User Comments, Suggestions, or Complaints | Privacy Policy | Terms of Service | Advertising
Buy MLB playoff tickets, plus 2011 World Series, 2011 ALCS tickets and NLCS game tickets. We also have Texas Rangers playoff schedule, tickets to Red Sox games and Yankees game tickets. Plus, buy Phillies baseball tickets, Tigers playoff tickets and the biggies like ALDS baseball tickets and 2011 NLDS tickets. |
Demarini, Easton and TPX Baseball Bats
|
AllianceTickets.com has cheap MLB Tickets. Get all your Colorado Rockies Tickets, Seattle Mariners Tickets, San Francisco Giants Tickets and all your favorite baseball tickets here. We also carry cheap Denver Broncos Tickets, Seattle Seahawks Tickets and Denver Nuggets Tickets. |
Page rendered in 1.6310 seconds
40 querie(s) executed
Reader Comments and Retorts
Go to end of page
Statements posted here are those of our readers and do not represent the BaseballThinkFactory. Names are provided by the poster and are not verified. We ask that posters follow our submission policy. Please report any inappropriate comments.
Brett Wallace awaits. It's time St. Louis, it's time.
I dunno ... He's hitting .337/.443 in Memphis. Throw in his defensive shortcomings, and I'm not sure he'd be of any help right now.
That's his overall line, but after an early adjustment he's smoking the ball down there. (paraphrasing today's Kevin Goldstein blog entry.) The guy's a hitter and his defense isn't going to get better as he ages. Might as well go for it now.
And still the team is 6th in runs in a 16 team league.
They also don't have any sinkholes. Joe Thurston's 86 OPS+ is the only one among the regulars lower than a 94. According to Baseball Reference, the OPS+ for the National League as a whole this season is . . . 94. So, they really have Pujols, Thurston, some below average bench guys, and six regulars who are average or very slightly above average. The team OPS+ is 96. It sounds about right that they are 6th in scoring.
The OPS+ for the AL right now is 99, so the overall average player is somewhere a bit below 100.
Chris Duncan and Ryan Ludwick turned back into pumpkins, didn't they?
MLB as a whole is hitting 260/333/415. Last year it was 264/333/416. That can't be it. Removal of pitcher hitting? That's probably it.
Duncan had the first disc replacement surgery in professional sports history, and Ludwick has been hurt all year. Factor in Glaus, Ankiel's injury, and now the potential loss of Derosa and you have a classic GOB pestilence with an epicenter in Busch.
The pitching staff has been modestly snakebit, but TLR's position players seem to be made of glass.
This is the least believable thing I have ever read on this site.
I think that OPS+ is calculated excluding pitchers' hitting, which would explain why the AL OPS+ is closer to 100 than the NL.
I guess in response to Tommy Herr and apparently Aaron Miles.
OPS+ is calculated excluding pitchers' hitting, but that doesn't explain the league OPS+ numbers on bb-ref. Whether you include pitchers or not, by definition the league average is 100 ... that's the fundamental premise of the stat. And yet I've been looking up a bunch of historical seasons on bb-ref lately, and time and again the league average is way below 100, like 93 or 94 or something, even long before there was a DH rule.
That just can't be right. There's something fishy in the league OPS+ figures on bb-ref.
Whatever he's doing, it's goofy ... take the 1954 AL: only 2 of the 8 teams with a team OPS+ above 97, and the league OPS+ at 93. 1954 NL, only 2 teams above 95, league OPS+ of 93.
That completely violates the notion of the league average being defined as 100. It just can't be right.
For example, NL 2009 Standard and AL 2009 Standard.
Both league's avg OPS+ are < 100. The NL was below 100 even before interleague play. Shouldn't it be 100? I guess imbalance in the number of games played could play a role, but the NL average still seems too low at 94.
Micah is a Texan, a big guy.
Jarrett is a very small guy, and from Mississippi.
Alan and Andy Benes, 2002.
There is? What? As you know, I'm basically born and raised in Arlington and been in Austin for the last 12 years, so I'm quite familiar with the Hill Country and the parts in between. I've at least gotten a 99 cent hamburger at every McDonald's on I-35 between I-20 and the Capitol, I know the best places in West to get kolaches (not that there's any bad places for kolaches in West, and no it's not the Czech Stop), the two times I've been pulled over in my entire life are within 80 yards of each other on the north end of Waco, I've seen all 11 years of construction in Hillsboro which has seemingly produced one raised section of highway a mile or two long and that's it.
