Thread: Span vs. Revere OPS

Originally Posted by Twins Twerp

I have never seen someone so good at getting picked off at first. Where do you put that into WAR? What part of WAR measures how a guy hits away from his home field, or how a guy can't hit because he is worried about being traded around the trade deadline.

Things I find amazing: That the analytical folks who come up with formulas for things like WAR somehow decided to only count home games (or is it only away games? Or is it that one should count differently than the other?). That those same folks think they should measure results of plate appearances rather than the alleged thought processes and worries in a player's mind while at the plate. That despite the constant attention and coverage and devotion to recording virtually every single thing that happens in every single baseball game that nobody has thought to record players getting picked off on the base paths.

OR

That someone can apparently think those things are actually true.

Originally Posted by USAFChief

Originally Posted by Brock Beauchamp

Read TT's post again.

No thanks.

It does seem rather inconsistent of you, though, to cite WAR in this thread and then pooh-pooh it in another--when it shows data you discount--all within the space of a few hours.

I was using it purely as a hypothetical in that post. Showing that if you subtract from one position and add more at another, you see a net improvement in team play. WAR is a nice easy way to compare pitchers and positional players, even if it's not terribly accurate in partial seasons.

Originally Posted by Nick Nelson

Originally Posted by Brock Beauchamp

Outfield:
Revere (3 WAR) - Span (5 WAR) = -2 WAR

Pitching:
Pitcher X (3 WAR) - PJ Walters (-1 WAR in a quarter of a season) = 4 WAR.

Net gain? 2 WAR. That makes the Twins a better team and those estimates are very conservative. I think the swing could be as much as 5-6 WAR if Parmelee/Hicks/Arcia pan out and the acquired pitcher is decent.

Honest question, Brock: Why do you put so much stock into WAR as an assessment of value when you've clearly demonstrated in the past that you have little faith in UZR and other defensive metrics? I agree with what you're saying but the way you've presented this argument isn't very compelling to me.

Anyway, like others have pointed out, no one is claiming that Revere is going to be able to replace Span without the offense missing a beat. But the Twins have outfield depth that is only likely to grow going forward, and they have significant needs elsewhere. At this point, Span is probably their most valuable trading chip (outside of maybe Willingham but he's just not going anywhere). Revere is, at worst, a decent placeholder.

Also, Shane, why you assuming that Revere is what he is at this point? He's 24, in his second big-league season, and he showed significant improvement from Year 1 to Year 2. When Span was his age, he was just finally figuring things out in Triple-A.

As I mentioned to USAChief, I used WAR because it's an easy way to compare pitchers and positional players and gauge their value, even if it isn't perfect over partial seasons (if at all). When putting together that hypothetical situation, I was demonstrating how bigger gains can be had by replacing a guy like Walters with Pitcher X than the drop-off you'd see from Revere replacing Span. WAR was the easiest way to do that without going into complex math and confusing the hell out of everyone (including myself).

In the end, all I was saying is "Pitcher X - Walters >> Revere - Span".

FWIW, I think WAR is overrating Span this season and underrating Revere so the net loss is even lower than the numbers I used in my original post.

Originally Posted by Brock Beauchamp

Originally Posted by USAFChief

Originally Posted by Brock Beauchamp

Read TT's post again.

No thanks.

It does seem rather inconsistent of you, though, to cite WAR in this thread and then pooh-pooh it in another--when it shows data you discount--all within the space of a few hours.

I was using it purely as a hypothetical in that post. Showing that if you subtract from one position and add more at another, you see a net improvement in team play. WAR is a nice easy way to compare pitchers and positional players, even if it's not terribly accurate in partial seasons.

Can you tell us why--if it's not terribly accurate in partial seasons--we should believe its accurate in full seasons?

When someone hits .400 for a month, everyone understands the player most likely isn't going to keep that up. Nobody, however, says "but he didn't really hit .400 for that month...that's not what happened."

Yet with WAR, that's exactly what people say. "Well, yeah, I know WAR says Denard Span is the 13th best player in the AL, but that's not accurate. What we need to do is take lots of these innacurate chunks and add them together, and presto change, they become accurate."

Thats like saying "I'm pretty sure one plus one doesn't equal three, but if I add together enough one plus one equals three samples, it becomes good math."

Originally Posted by USAFChief

Can you tell us why--if it's not terribly accurate in partial seasons--we should believe its accurate in full seasons?

When someone hits .400 for a month, everyone understands the player most likely isn't going to keep that up. Nobody, however, says "but he didn't really hit .400 for that month...that's not what happened."

Yet with WAR, that's exactly what people say. "Well, yeah, I know WAR says Denard Span is the 13th best player in the AL, but that's not accurate. What we need to do is take lots of these innacurate chunks and add them together, and presto change, they become accurate."

