Posted by 06gsp on 1/24/2022 7:16:00 AM (view original):
just4me you're putting too much faith in the results of that one test league. there is no doubt the extra errors you get with deadballers make a big difference, and I certainly don't think deadballers are better than modern pitchers across the board, but before dynamic pricing Summers and his ilk were underpriced because of their HR rate relative to the rest of the league.

compare 1909 Summers to 1967 Fergie Jenkins. similar WHIP#, similar salaries pre-dynamic pricing, and Jenkins has a HR/9+ of 73 to Summers' 70. Except Jenkins gives up 37 HRs on average in the performance history and Summers gives up 10. There is no way in hell the worse defense behind Summers is going to come close to making up for 27 extra home runs.

The one test league was just one where others finally got to see results, but I've tested it pretty thoroughly since the last update in like 2009 or whenever it was... and why I wanted to use a TWISL league for a similar test to help others understand the same thing. Especially as so many newer owners had come onboard since then that were starting down that same path on deadball pitcher dominance and not understanding the effects of normalized fielding.

I'm not saying the pricing is (was) perfect, and agree it should look more at # instead of +, but as a whole, it does compensate well for the errors against HR, $ being equal. That said, looking at a R/9 factor for Summers vs Fergie and their PH, Summers has a 1.39 WHIP and 3.88 ERA and Fergie has a 1.40 WHIP and a 4.53 ERA. Just doing quick monkey-math, the increased errors on Summers over Fergie should result in somewhere around 11-12 extra unearned runs over Fergie's unearned runs, so on a R/9 basis, that would put him around 4.12-4.15 + their common UER, where Fergie is at 4.53 + their common UER. So roughly a 0.40 gap on R/9.

However, if your opponent is more likely to have Summers, you can decrease that gap and maybe even erase it by stacking deadball hitters on your teams with pitchers like Fergie (and since you won't be hitting HR anyway a - HR park), if they also have deadball hitters playing defense behind Summers, you could easily swing that gap in your favor, but even if they don't and have good glove modern fielders, you playing deadballers will likely add another 0.33-0.5 R/9 to Summers total. Bringing him right in-line with Fergie pre-ballpark impact. It's much closer than it looks even with this and I'd still go with Fergie over Summers personally (though, honestly, neither fit the profile of a pitcher I would draft).

Just4me, did your test league include position players with a perfect fielding percentage? I dont think those players will make an error -regardless of who is pitching and hitting- if they aren't fatigued.

One of my OL teams just finished the regular season.
(Motorboating Morganna)

Target Field, rotation of Willett, Dutch Leonard, Bob Rhoads.

99-63, committed 196 errors, 129 unearned runs, only 122 + plays. Team ERA 3.14

Sam Wise committed 47 errors at 2B, Max (or was it Drew) Carey committed 38 errors in CF...

I don't know what to think of this. Would modern pitching really help me cut down significantly on errors? I know D fielders aren't preferred, but yeah, I need a few better gloves for sure...

We ARE going to have more test leagues, I've been too busy to put something together...in TWISL we only tested Deadball vs Modern pitchers, we haven't tested Deadball vs Modern fielders yet... haven't seen anyone moving on this yet either...is on my list.

We'll need to do Identical Pitching Staffs (deadball and modern) vs All Deadball Fielders vs All Modern Fielders

Posted by DoctorKz on 1/24/2022 7:20:00 PM (view original):
One of my OL teams just finished the regular season.
(Motorboating Morganna)

Target Field, rotation of Willett, Dutch Leonard, Bob Rhoads.

99-63, committed 196 errors, 129 unearned runs, only 122 + plays. Team ERA 3.14

Sam Wise committed 47 errors at 2B, Max (or was it Drew) Carey committed 38 errors in CF...

I don't know what to think of this. Would modern pitching really help me cut down significantly on errors? I know D fielders aren't preferred, but yeah, I need a few better gloves for sure...

Infielders performed better while outfielders performed worse.

