Will the real Jason Heyward please stand up?

zitothebrave said:
And you're soooooo well versed on them. You're a regular John Dewan.
Click to expand...

Forgive me for not putting stock into stats that are deemed only "accurate" based on a 3 yr sample. Yet we want to use a single season sample and factor it into actual value for a player? In what universe does that make sense?

Not to say that defensive stats are completely useless. Give a large enough sample you surely can tell to a certain degree how good/bad a person performed defensively. But there's way you can put an exact value on that defense and there is too much variance and personal bias/subjectivity to call defensive stats accurate. Certainly not over a single season (or 1 month, in this case).
 
yeah, that's not how they work. Maybe you should go do some reasearch. Obviously in larger samples like offensive stats, defensive stats normalize longer time. Heyward's defensive stats are static in his first month, but so are his offensive stats. People don't go we're not gonna count their offensive contributions so far because you need a larger sample.
 
zitothebrave said:
yeah, that's not how they work. Maybe you should go do some reasearch. Obviously in larger samples like offensive stats, defensive stats normalize longer time. Heyward's defensive stats are static in his first month, but so are his offensive stats. People don't go we're not gonna count their offensive contributions so far because you need a larger sample.
Click to expand...

Except that defensive stats are subjective and can be subject to bias. Not to mention different metrics might have different rankings of the best/worst defenders. Defensive stats are nowhere near definitive, unlike (most) offensive stats. If a player goes 2-4 at the plate, it is a mathematical fact that his average is .500. That can't be argued.

And I never said you don't count his defensive contributions. It's just not possible to accurately calculate and value his defensive contributions, especially in such a short sample size.
 
Heyward said:
I agree you dont get it.

He's saved a ton of runs.

Defense matters greatly, ask our pitchers if his defense makes no difference the way you make it seem.
Click to expand...

I get it just fine. If you believe that he's the sixth best player in the NL based on that silly-ass metric, I don't know what to tell you. He's not.

The metric is not even close. How you and Zito can continue to spout WAR numbers like they mean....well, anything - is completely beyond me. The metric is useless. It's wrong. It's not close to right. It's crap. It tells me nothing. The moment somebody signs Gerardo Parra to an eight year, $200m deal because his WAR is really up there, I'll concede you're right.

Until then, Zito and Heyward, how about backing off the WAR throttle a bit? You lose credibility every time your "analysis" includes it.

It's not more valid than Giles' or Carpe's or Orphan Black's gut feeling just because it's numeric. Your number is crap.
 
Carp said:
Except that defensive stats are subjective and can be subject to bias. Not to mention different metrics might have different rankings of the best/worst defenders. Defensive stats are nowhere near definitive, unlike (most) offensive stats. If a player goes 2-4 at the plate, it is a mathematical fact that his average is .500. That can't be argued.

And I never said you don't count his defensive contributions. It's just not possible to accurately calculate and value his defensive contributions, especially in such a short sample size.
Click to expand...

Certainly sample size, Carpe, but there something missing in terms of the relationship between the value of offense and defense. I think Giles is on to something when he says there just aren't enough balls hit to RF to make that much difference in the game. He's right. And yet a .194 hitter has a 1.2 WAR 15% of the way into the season because he has ostensibly saved 15 runs? No.

I think for a RF, defense is maybe 15-20% of the job. In center it's more and at short it's a big chunk of the game. For instance, I can believe Simmons is among the best at short because of his glove even with middling offense (last year).

I'm hoping this new defensive measurement system I heard about in spring training can clean up some of this foolishness that Zito and a few other heavy statheads throw out as gospel.

It'd be great to have one number with which to accurately value a player against his peers, but WAR - as currently configured - ain't it.
 
Carp said:
Except that defensive stats are subjective and can be subject to bias. Not to mention different metrics might have different rankings of the best/worst defenders. Defensive stats are nowhere near definitive, unlike (most) offensive stats. If a player goes 2-4 at the plate, it is a mathematical fact that his average is .500. That can't be argued.

And I never said you don't count his defensive contributions. It's just not possible to accurately calculate and value his defensive contributions, especially in such a short sample size.
Click to expand...

But the valuation of that .500 varies. If it's 2 singles, that's not the same as 2 doubles. Or 2 singles and a walk. Then you slide to OPS which on the OBP value values hits equally to walks, and on the SLG side double counts hits as every hit registers a base which factors into OBP. You can slide to wOBA which is the best offensive rate stat that we have. But Linear weights are not 100% perfect. The value of a walk is not necessarily the same every single year and a hundredth or thousandth of a point move may not have a gigantic effect but it's screwing up the valuation.