Thats like saying "I'm pretty sure one plus one doesn't equal three, but if I add together enough one plus one equals three samples, it becomes good math."

I have two issues with your assumption of WAR. First, that people say "that didn't happen" regarding a defensive metric. I think it's more "I don't believe that player is good defensively" is more the case. Even Delmon Young has had good weeks in the field, just as Drew Butera has had weeks where he posted an OPS over .800.

Second, I believe the problem lies in the absolute versus the abstract. At the plate, we deal with far more absolutes than we do in fielding. A homer, a strikeout, a walk. Those are absolutes that are not very open to interpretation. But even in batting, we see the weak double where the batter got lucky. We see a squibber the fielder should have scooped up but didn't and it resulted in the scorer calling it a "hit". That weak double counts just like a true gapper but no one would argue that it is just as indicative of a player's talent as a liner off the left-centerfield wall. That's where WAR suffers a bit. The defensive side of the metric is open to interpretation instead of hard results that are tracked by stat sheets. There will be aberrations that make a player look better than he should over a short period of time. An unusual number of balls may be hit into part of his "zone" that are difficult to reach. An unusual number of balls may be hit into part of his "zone" that are easy to reach. Over time and repetition, these anomalies should even out. It's not that the math is bad, it's that it's more easily influenced by statistical anomalies. Add in the fact that while a batter gets 4-5 chances at the plate every night, he may only have the chance at 2-3 putouts in a night. Other nights it might be 5-6. In the end, all of that combines into a metric that is unreliable in the short-term but fairly decent in the long-term.

And I haven't met anyone who argues that WAR is perfect... Far from it. People usually say that it's a good guideline to use for a player's performance but should never be used as the sole indicator of his play.

I think you can count coup Chief.

Originally Posted by USAFChief

Originally Posted by Brock Beauchamp

Originally Posted by USAFChief

Originally Posted by Brock Beauchamp

Read TT's post again.

No thanks.

It does seem rather inconsistent of you, though, to cite WAR in this thread and then pooh-pooh it in another--when it shows data you discount--all within the space of a few hours.

I was using it purely as a hypothetical in that post. Showing that if you subtract from one position and add more at another, you see a net improvement in team play. WAR is a nice easy way to compare pitchers and positional players, even if it's not terribly accurate in partial seasons.

Can you tell us why--if it's not terribly accurate in partial seasons--we should believe its accurate in full seasons?

When someone hits .400 for a month, everyone understands the player most likely isn't going to keep that up. Nobody, however, says "but he didn't really hit .400 for that month...that's not what happened."

Yet with WAR, that's exactly what people say. "Well, yeah, I know WAR says Denard Span is the 13th best player in the AL, but that's not accurate. What we need to do is take lots of these innacurate chunks and add them together, and presto change, they become accurate."

Thats like saying "I'm pretty sure one plus one doesn't equal three, but if I add together enough one plus one equals three samples, it becomes good math."

Logically, the sum is quite often more than the mathematical total of its parts.

Fallacy of composition = assuming that what is true of the parts must therefore be true of the whole. In this case, the parts do not encompass the total picture; putting them together does.

Someone Cue Edwin Starr... I feel a song coming on...

Originally Posted by Riverbrian

Someone Cue Edwin Starr... I feel a song coming on...

I see what you did there.jpg

Originally Posted by USAFChief

Can you tell us why--if it's not terribly accurate in partial seasons--we should believe its accurate in full seasons?

When someone hits .400 for a month, everyone understands the player most likely isn't going to keep that up. Nobody, however, says "but he didn't really hit .400 for that month...that's not what happened."

Yet with WAR, that's exactly what people say. "Well, yeah, I know WAR says Denard Span is the 13th best player in the AL, but that's not accurate. What we need to do is take lots of these innacurate chunks and add them together, and presto change, they become accurate."

Thats like saying "I'm pretty sure one plus one doesn't equal three, but if I add together enough one plus one equals three samples, it becomes good math."

Because of variance and the fact that not everything that is measured has the same amount of it. What we are trying to measure with a defensive metric is essentially "how likely is a player to make a certain type of play." Is it 70%, 90%, 99.4%...? Well the recorded measurements are only "Did he make a given play or didn't he?" He never makes 70% of the play. There is variance in the observed measurement. Throw in that on the defensive side, most plays are "routine" and the number of games needed to get a reliable estimate is large. Nobody (reasonable) claims that the data accumulated in smaller samples isn't true. They simply say/know/claim that the amount of data to get a reliable estimate is larger than the amount of data that they have.

