, and it's not unreasonable to expect an age-related decline.

So, sure, I'd take that too -- for the right price. 1 year at $5 million? Sure. 2 years at $12? Maybe, maybe not. 3 years at $25? Uh, no thanks.

FWIW, PECOTA's MORP for Pierre:

2006 $4.5 million
2007 $4.1 million
2008 $3.8 million
2009 $4.0 million
2010 $2.5 million

FWIW, PECOTA's MORP for Pierre:

2006 $4.5 million
2007 $4.1 million
2008 $3.8 million
2009 $4.0 million
2010 $2.5 million

The other thing to keep in mind is that real teams don't have generic "replacement players" in mind, they have specific replacement players, and in the case of Pierre, his replacement is Felix Pie. If you think Pie is worth $1 million above the minimum salary in 2008, then Pierre's marginal value to the Cubs, assuming PECOTA is entirely accurate, is only $2.8 million that year, which is probably at least $5 - $6 million less than it'll take to sign him.

The other thing to keep in mind is that real teams don't have generic "replacement players" in mind, they have specific replacement players, and in the case of Pierre, his replacement is Felix Pie. If you think Pie is worth $1 million above the minimum salary in 2008, then Pierre's marginal value to the Cubs, assuming PECOTA is entirely accurate, is only $2.8 million that year, which is probably at least $5 - $6 million less than it'll take to sign him.

Good point. It may or may not be the right decision to start Pie in CF in 2007, but if they did, he's a decent bet to match Pierre's performance at the plate and in the field (though not on the basepaths). PECOTA puts Pie at .268/.318/.447 for 2007, which seems roughly right as a weighted mean (median performance would likely be lower, but there's some significant upside built in). If that's in the ballpark, Pie will be roughly equivalent in value to Pierre at -- what? -- five or ten percent of the price.

Well, excuse the snark, but that's two players, not three.

That's because I was originally going to make up a third line, then decided against it and didn't change the text until too late.

Sure your two players are probably roughly equal in value because they have identical OBP and SLG - although of course a more precise breakdown of their lines could reveal significant differences in value. Now, then you raise the question as to which is less "variable." That raises two questions. What do you mean by variable?

See my post #51 for a better idea of what I was getting at.


Are you saying Player A and Player B just put up those seasonal lines, or are you saying those lines are their "true talent level"?

Do we really know? Let's just say that I'm talking about two players the same age, playing in the same ballparks, eras, etc., and that these are career stats rather than single season. I'm also assuming other numbers are kept constant, such as SBs, Ks, RBIs, etc.

Just from grins, let's look at a couple better illustrated hypotheticals, again assuming the same age, position, ballpark, era, handedness, strikeouts, stolen bases, caught stealings, defensive ability, etc.

Suppose we have these two players (again, all other things being equal):

           <u>AB</u>  <u>BB</u>   <u>H</u>   <u>2B</u> <u>3B</u>  <u>HR</u>    <u>BA</u>   <u>OBP</u>  <u>SLG</u>
Player A  600  85  165  30  5  20   .275 .365 .442
Player B  600  30  200  30  5  20   .333 .365 .500

Which player would you prefer? Why?

I could make a similar question and have it swing on ISO rather than SEC. Suppose we have these two players:

           <u>AB</u>  <u>BB</u>   <u>H</u>   <u>2B</u> <u>3B</u>  <u>HR</u>    <u>BA</u>   <u>OBP</u>  <u>SLG</u>
Player C  600  60  155  35  0  25   .258 .326 .442
Player D  600  60  200  25  5  10   .333 .394 .442

Who would you prefer? Why?

Which player would you prefer? Why?

Depends on the composition of the rest of my team. If I've got a bunch of Sean Casey's, then I'll take the higher SLG rather than OBP. If I've got a bunch of Jose Guillen's, then I want the OBP at the expense of SLG.

But now he's made the adjustment and he's doing his normal thing. Even if you insist on including the first two months, he's hitting .276/.323/.373

It's not just the first two months. It's the first two months and the entirety of 2005.

Yes, he's been back to his old self again. But as others have pointed out, Pierre's BA is largely a product of his speed; if he is gimped by injury, bye-bye BA. He has had some nagging injuries the last couple of years, and at his age, we should probably expect those sorts of things to increase and his speed to decline. Thus, his BA is less and less reliable.

Player B vs Player A (unless there's some significant reason why A had more PAs than B)

Player B is clearly better. Why? Because he advances baserunners better. He had more TB in fewer PAs, while making outs at the same rate. A linear weights-style argument will also obviously demonstrate B's superiority in value. In addition, batters of player B's type tend to do better with RISP; player A-type batters are more likely to suffer depressed BA and SLG, inflated OBP with RISP.

