SB Nation - Login for mobile commenting

Lookout Landing

Today's Fun Fact

Most of you should be familiar with the CHONE projections, available at Fangraphs and BaseballProjection.com. Of the four different projection systems that Fangraphs offers, CHONE's my favorite, as it includes more information than Marcel while generally avoiding the inflation that seems to plague both Bill James and the Fan forecasts. In Earth collision disaster movie terms:

Bill James: Deep Impact
Fans: Armageddon
Marcel: Asteroid
CHONE: When Worlds Collide

According to Fangraphs, CHONE has 2010 projections for 1367 different hitters, from Albert Pujols to some guy named Adrian Casanova. The two worst projected BB/K ratios:

1. Eliezer Alfonzo (0.112)
2. Greg Halman (0.115)

Go Mariners!

(Alternate optimistic perspective: Greg Halman has a better projected BB/K than a Major Leaguer.)

0 recs  |  26 comments

Comments

Pedro Serrano?
Quick, get me a live chicken!
When Worlds Collide

Awesome movie – with possibly my favorite riot scene ever – but technically that was a solar system-solar system collision film. Wait, should that be solar system-belar system? Either way, it sure was nice of that star to bring a spare habitable planet along for its Earth-destroying voyage.

Without wanting to derail this front page post too much, Jeff has just inspired me to find and download the asteroid disaster movie "Without Warning" which seriously scared me as a kid.
Not turning this in a discussion about how awesome Armageddon and Deep Impact are is one the hardest challenges that this site has ever presented me.
Do you like PECOTA more than CHONE?
I rarely use PECOTA due to the pay wall and my feeling is that it probably adds very little
Yeah I can't justify paying to see it but it interests me just because of its drastically different way of projecting

I was assuming that it added quite a bit of new insight because of this. For example, there is absolutely no reason to use Marcels instead of CHONE however since PECOTA doesn’t use the same methodology, both PECOTA and CHONE should both provide useful information. It might be splitting hairs at that point though so it might just not really matter.

How is PECOTA's methodology different?

I thought their process was proprietary and unknown to the public.

PECOTA works by creating similarity scores with past players

to find players that are most similar and then uses their performance to build a projection. It tends to work pretty well (for example: speedy players age differently and prospects can be difficult for Marcels to project). It breaks down though at the extremes because there aren’t any players like Ichiro so it kind of freaks out.

Please, somebody let me know if that is incorrect.

Beyond Marcel, it's been demonstrated that other systems have very little marginal value anyway

It’s amazing how far you can get with 3 years of data and basic common sense.

Unless there have been new efforts to look into this,

the studies showing this have been done on bulk projections. Which is fine, but I think it doesn’t expose the potential weaknesses of the systems. Most baseball players are easy to project once they establish themselves (which is why even Marcel can do well) but I’d love to see a study comparing how accurately the systems were able to project players with less than 600 major league plate appearances to their name, or those who are over 36 for example.

Does this mean that...

Fan projections make the Criterion Collection?

Hey, I like Deep Impact.

It’s the only asteroid impact film with much in the way of real science in it.

Armageddon is a steaming pile.

Yes because the biggest difference between Deep Impact and Armageddon was the actual science in Deep Impact
I'm a scientist and I thought Deep Impact was boring. Armageddon was much better.

I watch a movie about asteroids hitting Earth cause I want to see stuff blow up.

Maybe I like crappy movies but there is a continuum between serious and action movies. Either extreme are worthy movies however I feel like Deep Impact tried to straddle a fence and it just doesn’t quite work right. This happens a lot with science fiction where they try to explain how crazy technology works. If you don’t explain it, I can just buy into it and “believe” it works however if you try to explain it and you are wrong then that just annoys me.

And Armageddon is somehow better in this respect?

Even if you don’t understand science in the slightest, there is a whole shitload that would make you go “huh?”.

Truthfully I haven't seen either movie in a long time

But I remember Armaggedon having more memorable and enjoyable things happening than Deep Impact. It could be that I haven’t seen either since high school and just wanted to see an absurd cast of actors want to blow stuff up. I would check my memory by watching both of them again but I’m not really sure its worth the time.

I am a professional rocket scientist
what is your professional opinion of the carrot?
This seems interesting, but I'll wait until I read it from a better source.

This “Jeff Sullivan” fellow seems a bit too much like Jeff_ for my comfort.

This guy seems like a lil bitch
I rec'd this in the hopes that "this guy" was Jeff Sullivan and not me.

If it was me, I wish to rescind my rec.

You must Login with your SB Nation account and be a member of Lookout Landing to post a comment.