As an analogy that may make more sense, if we are trying to find the average height of men in St. Paul and we measure one fellow at 6'1". We don't discount that he is actually 6'3", but we also don't know yet if the average height of men in St. Paul is 6'1". We will keep measuring other men to deal with the fact that there is variation. The more measurements we gather, the more accurate our estimate can become. None of our men have the "wrong" height, but none of their individual heights is the answer we are looking for.

I don't know if that explanation makes things more clear or less.

Originally Posted by Twins Twerp

Don't quote WAR, it is such a scam. Put this into your stupid I-pad/computer/I-phone, Our pitching is downright terrible. The differnce between Span and Revere is small. What you lose with Revere in OPS and arm strength, you gain in baserunning and taking hits away. Revere is faster than Span, and is especially better at baserunning. Span does not get the whole baserunning thing. I have never seen someone so good at getting picked off at first. Where do you put that into WAR? What part of WAR measures how a guy hits away from his home field, or how a guy can't hit because he is worried about being traded around the trade deadline.

Let's just stop playing games, and start putting numbers into a computer and name the Oakland A's the Wolrd series champs, because they get on base alot and made a movie with Brad Pitt about it.

Half of the hitting in War measures how a guy hits away from his home field. And WAR includes baserunning too.

Data replicates at some point and you will see the same results repeat itself eventually. However... This is more reliable for the collective.

The data changes for each individual player based on his personal improvement or decline or flatness.

I'm intriqued by WAR but there isn't a single stat that will make me bang my gavel down and say case closed.

As evidence I offer up Carlos Ruiz. Statiscally you can't explain Cooch at age 28, 29 and at age 33. Or Chone Figgins at 29 and age 33.

Ben Revere is young. Lots of data needs to be collected yet. Denard Span isn't as young but age 24 is different than age 26 and 28 is different than 26.

Sweet! Another WAR argument.

what is the argument about anyway? Doesn't WAR say that span is a fair amount better currently and don't the eyes agree?

Originally Posted by Shane Wahl

.750 to .676 right now. I understand that there is a 21 stolen base differential that mildly offsets the slugging difference, but still a .74 OPS difference does NOT mean replacement to me. I still do not like the idea of trading Span before Aaron Hicks is fully ready to step into Span's role.

Let's do this a way that isn't really involving somewhat convoluted stats....

Essentially their OBP has been about the same all year. Revere is down a little now, Span is up a little, but I think they're still within 10 points or so of each other. The difference in OPS is primarily driven by slugging percentage difference. Span is about 60 points in slugging higher than Revere.

60 points of slugging over 500 at bats is a difference of 30 bases. Revere has about 20 stolen bases over Span and more than that if you deduct the damage done by being caught stealing or being picked off.

Add in the additional defensive ability (UZR says Revere is somewhere between five runs and 20 runs better defensively over a year) and I'd have trouble saying the two aren't at least comparable in value. And, of course, that isn't taking into account Span's durability and the risk of him being out for a large chunk of the season.

I guess I might agree that the fan base is generally underrating Span and overrating Revere, but I think Revere is a suitable replacement for him. And I'll take it yet one step further - essentially, moving Span (or Revere) allows the team to plug Parmelee into the roster. That also appears to be a net gain.

What we need to do, is get into a room, rent "Money Ball" and the new Clint Eastwood Movie. They seem to be two 100% fictional movies. Which everone is better, shall rule the day. Money Ball is the WAR, sabre metrics movie, while Clint Eastwood's is the scout, humanistic, have a conversation with an empty chair one.

I haven't seen Eastwood's...but I would say that movie looks to have the edge "on paper."

Sigh.

Originally Posted by Brock Beauchamp

Sigh.

x1000

Span is a good player...not great...who has some value & will get us a decent return. There has always been something about Denards attitude that rubs me wrong.
On radio today, they talked about asking Span about Revere's highlight style catches & why.....despite bein a very solid OF...he never makes them himself. He replied he's too old for that
WHAT

Originally Posted by John Bonnes

Let's do this a way that isn't really involving somewhat convoluted stats....

Revere has about 20 stolen bases over Span and more than that if you deduct the damage done by being caught stealing or being picked off.

Just because I'm that guy...their difference in Picked Off + Caught Stealing is 1. Over 150 games. And in Span's favor.

Originally Posted by greengoblinrulz

Span is a good player...not great...who has some value & will get us a decent return. There has always been something about Denards attitude that rubs me wrong.
On radio today, they talked about asking Span about Revere's highlight style catches & why.....despite bein a very solid OF...he never makes them himself. He replied he's too old for that
WHAT

I heard that on the radio as I was driving back to work at lunch today. I thought to myself, I bet some...ummm...person...is going to take that literally and be all up in arms about Span making a self-deprecating joke. Then I thought, "Nah...nobody would go that far." Then I thought, "I spend too much time reading comments from people on the interwebs when I react to their theoretical reactions."