However, Player A might be a better bet for the future, because he has better K:BB ratio.

Player D vs Player C

(Note: This comparison strikes me as strange; Player D is blatantly much quicker than Player C, so it's highly unlikely to me that their other attributes will be identical. However...)

Obviously this is going to be affected by the era. I assume we're in the current era; intuitively I think Player D is much better because of his huge superiority in OBP. But let's check, using linear weights. Ignore walks as they're equal for both.

Player C has 95 singles, 35 doubles, 25 HRs, and 445 outs. Linear weights says that makes him worth;

95*0.49 + 35*0.79 + 25*1.42 - 445*0.30 = 46.55 + 27.65 + 35.5 - 133.5 = -23.8

Player D has 160 singles, 25 doubles, 5 triples, 10 HRs, and 400 outs. By linear weights;

160*0.49 + 25*0.79 + 5*1.06 + 10*1.42 - 400*0.30 = 78.4 + 19.75 + 5.3 + 14.2 - 120 = -2.35

So linear weights confirms the intuitive view; Player D is better, and it's not even close.

Question: What are these examples supposed to prove? That if OBP is equal, greater SLG is (generally) better? That is SLG is equal, greater OBP is (generally) better? No something, sherlock.

What i have never gotten is why people think that BA-dependent hitters are so much more variable in value than others.
ALL hitters see variation in their BA walks n homers guys see their BA jump around just as much as a .300 hitter with no walks or homers. BA-dependent is seen as a criticism for some reason.
If these two hitters are equal in value to start with the walks and homers guys value will but just as negatively affected as the slap hitter's if both their batting averages "randomly" drop by the same amount. With two players with the same value the BA guy will have a much higher range of variation so when he is screwed by "random fluctuations" his BA will still be, say, .270 instead of .230 like Sluggo McBB

Question: What are these examples supposed to prove? That if OBP is equal, greater SLG is (generally) better? That is SLG is equal, greater OBP is (generally) better? No something, sherlock.

Neither, really. You were the one saying there is a skill to BABIP; I'm just trying to come up with examples to flesh it out (and apparently did a poor job of it).

Specifically, I wasn't trying to look at either OBP or SLG per se. In the first case (Player A versus Player B), the difference is purely BABIP -- Player A has achieved his OBP by virtue of extra walks (and a better SEC), while Player B has walked a lot less, but hit more singles and is more BA-dependent. In the second case, I was just noodling around trying to judge a guy who achieved his SLG by hitting for power compared to a guy who hit more singles.

I wasn't really trying to say who had the better season as much as I was trying to figure out who has more value on a going forward basis. Is there anything to think that a BA-dependent hitter who doesn't walk and relies on a high BABIP (i.e., Juan Pierre) is less valuable than guys who depend on BA a lot less? Why do we tend to discount the value of pure BA and tend to assume that it is more subject to luck?

Like I said, though, I was just noodling with some numbers to try to reinforce your point. Apparently, I didn't do a good enough job. Next time, maybe I should let you just back up your own points yourself.

ALL hitters see variation in their BA walks n homers guys see their BA jump around just as much as a .300 hitter with no walks or homers. BA-dependent is seen as a criticism for some reason.

A BA dependent guy needs to hit for a high average to be valuable and if he loses 30 points in average he's not going to be too good. Someone like Pierre.

A guy with a lot of walks and XBH could lose 30 points of average and still provide a lot of value. Someone like Dunn.

Or at least that's how I understand it.

I was saying with equally productive hitters...Dunns usually good for 50 more points of OPS+ than Pierre.

Even if you want to use Dunn and Pierre doesnt Dunn lose just as much value as Pierre when losing 30 points of BA, Dunn would still be better just because of the simple fact that he IS a better hitter

Note-im ignoring the lost basestealing opps losing 30 points of BA would have on Pierre. Im talking only about a players hitting contributions..and anyway i doubt the loss of a few steals would have a significant impact

Jonny Gomes is a good example of a walks and power guy being just as affected by BA as a BA-dependent guy.

Last year he hit 282/372/534...this year 212/319/439. Both in 350 ABs. His career average is now 242. His career minoe league average is 271. His seasonal BAs in levels above A-ball are 249 (AA in 03) 256 (AAA in 04) and 321 at AAA in 05. He has 579 Ks and 260 BBs in 1665 minor league ABs and 96 BBs and 233 Ks in 730 big league ABs

I think its reasonable enough to say his "true talent" is something like a 245 average. Now thats not based on the most indepth analysis but i think its accurate enough, and im too tired to try harder-sorry.