I'm unfamiliar with Hoffpauir ____ (Chevy, Chrysler, whatever), though.
(EDIT) Of course, Bob knows as well as anyone that "Central Texas" is somewhere between a misnomer and useless. Everything between DFW and San Antonio (N/S) and Abilene and Houston (W/E) has been, can be, or is labeled Central Texas. For those of you from parts hither and yon, that's an area basically the size of Pennsylvania, or Maine, or all of VT, NH, MA, RI, and the eastern portion of NY starting with the Adirondack Reserve. 90% of Missouri, 50% of Wyoming, and almost exactly a match for the width of CA with a northern edge of San Francisco and a southern of LA.
The only thing that's really been wrong with Ludwick this year is a .237 BABIP. Everything the guy hits gets caught.
That completely violates the notion of the league average being defined as 100. It just can't be right.
I'm all but certain AROM is correct--the baseline for OPS+ excludes pitcher hitting, but the overall league and team averages listed on BB-Ref include it, so the true league average OPS+ in non-DH leagues will always be under 100.
So the question as always is: Why? If the baseline for OPS+ excludes pitcher hitting, which is justifiable, the league/team averages including it, which is also justifiable, makes no sense. Either approach is fine, but why combine the two? There must be some benefit that outweighs the rather massive philosophical problem with the baseline not being 100. We've discussed this many times before, but I've yet to see where the upside is.
I noticed something similar with the 1986 NL and e-mailed Sean Forman about it a while ago. Apparently, b-ref's league OPS+ numbers includes the pitcher offensive contributions, and the 100 baseline doesn't include pitcher. I would've figured it was fixed by now.
Exactly. It's nonsensical.
Yes, it is. If there is no league average of 100, how in the world are we supposed to assess any particular team OPS+ figure?
Right, as long as it's clear which algorithm is in use for which statistic. The new B-R does a pretty good job of this, imo.
I really do think the GOB want The Walrus in the majors this year, btw. Or maybe Allen Craig, who for some reason is getting zero consideration at either LF or 3B. Injuries to cardinals 3B this year: freese (car accident), Mather (wrist), Glaus (uh?), now derosa, before he even gets a hit?
How many other teams have had THREE season-ending injuries plus whatever is going on with DeRosa, at one single position they got 5.3 WAR from last year and/or traded the closer-of-the-future for, and are still in a position to compete?
Yes, it is. Forget Steve's argument for a second (sorry Steve), the average person that knows about OPS+ thinks of 100 as the baseline, definitionally. 101 is above average, 99 is below. You mess with that and you greatly lessen its adoption and/or usability for that guy, and you make it that much harder to explain to someone new to it. This is not remotely an inconsequential issue, and anything that ##### with that is a massive philosophical problem, in that advanced statistics should strive to be as understandable and relatable as possible within the confines of acceptable tradeoffs to accuracy.
Here there is no accuracy downside to using just one method.
So a little smaller than Northern California...
The only thing that's really been wrong with Ludwick this year is a .237 BABIP. Everything the guy hits gets caught.
yep, the other night I said in a game chatter that Ludwick has looked good recently and got harrassed because he was 2 for 20 previous to that point, and I responded, doesn't matter the result, he has looked good for most of the past two weeks, the results haven't been there, but he has looked a lot better recently.
The way I would tend to use the phrase "Central Texas": the I-35 cities from San Antonio to Waco and a little beyond, most of the interior of the Texas urban triangle (with corners at DFW, San Antonio, and Houston) - but by interior I mean outside the influence of either Dallas or Houston. And only as far west of I-35 as the strong influence of those cities extends. So I'd accept Kerrville as Central Texas but not Abilene - Abilene is West Texas. I guess Bryan/College Station counts as Central Texas, but that's pushing it a little.
Of course the biggest problem with the phrase "Northern California" is that it's universally applied to the Bay Area and to Sacramento and Stockton. If those places are Northern California, what are Chico and Eureka? The same problem exists with West Texas, but in my mind El Paso doesn't belong to West Texas - it belongs to the Trans-Pecos, which is a whole different place.
Southern Oregon?
Just joking as a former Chico resident of 25 years, but to many Bay Area residents, when I told them Chico was 180 miles north of San Francisco, they immediately assumed I lived on the Oregon border.
Of course, the downside to using different baselines in different places is that it can get confusing; people may not remember which baseline is used in which place.
One option is to compare pitchers to other pitchers. Then, when computing totals (team and league figures), the baseline should be weighted by pitcher PAs and non-pitcher PAs. But this is far more complex, and wouldn't make much practical difference.
