Lookout Landing

Sounds like suitable work

Fett42 - August 14, 2008 reply

Atta way to go, boys!

Good stuff.

;)

dnc - August 14, 2008 reply

Congratulations!

And thank you

Where could I find league average wOBA?

It will be added soon a la lgTRA is listed

but for now, a good proxy is just the straight league average OBP, somewhere around .331/.332 currently.

Cool thanks

With that are you considering something like wOBA+?

Yes.

And park-adjusted wOBA as well.

You're my hero

I suppose you'll also eventually have leaderboards

Oh and small things I noticed:

There’s two Sabathias: C.C. and CC

And there’s no combined 2008 for him or any player traded midseason. (You probably know these things already)

Stupid prima donna baseball players changing their names. In the middle of the season, no less!

He did it to spite his old employer.

They wouldn’t let him keep playing.

Fin - August 14, 2008 reply

Yes

Question

I aways wondered how sites like this, fangraphs, etc update every day. Does it have to be done manually or can it be coded to take the info from elsewhere automatically?

Fett42 - August 14, 2008 reply

Coded.

For instance, it’s updating as we speak :)

Holy crap you guys are good at what you do

just one suggestion – seems like you have some LL-centric memes in your writing (the tRA intro post for example calling Matthew the Data Fairy). Might seem a bit more professional, for lack of a better word to clean it up so non-LLers reading the site won’t think “wait wtf?” when they come across something like that.

The site looks fantastic though – it’s already been bookmarked

Dammit Graham. :P

Thanks for the catch.

baller

This is awesome. Congrats, guys.

I’m excited to be able to find tRA on my own, and that’s what I was expecting with this announcement, but I didn’t know you guys were going to bring wOBA, too. I understand it was available elsewhere (FirstInning, maybe?), but it’ll be nice to have great hitting and pitching metrics at the same site.

So, a somewhat political question: How much do you guys believe in wOBA? Is Tango’s formula, right now, the best “one-stop” offensive number that we have?

I think so yes

and you were right about FirstInning having wOBA. I always forget they have MLB stats as well, I spend so much time there looking at minor league numbers.

Right on.

Well, good luck with the site. We’re spoiled.

I'm already excited for StatCorner 2.0

Wilder. - August 14, 2008 reply

I'm excited for StatCorner 1.0

This thing is still in beta, remember?

Oh yeah, whoops.

But com’on, you know 2.0 is going to have some amazing new features.

Wilder. - August 14, 2008 reply

But I like it...and Jeff has come to love it. I don't see how this is a bad suggestion.

kentroyals5 - August 14, 2008 reply

This is really cool.

Good job guys!

Fin - August 14, 2008 reply

Another question:

As far as the batted-ball data, you guys use MLB’s stuff and not Fangraphs’ data, right?

Where does Fangraphs get its data? BIS?

Right.

Yes.

Felix Hernandez, 2008:

RV600: 336.0

Okay, so this is awesome

I’ve just started reading the book, so appreciation is at an all-time personal high.

The site’s great guys. Functionally fantastic. I love being able to compare tRA+ season-on-season (GS52 dwon from 160 to 103?!). Having all this stuff centralised is great. Oh, and being able to flip between RP and SP data is a godsend.

I might have missed this in one of the writeups, but RA stats only seem to appear for 2007 data, no other years. Presumably that’s intentional?

Keep up the fab work, looking forward to continued development.

MarkE - August 14, 2008 reply

SB 2.0wned

can’t control the itcalic function

MarkE - August 14, 2008 reply

RA will end up being fully listed.

You are superheroes

Nick S - August 14, 2008 reply

Congratulations Matthew.

I sent the site to my uncle. He’s a statistician at UWisc and has been writing papers for the journal of quantitative analysis in sports (about baseball). He should appreciate all you guys do.

CapSea - August 14, 2008 reply

Wow Awesome guys.

Good work!

brayden04 - August 14, 2008 reply

I love it.

Navigation is extremely quick and easy!

ThundaPC - August 14, 2008 reply

Thank you.

Woot

Milendriel - August 14, 2008 reply

The site is amazing!

One question – did you explain somewhere why your traditional stats appear to differ from what’s at other sites, like baseballreference.com?

