Comments on: Colby Lewis: 2.22 ERA in 7 postseason starts This and that about baseball stats. Tue, 16 Jul 2013 17:01:55 +0000 hourly 1 By: PhilM Fri, 21 Oct 2011 22:52:55 +0000 A bit off-topic, but I'll offer:

Career ERA+ is tricky, and I've posited a version of this discussion when the change in the formula was being contemplated. Because the denominator is the individual ERA, summing is complicated. I do it this way:

- find the "percentage better than average " by calculating (ERA+ minus 100)/(ERA+) -- so an ERA+ of 200 is 50% better than average, since an ERA of 2.25 in a league of 4.50 is 50% lower
- weight the "percentage better" by innings pitched
- sum it all, divide by total innings pitched, for "overall percentage better"
- translate that back into the ERA+ formula we recognize by dividing 100 by 1 minus the "overall percentage better"

This is often slightly different from the "career" value found on B-R pitcher pages, because that's calculated using total earned runs and total league runs (which are just earned runs times ERA+ divided by 100 for each season) -- it can be skewed by extreme high-run or low-run environments, as runs in those seasons get more than their fair impact on the sums.

I'm not sure that's in English, but that's how I figger it!

By: DavidRF Fri, 21 Oct 2011 20:45:14 +0000 @12
I don't know. You'd have to figure out the 'run context' of the series.

NL ERA: 3.81

AL ERA: 4.08
TEX PPF: 109

... but Park Factors are deceptively complex. They assume a team plays road games in all the other parks in the league and doesn't have to face its own batters, etc. And they are calculated over multiple years to smooth the data. I don't know how to apply them to a short series played by two teams.

By: Lawrence Azrin Fri, 21 Oct 2011 20:24:48 +0000 @11/ DavidRF -
Thanks, I was trying to say something like that, but in English, not mathmatically. So how do we do this for the postseason,or is it not defined yet?

By: DavidRF Fri, 21 Oct 2011 19:57:41 +0000 @10
I believe an equivalent "lgER" for the pitchers IP for each year has to be calculated. Then the career contextual "lgERA" for a pitcher is 9*sum(lgER)/sum(IP).

I remember reading something like that somewhere, but I can't seem to find it.

By: Lawrence Azrin Fri, 21 Oct 2011 19:27:44 +0000 BSK/ - "@9/ LA- I sort of assumed that career ERA+ is simply the compilation of the individual season ERA+'s. ... "

Yes, I also assume that you scale career ERA+ proportionately by IP, and not just add the season totals up and divide by the seasons total,in the same way that career BA is not figured by that method. However, career BA is simply {hits/at bats}; career ERA+ is more complicated.

Any place to go here on B-R for a definition?

By: Lawrence Azrin Fri, 21 Oct 2011 19:15:31 +0000 John Autin -

"@6, Lawrence -- Perhaps John Q. just meant Boston's poor play in the '88 ALCS, when they were swept by Oakland..."

John A., I see your point. My apologies to John Q.; it would be easy to ignore Hurst's fine pitching in the 1988 ALCS when he went 0-2, despite a 2.77 ERA (4 runs/ 13 IP).

That 1988 ALCS loss was a disappointing coda after "Morgan Magic" after the All-Star break.

By: BSK Fri, 21 Oct 2011 19:06:49 +0000 LA-

I sort of assumed that career ERA+ is simply the compilation of the individual season ERA+'s. So a 200 IP season with 150 ERA+ and a 100 IP season with 100 ERA+ would work out to 133 ERA+ for career (200*150 + 100*100 / 200 + 100). Is that how it is calculated?

Postseason numbers seem much tougher, for the reasons you offered.

By: John Autin Fri, 21 Oct 2011 18:38:26 +0000 @6, Lawrence -- Perhaps John Q. just meant Boston's poor play in the '88 ALCS, when they were swept by Oakland.

Speaking of Bruce Hurst ... Wow, he won 2 CG in the '86 postseason in which he allowed 10+ hits. There've been just 11 such games in the past 60 years. Since 1920, only Hurst, Tommy Bridges and Waite Hoyt have more than 1 such game (Hoyt has 3).

By: Lawrence Azrin Fri, 21 Oct 2011 18:18:24 +0000 @2/ BSK -
"Can ERA+ be calculated? Or is that too hard because the games are happening across several postseasons?"

ERA+ for regular seasons is also calculated for more than one season, as a career stat.

But the problem is, would the baseline for postseason ERA+ be:
- regular season ERA that year?
- postseason ERA that year?
- postseason ERA in just that series that year?

I think I just answered your question.

@3/ John Q -
"... Bruce Hurst's terrific post-season pitching just gets lost because of the Red Sox's terrible play. His contributions to the '88 Red Sox are completely forgotten."

John, Red Sox fans certainly appreciated Hurst when he was with them; a lot of them did not want him to leave as a free agent after the 1988 season. He would have been the 1986 WS MVP, if not for the 10th inning of Game 6. As I matter of fact, I distinctly remember the Shea Stadium scoreboard flashing "Congratulations Bruce Hurst, 1986 World Series MVP" for a few seconds just before you-know-what happened.

I am not sure what you mean by "because of the Red Sox's terrible play", even if you are specifically referring to their collapse this September. That month doesn't negate what Red Sox players did before that for 111 years.

By: Ed Fri, 21 Oct 2011 18:05:00 +0000 @4 Doug

Last year Lee made 5 post seasons starts, striking out 47 and walking only 2. How's that for incomprehensible? A 23.5 K/BB ratio while facing the best teams in baseball.