A Realistic View at 2022: Part I

[mvp 78]

Comfort level? Yes.

 

CERA? Highly doubtful.

 

If that stat had any meaning, ESPN would still be using it on their leaderboards. They realized like 3 years ago that it was useless. https://www.espn.com/mlb/stats/fielding/_/position/c/sort/catcherERA

[moonslav59]

Comfort level? Yes.

 

CERA? Highly doubtful.

 

If that stat had any meaning, ESPN would still be using it on their leaderboards. They realized like 3 years ago that it was useless. https://www.espn.com/mlb/stats/fielding/_/position/c/sort/catcherERA

 

Who cares what the reason for CERA to be so different?

 

I never argued about knowing the reasons for such wide and consistent variances, and in fact I have often said it's "not just about pitch-calling."

 

If it's 100% comfort as to why a certain pitcher does better with catcher A over B, it's a reason to keep the better tandem together, and the CERA will show why.

[moonslav59]

Comfort level? Yes.

 

CERA? Highly doubtful.

 

If that stat had any meaning, ESPN would still be using it on their leaderboards. They realized like 3 years ago that it was useless. https://www.espn.com/mlb/stats/fielding/_/position/c/sort/catcherERA

 

Comparing CERA of catchers from different teams exposes a complete lack of knowledge of what CERA means and is useful for.

[mvp 78]

Who cares what the reason for CERA to be so different?

 

I never argued about knowing the reasons for such wide and consistent variances, and in fact I have often said it's "not just about pitch-calling."

 

If it's 100% comfort as to why a certain pitcher does better with catcher A over B, it's a reason to keep the better tandem together, and the CERA will show why.

 

CERA doesn't "show" anything. The sample size is too small. It's just noise and random.

[mvp 78]

Comparing CERA of catchers from different teams exposes a complete lack of knowledge of what CERA means and is useful for.

 

It has no use. That's why they scrapped it.

[moonslav59]

CERA doesn't "show" anything. The sample size is too small. It's just noise and random.

 

career sample sizes are over 100 IP'd by multiple catchers. That's not small.

 

It's been "scrapped" because it is useless comparing players from different teams, which is what stats are mostly used for.

 

Not many people are interested in what a back-up catcher on a team might be adding value on?

 

We've been giving this way too much air time.

 

It's just a factor- not a majority one. Not even a plurality one.

 

I think it is significant. You think it's zero.

 

It's probably somewhere in between- like minimally significaant.

 

[mvp 78]

career sample sizes are over 100 IP'd by multiple catchers. That's not small.

 

It's been "scrapped" because it is useless comparing players from different teams, which is what stats are mostly used for.

 

Not many people are interested in what a back-up catcher on a team might be adding value on?

 

We've been giving this way too much air time.

 

It's just a factor- not a majority one. Not even a plurality one.

 

I think it is significant. You think it's zero.

 

It's probably somewhere in between- like minimally significaant.

 

 

Is Vaz the same catcher he was in 2015? No, he's probably better. Comparing random years of Vaz's career to random years of Tek's just doesn't make much sense. Pitchers change over time. These aren't apples to apples comparisons.

[moonslav59]

Here are some interesting reads on CERA. There is not much talk about OPS against among different catchers on the same team.

 

https://members.tripod.com/bb_catchers/catchers/cera1.htm

 

https://tht.fangraphs.com/classic-tht-annual/do-catchers-have-an-era/

 

This one argues it is not significant, but it attempts to compare catchers against other team catchers:

https://www.baseballprospectus.com/news/article/1489/aim-for-the-head-simulating-catchers-era/

 

 

 

[moonslav59]

Is Vaz the same catcher he was in 2015? No, he's probably better. Comparing random years of Vaz's career to random years of Tek's just doesn't make much sense. Pitchers change over time. These aren't apples to apples comparisons.

 

It's not random. I get how comparing a pitcher from 2015 with the same pitcher in 2017 is fraught with problems, but I have done year by year- same pitchers- same years- 2-3 catchers. True, there are other variables at play, such as who was the opposing line-up and what field were the games played on with one catcher vs the other, but when you do a study based on 5-6 years and nearly every year shows vast disparities all in one direction, I think one can determine the trend has to do with something one catcher is doing better than the other.

 

To say there is zero effect from one catcher and another really blows my mind. Sorry, it just does.

 

I can see you and others thinking I value it way to much as "mind-blowing," too, but I will argue that just because I talk about it a lot does not mean I think it is the most important thing a catcher brings to his overall value.

 

I do sometimes wonder, if a 1.50 ERA disparity is indeed linked totally to the catcher's CERA related skills, I might actually be understating it's value to certain pitchers - not all pitchers, of course.

[mvp 78]

Here are some interesting reads on CERA. There is not much talk about OPS against among different catchers on the same team.