You don't understand defensive stats. I get it. Instead of blurting out that they're invalid, maybe you should just ignore them (which you do already) and not make ignorant statements about them. No stat that applies value is perfect. And no valuation is even remotely close to perfect. If you think that based on what you want to believe that you can accurately prescribe value, then you're a fool because you're certainly not a better judge than advanced metric.
 
GovClintonTyree said:
I get it just fine. If you believe that he's the sixth best player in the NL based on that silly-ass metric, I don't know what to tell you. He's not.

The metric is not even close. How you and Zito can continue to spout WAR numbers like they mean....well, anything - is completely beyond me. The metric is useless. It's wrong. It's not close to right. It's crap. It tells me nothing. The moment somebody signs Gerardo Parra to an eight year, $200m deal because his WAR is really up there, I'll concede you're right.

Until then, Zito and Heyward, how about backing off the WAR throttle a bit? You lose credibility every time your "analysis" includes it.

It's not more valid than Giles' or Carpe's or Orphan Black's gut feeling just because it's numeric. Your number is crap.
Click to expand...

WAR is absolutely important. Is it important to everyone? No of course not. But you can bet your asses that GMs are looking at that stat when making moves. I think there was a heck of a reason why you don't see "underrated" players sit on the market. I think there was a great reason why Ervin sat on the market so long because people considered advanced stats. If you don't think that GMs utilize WAR, UZR, DRS, etc. then you're a fool. Of course with Field f/x they have something we dont' have access to which is better than advanced metrics we have now. But saying because UZR and DRS isn't perfect so we should throw them out

I hate analogies, but here's the best one I can come up with to explain why defensive stats matter and your believe that "nothing is close" or whatever is foolish. Do you drive a car or use a computer? Neither of them are perfect, new things constantly come out. Do you just live in the stone age waiting for something perfect to come out? Or do you utilize the best available tools that are on the market?

As far as Parra goes, no one would pay him 200M based on WAR. Should Chris Johnson get paid 200M cause he hit .340? Players in one year samples create noise. Parra had a fluke high defensive year. So yeah his value was skewed, but was Chris Johnson's not skewed last year? Will potentially Aaron Harang's not be skewed this year? I mean this idea that a stat is flawed because of 1 year variances is entirely moronic as all stats have variances. Parra is a solid player. He's not a world beater but he's a complimentary player. Probably similar to a less versatile Prado. Good guy to have if you need to fill a hole and can't afford to spend over 10M but can spend 6-8 or so on him.

You can keep making tired arguments, you have no case. You don't get defensive metrics, I get it. It's OK to not get it. I don't get string theory but I don't walk around calling it a silly-ass theory. If you understood advanced metrics more you'd realize why they're used, because they're the best we've got. I'd love field f/x. but that data may never be available to us. So we'll have to rely on UZR and DRS to keep improving.
 
GovClintonTyree said:
Certainly sample size, Carpe, but there something missing in terms of the relationship between the value of offense and defense. I think Giles is on to something when he says there just aren't enough balls hit to RF to make that much difference in the game. He's right. And yet a .194 hitter has a 1.2 WAR 15% of the way into the season because he has ostensibly saved 15 runs? No.

I think for a RF, defense is maybe 15-20% of the job. In center it's more and at short it's a big chunk of the game. For instance, I can believe Simmons is among the best at short because of his glove even with middling offense (last year).

I'm hoping this new defensive measurement system I heard about in spring training can clean up some of this foolishness that Zito and a few other heavy statheads throw out as gospel.

It'd be great to have one number with which to accurately value a player against his peers, but WAR - as currently configured - ain't it.
Click to expand...

Well it's simple, it doesn't matter how many plays a RF makes. Lets say to keep it simple, jason makes 4 plays that no other RF makes, all of them save triples. We know that triples are worth about a run a pop so we know that Heyward saved 4 runs. Of course we know in advanced metrics things aren't that simple. Because even the plays Heyward makes that others make, not everyone makes os he deserves credit for that and he deserves knocks for ones that he doesn't make but others do as well.

Read the fielding bible. just read it. No one I know who didn't read it still says fielding stats are crap. They're not perfect. There's some variance but we're not talking about a full WAR swing in variance. These stats take much much more into account than you give credit for.

And FWIW to your and gilesfan's beliefs. heyward has already at least had 54 balls to make a play on as that's the number of putouts he's had. What advanced stats do is thne take that number and correlate many things. For example some RFs may have worse stats because even if they get more putouts, it's because they're getting mroe opportunities.

Stats are muddy whenever you try to put a valuation on them. But to just ignore stats because of that muddiness we would have nothing to value a player on because literally every stat has valuation issues. If you don't want to use WAR that's fine. But calling it a crap stat is foolish, just because in your mind it doesn't make sense.
 
You must log in or register to reply here.
Back
Top