Either way, once you start using different baselines, you'll see oddities like a player hitting 267/345/410 with a worse OPS+ than his team that is hitting 264/341/409.
I think the way Sean is doing it now is pretty defensible. Each option (including not excluding pitchers) has some serious flaws.
When looking at non-pitchers, OPS+ works just like we'd expect it to: average is 100. The 94 is irrelevant. The Cards' starters are average to slightly below average hitters at those six positions. (And starters tend to hit a bit better than league average, so the Cards starters are a few points even lower than average starting players.)
Actually yeah, but the two are almost exactly the same, as measured by me off Google Maps using a pencil and a piece of paper pressed to my LCD screen. (What can I say, I may be in IT, but I'm analog at heart.) Draw a line from SF directly east to the border, and the Texas area covers all but a tiny part of everything north of that. There's a strip about 15-20 miles tall, running the width of the northern border, that's not covered. Eureka is mentioned, the uncovered strip is about 1/3 the distance between Eureka and the border.
I'd accept Kerrville as Central Texas but not Abilene - Abilene is West Texas
Not that I disagree in the least with your definition, it's pretty good, but the funny thing here is that if you had to pick a city that was actually the center of Texas, it'd be somewhere in the triangle of Abilene, San Angelo, and Brady. As for areas "outside the influence" of the big'uns yet in the triangle, with the expansion of Dallas and Houston, and the Hill Country commuters to SA, they damn near eat up the whole triangle. Houston goes to Sealy on the west and at least Conroe (if not Huntsville) north, DFW goes Hillsboro SW and at least Ennis SE, and SA goes NB (if not SM) NE and Seguin east. What's that leave, Giddings/La Grange/Brenham/Bryan-CS?
With the explosion of Frisco, Sanger, and Pilot Point in the last 10-15 years basically connecting Gainesville to DFW continuously, a fairly conservative imagining of DFW's sprawl is still freaking mind-boggling. Weatherford west, Kaufman east, Cleburne/Alvarado/Ennis south and Gainesville/Sherman north are all pretty inarguable unless you want to mark Denton as the north edge. Each describes an ovoid that's 88 miles wide and either 70 miles tall or ~100 miles tall. With rounded corners, call the latter 7500 square miles. A quick Google of "7500 square miles" tells me that's the exact amount of land area in New Jersey. In 50-75 years, Austin will connect with Waco via Temple/Belton solidly (it's halfway there), DFW will via Alvarado/Hillsboro/West (ditto), and Houston will be 2/3 the way up 45 to Dallas and west on 10 to SA.
Good points, but don't neglect the fact that we're taking Pujols (and hence 1b) out of the equation. That's going to drop the OPS+. It ain't 100 for C-2b-SS-3b-LF-CF-RF. Well, not usually. Let's see what Mr. Eightmanout has to say...
Okay, 2009 NL splits are 102 tOPS+ for starters, 72 for subs, but that's in-game. By position, okay, we get 131 for 1b, the next highest is...wow, 110/107/110 for the OF (121 for DH, but 10% of the PAs.) sOPS+ as applied to league splits by position is confusing me and making my head hurt, so perhaps someone else can parse that.
I don't think they are responsive to your initial point.
Just as much as you may want to "take pitchers out of the equation"; you may have others that want to "take 2b out of the question" or "take C out of the equation". Conversely, someone may want to "take 1b out of the equation." As everyone knows, the different positions tend to have different hitting profiles.
You even have alphabet soup stats that do those things--they compare players to their positional equivalents.
That leads to the question that was posed by Jeff and Peaches --- do you want OPS+ to ALSO do this or do you want OPS+ to have an offensive baseline of 100. That conversation should be shaped on:
(1) the ability to understand the metric without prolonged study
(2) the ability to use it as a comparative metric across different performance spaces.
IT IS NOT, AND SHOULD NEVER BE, AN ARGUMENT TO JUST WRITE ITS DEFINITION.
One of the reasons that batting average is accepted is because its easy to explain and easy to comprehend. You can have convuleted stats that are easily excepted. For instance, ERA takes some advanced understanding of the differentiation between earned and unearned runs.
As Peaches, et. al. has pointed out, OPS+ intutive value is its scale relative to 100. If that scale starts moving so that 100 is not a universally comparative number (e.g. when a league has an OPS+ of 94), its going to get a WTF reaction. The fact that it has the "+" in its name also gives it a very basementy feel.