For example, you have Ichiro’s Avg. for 2002 at .297, whereas B-R has it at .321. Is this one of the things that’s still in Beta mode?

Nadingo - August 14, 2008 reply

Hah.

I kind of wish we’d noticed this earlier… that needs fixing.

Well so much for being productive at work today, I have a shiny new distraction.

Congratulations, fellas! Hope this venture is a success all around.

some work still to do

Hey I love the site (in fact it’s already made my list of bookmarked baseball sites along with LL, USSM, and Fangraphs), but there are some issues.

For instance, Ichiro’s stats are wrong. His traditional stats thing he hit below .300 his first two years in the league, which clearly didn’t happen. Might want to check that sort of thing.

Otherwise, great work and great site, I will be a consistent visitor.

abender20 - August 14, 2008 reply

This is tits.

Genuinely impressive

I have no idea how most of those numbers work, but I look forward to the offseason to actually learn.

kericr - August 14, 2008 reply

What kind of additional functionality are you looking to add before you call it production-live?

kericr - August 14, 2008 reply

phenomenal

Do you have an ETA on regressed tRA (seems like all of the related studies would be a first in the sabrmetric community)? Hoping you take into account stuff like GB pitchers allowing fewer hits on GBs than FB pitchers.

Any chance of a tRA number on the ERA scale for easy comparison?

Any chance of a tRA for hitters (like PrOPS) that’s properly regressed? (Obviously different regression for hitters than for pitchers.)

I believe tRA is meant to be read like ERA

5+ = ugh
4-5 = meh
3-4 = not bad
lower 2 = WOOOOOOOOOOOOOOO!

league-average tRA appears to be about a quarter run higher than league-average ERA

because well, it’s tRA, tERA. and that makes sense because the whole earned run/unearned run things DOESN’T make a lot of sense.

the mental adjustment isn’t difficult, but, you know, most people aren’t going to do it.

Yeah, it's on a R/9 scale.

We’re having philosophical difficulty with putting it on an ERA scale, but it’s something we’re thinking about.

tRA+ kind of does the trick

Regressed tRA (tRA*) already exists, but it's such a pain to code that we haven't got around to putting it on the site.

It doesn’t look at the change in out% on things like FBs and GBs for different pitcher types, but that’s a good suggestion and we’ll take a look at it.

Maybe you’ll get tERA but I don’t really like the idea.

The problem with hitters is that they throw off run and out values when it’s the individual hitter over an entire year. Ichiro does b etter on GB, worse on FB, etc. So using a tRA type measurement may not be very sensible.

can you point me to an explanation of tRA*? thanks

I don’t get your point about hitters being different from pitchers. I realize you don’t really want hitters’ tRA, you want to use the pieces of tRA to get tOPS or twOBA. But just like not all pitchers have the same run values for FBs (isn’t this something you would regress in tRA*?), neither do hitters. you just regress pitchers different from hitters, because hitters have more of a spread in skill than pitchers in most classical DIPS things like BABIP, IFF/FB, HR/BIA.

Each hitter is going to have a ~unique run value for his GB, LD, and FB's

Let’s say that, over the next month, Ichiro and Albert Pujols each put up the following BIP distributions:

60% GB
20% FB
20% LD

What you are proposing would treat both players equally, even though it should be pretty obvious that their numbers in the end would be substantially different.

By facing a large sample of hitters, most of the noise gets teased out for pitchers, and the run values of their balls in play approach the average. It is not so for hitters.

Bookmarked.

Looking forward to the development of this site.

However, there are layout issues that I think could be improved going forward. And not to nitpick, but why, under the Brewers team entry, is CC Sabathia listed twice? (Once with and once – properly – without the periods after his initials.)

esoteric - August 14, 2008 reply

because he appears twice on MLB's XML rosters.

I’m trying to stop that from happening, but if you could tell MLB to stop listing him twice, that would help :)

The park factors are nifty.

The Trop's HBP factor is 1.16

Damn kids and their aggressiveness.

Ironic.

Oh man not cool.

That sucked.

Why?

Unfocused rage

As much as you've worked on this, I didn't expect you to roll out the live beta so soon

Nicely done, men.