So in 05 he gets a little BA-lucky and hits 282...this year it does the other way and he gets a little unlucky. Not so surprisingly he was a damn good offensive LF last year...this year no so much (and he even upped his walks this year!)

Say theres some fictional player who hit 330/370/530 last year but fell to 280/320/440..sounds sort of like a Garret Anderson type (who is overly maligned round these parts)..isn't Gomes just as likely to see his value drop off as Mr Andersons clone?

So why is Gomes-type often thought of as being more "stable" and less prone to variation than a player of comperable value whose value is built more on batting average?

Actually not even why, i know why, but rather...is there any good reasoning behind this thinking?
i doubt it

A BA dependent guy needs to hit for a high average to be valuable and if he loses 30 points in average he's not going to be too good. Someone like Pierre.

A guy with a lot of walks and XBH could lose 30 points of average and still provide a lot of value. Someone like Dunn.

Sure, a BA dependent guy won't be good if he loses 30 points in BA. Still, if a guy who gains the same OBP by walking 70-100 times has a season in which he draws half as many walks, his OBP will suffer just as much. Is there some reason why we have more faith in walk-dependent guys like Mark Bellhorn rather than more BA-dependent guys (say, Darin Erstad)? Is drawing walks a skill that varies less than hitting for BA?

Actually not even why, i know why, but rather...is there any good reasoning behind this thinking?
i doubt it

My guess is that it's based on a study that concluded that players heavily reliant on BA are much more prone to sudden drops in value than players less reliant on BA.

Sure, a BA dependent guy won't be good if he loses 30 points in BA. Still, if a guy who gains the same OBP by walking 70-100 times has a season in which he draws half as many walks, his OBP will suffer just as much. Is there some reason why we have more faith in walk-dependent guys like Mark Bellhorn rather than more BA-dependent guys (say, Darin Erstad)? Is drawing walks a skill that varies less than hitting for BA?

The guys you used with the same OBP and SLG had a difference in power. The lower BA guy accounts for more of his SLG with power, increasing things like doubles and homeruns, making him more valuable.

Pierre mentioned "Pierre to the White Sox" three different times during the offseason. The Trib has some kind of bizarre fascination with that idea.

That should be: Phil Rogers mentioned "Pierre to the White Sox" three different times during the offseason. The Trib has some kind of bizarre fascination with that idea.

Like I said, though, I was just noodling with some numbers to try to reinforce your point. Apparently, I didn't do a good enough job. Next time, maybe I should let you just back up your own points yourself.

I wasn't trying to be rude - I apologise if I came off that way.

The problem in this thread is that there seems to be a fundamental misconception with the "all other things being equal." And that's why I keep mentioning strikeouts. The point is this; a walks and power guy typically strikes out a lot. If you walk a good amount and you don't strike out much, you're basically a superstar. Here's the list of players with a K:BB ratio of 1 or less, and at least 60 walks:

Albert Pujols, Todd Helton, Brian Giles, Nick Johnson, Jason Giambi, Barry Bonds.

Note that there was no requirement for these players to be power hitters, but they all are. Expanding it to 45 walks (cos the season isn't over) includes the following:

Joe Mauer, Frank Thomas, Luis Gonzalez, Scott Hatteberg, Trot Nixon, Omar Vizquel

A couple of guys who aren't sluggers now creep in. But the point remains; walk a lot, and keep your K:BB ratio down, and you're sweetness itself.

The point I'm trying to make is this; strikeouts are really really important. Why? Precisely because of the BABIP which some people here think counts against Juan Pierre. Let's look at BA/SLG on balls in play, because that's more important.

Prototypical 3-true-outcomes star Adam Dunn is hitting .379/.826 on balls in play, compared to Pierre's .295/.398. So if BABIP/SLGIP is just luck, then it's Adam Dunn who's value is about to collapse, not Pierre. Of course, BABIP/SLGIP is not luck (at least for batters). To a large extent, it's a function of how hard you hit the ball. Dunn smacks the ball really hard, so when it goes in play he does well - Pierre not so much. However, Pierre does put the ball in play far far more.

So when we're looking at Players A,B,C and D in post #59, for example, we have to realise that typically, A and C will strike out more than B and D. And this is really important going forwards. Can these guys maintain their BABIP/SLGIP? How much room for growth is there - there's a ceiling for how well you can do on balls in play, and we know it's unlikely that they'll suddenly halve their strikeouts and double their walks. Whereas Players B and D, if they don't strike out much, can easily improve by putting on muscle, and hitting the ball harder.

