You are viewing the chat in desktop mode. Click here to switch to mobile view.
X
NLDS Games 4 Chat
powered byJotCast
Prev Page View Full Chat Next Page
Ben Clemens
8:45
Way better than middle/middle.
8:46
But eh, not as good as I first thought.
It's hard to call a pitch out of the zone a bad location, but I guess if there was someone you didn't want to miss high with, it's Zim.
8:47
That took me embarrassingly long to remember, had to look up his career ISO/BIP heatmap:
8:48
As well as an all-time favorite of mine, an article I wrote about how Zimmerman hits into all the double plays:
bosoxforlife
8:48
My problem with analytics is that the true believers can't understand when their predictions don't turn out as expected. Like how did Zimmerman hit that high heat? "You can't predict baseball Suzyn."
Ben Clemens
8:49
I would say that they're doing their true believing wrong, then. If you don't constantly doubt your own predictions, you're not doing analysis right.
8:50
As Mitchel Lichtman put it: The sabermetrician’s credo: “I’m not sure, this is why I’m not sure, and this is roughly how not sure I am.”
bosoxforlife
8:52
I wish I got that feeling when I read a lot of the analytics. I tend to read, between the lines, a much more aggressively certain tone.
Ben Clemens
8:52
Yeah, I will say that personally I struggle to find the right line in my articles between being interesting and being responsible. I think I tend to err on the side of responsibility but I empathize with the desire to say something meaningful.
Josh Herzenberg
8:53
I find the process of analyzing to be far more purposeful than the result it brings out
Ben Clemens
8:53
I think the real secret ingredient is humor, which makes your stuff fun to read without you constantly having to invent new things.
That's why Dan has been able to do this for so long.
Josh Herzenberg
8:53
Or dogs
Ben Clemens
8:53
Look the secret to having readers like you, not to FG internal politics.
Belli Flop
8:53
And Jeff
Ben Clemens
8:54
Yeah Jeff was the king of this. It really sucks that he left, even though that indirectly helped me get my job.
Derek
8:54
I think it's just that sabermetricians believe they have the right approach (which, as has been borne out time and again, they do) more than they think they are *right*
Josh Herzenberg
8:54
That is my perspective. Doing it the right way leads to a better chance of the result being right. Doesn't mean the results will be right any percent of the time
8:55
Predicting as such would be disingenuous
Wire Fan
8:55
How long does Scherzer need to go for the Nats to feel comfortable about winning this game?   9 innings?
Josh Herzenberg
8:55
At least
Ben Clemens
8:55
Yeah I'm with Josh. I'm a strong believer that my macro process is good, which means absolutely nothing in a given analysis, but hopefully will add up over time.
Dcrl
8:55
Analytics bashing aside, I gotta say Francoeur is a lot less bad and annoying than most of the other former players who do color on national broadcasts.
Ben Clemens
8:56
Yeah he's climbing in my estimation. I heard him on a few Braves broadcasts this year and liked him, so I was surprised that he sucked early on.
But I think he was just getting his bearings in a new booth with new teams to cover.
Though he doesn't love analytics he loves baseball.
And that is by far the most important thing to me.
Josh Herzenberg
8:56
Yeah, at the micro level, the unpredictability of human actions creates an almost impossibly predictive environment. But eventually, sample sizes turn into macro. That becomes predictive
Ben Clemens
8:57
Honestly Ernie Johnson has also grown on me a ton. He seemed early on like he was catching up with all the players (which he maybe was!) but he's warming up.
Adam2
8:57
Could also be preemptive defensiveness because the other side will highlight times the outcomes go against the analytics as evidence against the analytics
Josh Herzenberg
8:57
Yep. I can empathize with that, though. I used to not be able to as much, but there are a lot of dynamics in play. Humans are playing the game
Ben Clemens
8:58
I mean I do think that there's bad analytics. If you stubbornly believe that you can capture everything with data, you'll be shocked at the stuff you keep missing
Wire Fan
8:58
My biggest pet peeve is the expression "good process,  bad result", especially when someone is evaluating their own process (obviously they think it is good)
Ben Clemens
8:58
Misuse of tools is also super dangerous. People think that because they have numbers they are right automatically, even if they use those numbers wrong.
Josh Herzenberg
8:58
The challenge/follow-up here is "please define the difference between good and bad process"
8:59
Establish that baseline before conducting the analysis
Ben Clemens
8:59
There was a great Twitter thread between Tango and Nate Silver about the biggest errors people make in statistics.
Connecting…