Let me know if there’s anything I could possibly help with w/r/t compiling numbers for the minors.

This site blows!

And by blows, I mean totally rocks my socks.

You guys rule.

Goose - August 14, 2008 reply

I will visit often, and will inform my friends of this resource!

I’ve had a couple of questions on some things for a while, and this seems like a decent place to ask them.

1) Recently, Jeff was lamenting the need for human coding of batted ball types; this is something hit fx is going to fix, but for now, LD% is a huge determinant of any number of stats – from PrOPS to, obviously, tRA. So the question is, how do we know that this coding is being done well? I thought of this when scanning through the SF Giants rotation on THT. Each pitcher saw a distinct jump in his LD% in 2008 over 2007, even as their RA or FIP dropped (Matt Cain is waaay out of line with his career ave., but is having fairly normal results. Lincecum’s given up way more LDs, but lots fewer runs). This can easily be explained by random chance; it’s not like Lincecum’s giving up 50% LDs and an RA of 2.9. But it got me wondering about ‘park’ effects for LDs, which is really more like saying the effects of home scorers at each park.
I checked out the piece that David Gassko wrote on park effects for batted balls, which is great stuff, but it didn’t mention LDs. Maybe it’s in there but wasn’t mentioned in the article. If that’s the case, what’s the YTY correlation for LD rate? Do you see any large (define that how you will) swings in LD% by park from year to year? More basically, do you think that the ambiguity around LD versus FB versus ‘fliner’ hurts tRA in any way? Have you had a chance to compare tRA in 2007 to tRA (so far) in 2008? More or less stable than FIP?

2) Might there be a way to rank parks based on batted ball types? Gassko had those mini ‘leaderboards’ in the THT article; it might be nice just to have a sortable table where you could rank each park on IF or LD or whatnot.

marc w - August 14, 2008 reply

leaderboards on hopefully everything will roll out in phase 2

It’s on the agenda, big time. After fixing the stat numbers to be 100% perfect, it’s pretty much all I’ll be working on.

Congrats

The site works very well.

G_ - August 14, 2008 reply

Awesome site

Though I noticed that Tex is still listed as a Brave and not an Angel. I’m guessing it’s because you set the rosters before all the trades? If so, you might want to double check those.

Heh, they have like 50 stat-checkers for their beta just by posting the link here

Haha and Richie is still a Mariner.

It's sorted by team name right now unfortunately

and the players appear on every team they’ve played in that season for now.

Thanks for the catch on Tex though, helped me spot an error in the player loading algorithm.

http://www.statcorner.com/pitcherRP.php?id=406550

Heh

But you left out his Awesome Factor!

He has an AF of like 672.4 !

But pitcher losses!

5.6% swinging strike.

Thanks, Pudge.

There's a gif of that somewhere, right?

I believe so, yes.

Ahahaha yes I am so glad I went to that game.

This fills me with joy.

That terrible delivery, no velocity… yet he swings right on through it.

Thank you.

His delivery is awesome!

He throws that ball from right out of his ear hole! It’s better than a hitch in the leg kick.

dpseadv - August 14, 2008 reply

I don't know why everyone is criticizing Batista

per tRA he’s been unlucky this season!

Maybe tRA is kinda like BABIP for hitters.

It only applies if the pitcher really is major league quality. Batista might be completely done as a pitcher so advanced metrics don’t apply.

Per tRA he is one of the worst pitchers of the past 5 years

But he has a negative O-xO!

unlucky

His tRA is 7.67

I think it applies.

Ah.

I’m saying this to myself: WHOOSH!

Awesome job guys.

I was wondering if there is a place (or a place coming) where you could compare the team averages.

I’m just curious to see how certain teams that say they are interested in things like OBP, etc are doing at achieving that. And if teams that are more interested in “grit” are good because they lucked into guys who are good in those categories.

GhettoBear04 - August 14, 2008 reply

Thanks so much for doing this you guys!

Do you have any desire to jump into the Defensive Stats game?

Sec 108 - August 14, 2008 reply

We are contemplating using tRA to approximate defence.

Very cool idea.

Would the basic application of that be showing how the defense is affecting the difference between actual results and expected results?