OTOH, if strikeouts are even, then it cuts the other way. If player A is a high strikeout player we can forgive him because of the walks. If player B is a high strikeout player we should be very concerned about the sustainability of his value. If player B is a low strikeout player, good. But if player A is a low strikeout player, then he'll have superstar peripherals, and he won't need to do much better on balls in play to put himself up in the Pujols/Helton/etc territory. So basically it's complicated.

But we should get back to the main example - Juan Pierre. Pierre doesn't strike out much, and maintains a good K:BB ratio. He's a .300 hitter. So why is he having a bad season? The first two months. And what went wrong in those first two months? Bad luck on BABIP? Perhaps he was just unlucky then and lucky since, so he really is a .276/.323/.373 hitter? No. It is, as I stated above, about the strikeouts.

Pierre has 30 strikeouts all season. 20 of those came in the first two months, and only 10 since then. Strikeouts typically stay fairly constant, so what's going on here? The answer, to me, is obvious. Pierre had trouble adapting to his new stadium, and wasn't seeing the ball well - a lot of players complain about this at Wrigley. But now he's made the adjustment, and is back to his old self.

The guys you used with the same OBP and SLG had a difference in power. The lower BA guy accounts for more of his SLG with power, increasing things like doubles and homeruns, making him more valuable.

I'm dancing around something here. My question isn't really "who had the better season?"; it's "which guy would I want in the future?"

Should I have more faith that the low BA guy will raise his BA to the point that he'll be more valuable (or at least maintain his plate discipline) . . . or should I have more faith that the BA-dependent guy will draw more walks to push his OBP even further (or at least maintain his BA)?

Take Bellhorn, for instance -- he has some value if his BA is in the .250 range, but if it falls lower than that, he's useless, regardless of any power he may have. Why do we in the BTF community give him so many chances, while we're so quick to slag a guy like Darin Erstad (just an example) because we say something like "he needs to hit .300 to be any good"?

Isn't it the same thing? For that matter, is it easier for a .287 hitter like Erstad to hit .300 than for a .236 hitter like Bellhorn to hit .250?

More importantly, on the other side of the coin, why do we have more faith that Bellhorn will hit .250 than Erstad will draw 10 more walks? You commented that "[m]y guess is that it's based on a study that concluded that players heavily reliant on BA are much more prone to sudden drops in value than players less reliant on BA." Where is this study? We all seem to assume this is true, but where does this come from?

20 of those came in the first two months,

9 at home, 11 on the road. Doesn't really support your argument. OTOH, it is true that he performed much better on the road for the first couple of months. Maybe he just wasn't seeing the ball well enough to make good contact at Wrigley. He certainly hit enough relly weak groundballs back then to support that case.

Don’t forget that Pierre is also a favorite of White Sox manager Ozzie Guillen from their days together in Florida, where Guillen was third-base coach. Guillen could persuade Sox GM Ken Williams to push for Pierre if he goes on the market in November, depending on how the Sox feel about Brian Anderson.

Pierre might be seen as a replacement for Podsednik, whose contract is up (I think he has two more arb years left, but he's worn his welcome out with Guillén) but I doubt Guillén or Williams see him as a replacement for Anderson, who excels defensively, will hit just fine now that he's made the adjustment to the majors and plays for peanuts.

And the Cubs and the Tribune aren't the only ones who see lead-off batter as a position. The Sox do, too.

Prototypical 3-true-outcomes star Adam Dunn is hitting .379/.826 on balls in play, compared to Pierre's .295/.398. So if BABIP/SLGIP is just luck, then it's Adam Dunn who's value is about to collapse, not Pierre. Of course, BABIP/SLGIP is not luck (at least for batters). To a large extent, it's a function of how hard you hit the ball. Dunn smacks the ball really hard, so when it goes in play he does well - Pierre not so much. However, Pierre does put the ball in play far far more.

Excellent post. I got caught up in drawing hypothetical Players A-D, which got things sidetracked and didn't really explain my question as well as I did in Post #74.

Here's another question on your point, though: I don't think anyone would put Dunn side-by-side with Pierre because they are two separate players with different skillsets. Let's take a closer comparison -- say, Pierre and Dave Roberts:

Pierre: .296 BABIP/.399 SLGIP (career .322/.401)
Roberts: .351 BABIP/.486 SLGIP (career .312/.433)

Roberts has a bit more power than Pierre, but I don't think anyone will say he hits the ball *that* much harder than Pierre. Do we think that Roberts will regress? Pierre will improve? Both? Neither?

More importantly, how much of this is luck versus skill (obviously, there is an element of both)?

Just two more seasons at the most, and they can re-sign him!