They aren't, but he wasn't responding to me there, he was responding to the position of Mr. Michael Neal in #11 that they don't have sinkholes, and in fact with the NL average OPS+ at 94, they have Pujols, Thurston below average, and 6 average/slightly above guys. When talking about those 6 guys you can't use 94 because that includes pitchers and you're not talking about the whole lineup here. Additionally, starters are better than bench players so they should the average for starters is going to be over the league average. Both make sense. You can prove the former by looking at NL splits by position. The latter is pretty common sense. Counteracting that is the fact that we're taking 1b out of the equation in this (just talking about the non-Pujols/non-pitcher lineup) which is going to drop it back down some.
That leads to the question that was posed by Jeff and Peaches --- do you want OPS+ to ALSO do this or do you want OPS+ to have an offensive baseline of 100.
Well, that's not quite my sentiment. My sentiment is either include pitcher hitting or don't, but do it the same way every time you present OPS+. The way you get less than 100 as a baseline is when you have some OPS+ number that's inclusive of pitchers and you compare it to some set of stats that doesn't have pitcher hitting. Don't do that. Either way, including or not, you baseline 100, which is a valuable thing. Changing halfway through is where, as Vinay notes, "the downside to using different baselines in different places is that it can get confusing; people may not remember which baseline is used in which place."
To me, the baseline being at 100 is no small thing, but even bigger is that the baseline should always be the same, as presented on bbref. I'd much rather the baseline be at 95 across the board than 100 for 90% of presentations and 95 for 10%. I'm fine if pitchers are excluded always, even in team numbers, and there's just a note on the calculation page (which only those very curious would go to) that says "Hey, that 103 for the Cubs is for their hitters, which is what matters. If you calculate numbers including pitchers, you won't get that." I'd rather it always include pitchers, and you can use tOPS+ or something like it to stratify. But I damn well want it the same scale across the board and for all stats, because I should never see a stat with a definitional baseline of 100 figured by league, where a league doesn't hit 100. No more than I should see a league (pre-IL, or MLB after IL play) where RS and RA don't add up.
I agree with this. What you use OPS+ for when looking at an individual is not quite the same thing as you use it for when looking at a team. For the first, you don't want pitchers in the comparison. For the second, you do. Yeah, that makes it a little goofy conceptually, but it's a better application.
It's not? It is for me. I want to know how well the guy hit, normalized for park. For the team, I want to know how well they hit, normalized for park. I do look at other things, in that I can't think of when I looked at OPS+ for a whole team (that's what runs are for, on the team level), but if I'm looking at OPS+ for a team that's what I want. What are you looking for? Honest curiosity.
It's probably not a good idea to include pitchers in the pool of guys you're comparing him to, then. That would make hitters look better than they really are.
For the team, I want to know how well they hit, normalized for park.
And pitchers are part of that total offensive picture.
I don't know how else Sean could have put that together.
Well, no. That would only be the case if you included pitchers and then entered an 80s movie where a brick falls on your head and you get amnesia (with hilarious results!)
And pitchers are part of that total offensive picture.
Pitchers are part of the total offensive picture for a team, but not the league? This doesn't compute for me. Pitchers hit. Their slot comes up just as often as the other guys. OPS+ doesn't lay claim in any way to being a cross-era stat, what equals a 100 OPS+ in the 2005 NL has nothing to do with what does in the 1935 AL. When measuring how much a hitter is above or below the average hitter in the league (as represented by the league averages), you include all at-bats.
I have no idea what this means.
If you included pitchers as part of OPS+, you'd have a situation where a 100 OPS+ for a position player was below-average, compared to other actual hitters. It would be the same thing you're complaining about, except it would apply to individual hitters instead of teams.
It means that hitters would only "look better" if you included pitchers and then forgot you did so. Otherwise, they look equal. Or they have the same bump vis a vis not including pitchers, but you are aware that you included pitchers and are thus okay with this fact.
If you included pitchers as part of OPS+, you'd have a situation where a 100 OPS+ for a position player was below-average, compared to other actual hitters.
Except the pitchers are actual hitters. They hit. They comprise a significant portion of the plate appearances in the National League. And since when one is talking about OPS+ they are comparing all hitters to the same baseline, it doesn't give any one of them an advantage or disadvantage. When you look historically, including pitchers ends up being no different than looking at a year with a below-average RSE. A 100 OPS+, if pitchers are included in the baseline, is below-average wrt the population consisting solely of position players. That's much easier, to me, to grasp than the notion that a stat with a definitional baseline of 100 to be set at the league average that ends up being displayed as "League Average - 93"
Including pitchers in the calculation would give National League non-pitchers an advantage and American League non-pitchers a disadvantage.