I wish I had more free time because this stuff looks way too fun. Keep it up.

Sec 108 - August 14, 2008 reply

definitely!

tRA compared to actual RA would contain a bunch of noise, but if you compared outs and expected outs, you’d have something. also expected SLG_BIP and actual SLG_BIP. good idea.

What, no wOBA numbers for pitchers?

They hit too.

Llewdor - August 14, 2008 reply

Hmm?

Felix as a hitter

Are career totals coming soon?

Cool, but I don't appear to be able to reach that through the player name search.

Llewdor - August 15, 2008 reply

Ok, we'll put an option in to jump to hitting/pitching stats on the player pages if said stats exist.

Bear with me though, because Matthew’s in Chicago.

Graham MacAree - August 15, 2008 reply

Thanks for this fantastic resource.

Robert - August 14, 2008 reply

Seconded.

Hurry up and get to work. Lazy

I remember being ecstatic that I finished "D"

Than I remembered that dear god that’s only the fourth letter

Robert - August 14, 2008 reply

And I'd already done A-C.

Lazy!

You've tricked Robert into being your slave

it really is a sitcom.

We need a laugh track

That's so retro.

[LAUGH TRACK]

Laffey track

This episode was taped balfour a studio audience

You're not setting a very good example.

Oh dear.

I think you mean Deer

johnbai - August 14, 2008 reply

Huzzah!

mikeA - August 14, 2008 reply

And thank you for publishing league averages with your stats

that’s really helpful.

pdb - August 14, 2008 reply

This.

i agree -- awesome glossary

I've e-mailed this site to some friends

and already plan on having it replace baseballreference as my go-to site for stats. Right on, guys.

katal - August 14, 2008 reply

not that i disagree, but i think B-Ref is underrated if just for their BtWins stat

Great stuff. Just a few suggestions

Have outcomes only go to 2 decimal places.
Have BABIP only go to 3 decimal places.
On the hitters/batters page an option to sort differently would be nice. (IP, etc.)
Sortable stats for the league. Who has the best tRA? Even a leaderboard would be nice.

Thanks guys.

Edgar for Pres - August 14, 2008 reply

You spelled available wrong in your definition of RA

Now I just need a firefox plugin so it can replace baseball-reference in the search box

That would be nice.

Good call Corco.

Goose - August 14, 2008 reply

Nice piece of work!

I’ve been catching references to the work you’ve been putting into this all season, I thought it was an in joke or something. You should charge money!! I’m going to be geeking out on this until Christmas.

dpseadv - August 14, 2008 reply

also you will eventually need fielding stats if you hope to become a "go-to" place for baseball stats

Matthew's Subsequent edit: Stats for years before 2007 should now be fixed.

I thought Pujols' 2004 looked funny.

Grammar error

on the Authors page. “Currently residing in Seattle, he can reached via e-mail.”

Should read: “Currently residing in Seattle, he can be reached via e-mail.”

Sorry I’ve been fixing string errors at work for a couple days now.

bluemax - August 14, 2008 reply

No need to apologize.

Because so much of is written inside the code, it makes spell and grammar checking impossible except by eyeballing it. Please let me know any such errors.

Hardcoded strings

tsk tsk. Although maybe for websites the rules are different.

bluemax - August 14, 2008 reply

Shouldn't be.

pdb - August 14, 2008 reply

I always type out my text before coding it

though I should spell check it before doing so….

Awesome liftoff guys

just one little reaction…
So much green… is that really necessary? Any chance a graphic designer will be employed?

But definitely keep it to a single screen on the front page, and keep the same simple/effective vibe. That part rocks.

johnbai - August 14, 2008 reply

average vs. replacement question

If going from runs above average to runs above replacement, how many runs is each PA worth? I know Sean Smith’s shortcut is to assign two wins per 700 PAs. Anyone have a better number or good explanation how to figure it out?

league adjustment

Am I right that there’s no adjustment for the difference in league-quality between the AL and NL? (Not that any other stat sites do this.)

given the half-win difference that most people use for converting between NL and AL...

… how many points of wOBA would this be?

Would sortable leaderboards be possible?

Phase 2

They’re coming eventually.

Excellent.