 

https://members.tripod.com/bb_catchers/catchers/cera1.htm

 

https://tht.fangraphs.com/classic-tht-annual/do-catchers-have-an-era/

 

This one argues it is not significant, but it attempts to compare catchers against other team catchers:

https://www.baseballprospectus.com/news/article/1489/aim-for-the-head-simulating-catchers-era/

 

 

 

 

I've actually read the first two before. My conclusion was that CERA is random because of SSS.

[mvp 78]

It's not random. I get how comparing a pitcher from 2015 with the same pitcher in 2017 is fraught with problems,

 

So yes, it's random then. Adding a whole bunch of words after stating there are problems is just a waste.

 

It's a random SSS.

[mvp 78]

However, there is a problem with this straight forward approach, as noted by Keith Woolner in his study published in Baseball Prospectus. The problem is sample size.

 

Agreed.

[mvp 78]

For the past 10 years or so I have subscribed to Woolner’s conclusions, but I can no longer do so. I believe his sample sizes were too small.

 

Exactly!

[mvp 78]

And what happens when sample sizes are wrong or fraught with other sample differences (opposing lineups, health of players involved, pitchers either declining or improving as time goes on, etc.)?

 

Randomness!

[moonslav59]

And what happens when sample sizes are wrong or fraught with other sample differences (opposing lineups, health of players involved, pitchers either declining or improving as time goes on, etc.)?

 

Randomness!

 

Then never use any stats.

[mvp 78]

Then never use any stats.

 

I'll stick with the good ones.

[moonslav59]

I've actually read the first two before. My conclusion was that CERA is random because of SSS.

 

Even if about 15 out of 20 small sample sizes show the same thing for 5 straight years, and the bigger sample sizes show it even more?

[moonslav59]

I'll stick with the good ones.

 

They are just as random.

[mvp 78]

They are just as random.

 

No. With CERA, you are taking partials and dividing them further. You say that CERA is useless on a macro level. Things like OPS+ are made to compare players across the league. CERA, by your definition, cannot. Other stats can utilize a larger sample size. You say that CERA gets worse the larger you make the sample. These two things are not the same.

[mvp 78]

Even if about 15 out of 20 small sample sizes show the same thing for 5 straight years, and the bigger sample sizes show it even more?

 

Your strawman is not true though.

[Bellhorn04]

Hey, remember when Cora ditched all CERA concerns in the 2018 postseason? I do. It worked out.

[mvp 78]

Hey, remember when Cora ditched all CERA concerns in the 2018 postseason? I do. It worked out.

 

Is that the same time ESPN stopped tabulating it?

[moonslav59]

Hey, remember when Cora ditched all CERA concerns in the 2018 postseason? I do. It worked out.

 

I remember Tito ditching it in 2009 by switching from VTek to VMart in the playoffs with Beckett pitching.

 

It did not work out.

 

6.2 IP 4 ER

 

Season numbers:

 

3.17 with VTek (185 innings)

7.16 with VMart (16.1 Innings)

10.32 with Kotaras (11.1 innings)

 

career:

3.68 VTek (905)

5.48 VMart (90.1)

[Bellhorn04]

I remember Tito ditching it in 2009 by switching from VTek to VMart in the playoffs with Beckett pitching.

 

It did not work out.

 

6.2 IP 4 ER

 

Season numbers:

 

3.17 with VTek (185 innings)

7.16 with VMart (16.1 Innings)

10.32 with Kotaras (11.1 innings)

 

career:

3.68 VTek (905)

5.48 VMart (90.1)

 

Tito maybe left Beckett in that game too long?

[mvp 78]

Tito maybe left Beckett in that game too long?

 

He had a 1.50 CERA through 6 innings. Seems good!

 

If only he had a decent pen of Paplebon, Wagner, Bard, Ramirez, Okajima, Saito to give the ball to...

[mvp 78]

If the 2003 team had that pen, they break the curse.

[moonslav59]

Tito maybe left Beckett in that game too long?

 

Or, brought VTek in for the 7th!

 

LOL!

[mvp 78]

His previous 5 starts:

6 innings 2 ER, VMart 3.00 CERA

8 innings 3 ER, Tek

5 innings 1 ER, Tek

7 innings 3 ER, Tek

6 innings 4 ER, Tek 3.80 CERA

 

The best outing of those 5 starts was with VMart.

[notin]

His previous 5 starts:

6 innings 2 ER, VMart 3.00 CERA

8 innings 3 ER, Tek

5 innings 1 ER, Tek

7 innings 3 ER, Tek

6 innings 4 ER, Tek 3.80 CERA

 

The best outing of those 5 starts was with VMart.

 

By what measure? 5IP 1ER is a 1.80 ERA…

[moonslav59]

His previous 5 starts:

6 innings 2 ER, VMart 3.00 CERA

8 innings 3 ER, Tek

5 innings 1 ER, Tek

7 innings 3 ER, Tek

6 innings 4 ER, Tek 3.80 CERA

 

The best outing of those 5 starts was with VMart.

 

Total randomness and small sample sizes.

 

(Also, the 1 ER in 5 IP was better.)