Only if one fails to follow the very basic step of realizing that you must adjust when looking between the two. We accept that the two leagues have different RSEs, just like we accept that 2001 had a different RSE than 1968. Why is it such a big problem dealing with 2009 AL/NL differences conceptually and not 2001/1968? The same numbers in different contexts provide different contextual stat results.
Which is done by excluding pitcher hitting. Which is what Sean does.
It may hope to, pretend to, wish to. But it doesn't, really.
Is hitting 268/324/405 with an 18/54 BB/K rate in AAA as a 24/25 year old (turning 25 in two weeks) a good enough reason?
He is a decent prospect who can very well hit better than he is hitting but it is completely silly to pretend like there is no reason people don't think he would be hitting MLB pitching well right now.
How doesn't it? The premise of making cross-era comparisons using OPS+ is an implicit assumption that the average hitter in 1968 was just as good a hitter as the average hitter in 2001 (to use Jeff's examples) and differences in raw stat lines are because of the era. (this may not be completely true, of course, but it's the premise underlying OPS+)
We KNOW that the average hitter in the 2009 NL is worse than the average hitter in the 2009 AL, because the former group includes a bunch of pitchers who are, we know, terrible, terrible hitters (relative to major-league norms for non-pitchers). Now, the run-scoring environment may also differ between the two leagues - maybe the NL has more hitter-friendly ballparks (Coors, Arizona, Wrigley, etc.)?
But, the premise that matches the premise of my first paragraph is the the average non-pitcher hitter is just as good a hitter in the 2009 NL as the average non-pitcher hitter in the 2009 AL. And you get to that by assuming that the differences in the batting lines <u>of the non-pitchers in the two leagues</u> are the difference in the relevant context.
If the AL and NL are equal leagues, we would expect a player to hit equally well in both leagues. Of what possible good, then, is a metric that would say that a 100 OPS+ player in the AL would become a 106 OPS+ player in the NL, not because the AL is better than the NL, but because the AL uses the DH rule and the NL doesn't?
The problem with what Sean does is that he doesn't re-construct the OPS base at the team/league level to include pitchers. What he does with respect to individual players, on the other hand, strikes me as exactly correct.
OK, then we entirely agree. My issue is with the problem you articulate.
My issue is entirely with the team/league OPS+ figures, not the individual numbers.
If the AL and NL are equal leagues, we would expect a player to hit equally well in both leagues. Of what possible good, then, is a metric that would say that a 100 OPS+ player in the AL would become a 106 OPS+ player in the NL, not because the AL is better than the NL, but because the AL uses the DH rule and the NL doesn't?
Well, I can think of two possible goods:
1) It's the truth. If you just take the formula that gives you OPS+ and plug all the numbers in, the AL will come out above the NL. This isn't a value judgment in the sense that someone's laying blame or anything. You say "...OPS+ is an implicit assumption that the average hitter in 1968 was just as good a hitter as the average hitter in 2001". I say no it isn't. It isn't even an assumption that the guy with a 130 in '68 is as good a hitter as the guy with the 130 in 2001. It definitely isn't any sort of claim that by setting both to a baseline of 100, that hitting skill is equal among the two leagues.
OPS+ measures a player's contribution over league average, adjusted for park, and normalized around a certain point. That's it. If 30 planes all crash into each other tomorrow and 2010 starts with all new (and worse) players, the baseline for OPS+ will still be 100. On top of that, even if it's granted that two leagues are of equal hitting skill on average, the 130s don't mean the two guys who did them are of equal skill, or even had equally valuable production. If one of the leagues has pitchers twice as good as the other, the offensive numbers will be down overall, and it will be easier to produce a 130 (and less valuable) than it is in the other.
2) That 6 difference goes a long way to noting what the effect of the DH rule is on offense. Even absent everything else, there's a good.
My issue is entirely with the team/league OPS+ figures, not the individual numbers.
My issues in order are:
1) The two don't match/aren't calculated the same, allowing for +/- 100 league OPS+
....
....
14) Pitchers shouldn't be excluded.
15) BBRef doesn't make me breakfast.
If you're going through the trouble to provide sOPS+, it would seem you would want to compare batters only to other batters.
You must be Registered and Logged In to post comments.
<< Back to main