Name /Fielding# /Motorboating Fielding
3B Bill Hassamaer 1894 /.916 /.958
1B Cecil Cooper 1981 /.990 /.997
SS Ozzie Smith 1993 /.967/ .967
2B Sam Wise 1887 /.939 /.949
C Steve O'Neill 1920/ .980 /.989

OF Tommy McCarthy 1893/ .948/ .933
OF Max Carey 1921/ .962/ .925
OF Larry Walker 1994/ .967 /.958

Posted by ncmusician_7 on 1/24/2022 6:10:00 PM (view original):
Just4me, did your test league include position players with a perfect fielding percentage? I dont think those players will make an error -regardless of who is pitching and hitting- if they aren't fatigued.

I did not directly test that, though in the test league referred to if anyone had a 1.000 FLD player, they would have had them on both their modern and deadball teams. And I definitely used '89 Hojo on some of my old normalized fielding tests which gave me the confidence to tell zubinsum (whom was one of the best numbers guys here) his value charts were missing key data that erased the over/under characteristics.

It would be fun to directly test that though. I'm running some other tests at the moment, but maybe when I finish those up I'll take a league and run a 12/12 with all 1.000 FLD guys with modern pitchers and deadball pitchers teams evenly distributed... etc...

Here's an example of what j4m is talking about and doctors, I think answers your questions on if it would take down errors - and is also something I'd be down to test again in another TWISL league.

League #: MLB131914

My team: Making You Earn A++ V.19
sford's team: Petco Ten Two

Before I break this down, let it be known I am not saying anything is wrong with sfords team (they won 111 regular season games, and is facing off against my team in the WS as we speak), but am rather just highlighting some of the things J4M mentions above.

The two teams had very similar offensive/defensive makeups:

MYEA++- Chargingryno			PTT -sford
Clements, Jack	1891	C	Schang, Wally	1921	C
Connor, Roger	1894	1B	Connor, Roger	1894	1B
Kelly, King	1887	2B	Kelly, King	1887	2B
Baker, Frank	1910	3B	Joyce, Bill	1897	3B
Templeton, Garry	1978	SS	Bancroft, Dave	1923	SS
Carey, Max	1921	CF	Carey, Max	1921	CF
Smith, Reggie	1969	LF	Chapman, Ben	1936	LF
Seymour, Cy	1904	RF	McCreery, Tom	1899	RF

Both teams featured primarily deadball hitters with high range. The only difference in the makeup of our teams were the pitching staffs. My team features all modern pitchers; versus sfords featuring all deadballers. This isn't the main point of this, but I'll also notate the defensive differences between our common players...Max Carey and King Kelly primarily:

Making You Earn A++ - Chargingryno									Petco Ten Two - sford
Player	SN	Pos	G	E	+	-	Fld%		Player	SN	Pos	G	E	+	-	Fld%
Kelly, King	1887	2B	162	24	24	0	0.972		Kelly, King	1887	2B	147	39	24	0	0.957
Baker, Frank	1910	3B	162	30	22	1	0.949		Joyce, Bill	1897	3B	152	46	16	0	0.922
Templeton, Garry	1978	SS	162	49	27	0	0.943		Bancroft, Dave	1923	SS	143	61	21	0	0.932
Carey, Max	1921	CF	162	8	50	0	0.982		Carey, Max	1921	CF	151	19	32	0	0.959

In each position listed, you see a pretty clear difference in errors committed, despite having similar (or identical) players.

Defensively, sfords team was the worst in the league with 252 errors and 18 - plays. Accounting for their 141 + plays, they were -129 on defense.
My team ranked just below league average in errors with 169 and 12 - plays. Accounting for my 183 + plays, we were +2 on defense (best in the league)

sfords team ERA was 2.38 playing in petco, but had an R/9 of 3.23. Taking into account the park factor (.82), you can estimate his neutral park R/9 to be 3.93 R/9.

My teams ERA was 3.80 playing in Hilltop, and had an R/9 of 4.48. Taking into account the park factor (1.14), you can estimate my neutral park R/9 to be 3.93 R/9.

Salary for sfords pitching $40.2m
Salary for my teams pitching $39.4m

Sford is one of the best owners on this site, and he knows how to build a good team. But after all was said and done, $ for $, his team that led the league with a 2.38 era had the same neutral R/9 as my team that had a 3.80 era.

Example 2:

From the same league - thunder1008's team had a similar theme:

BMR -Thunder1008
Kerins, John	1887	C
Rose, Pete	1981	1B
Frisch, Frankie	1925	2B
Wills, Maury	1967	3B
Templeton, Garry	1980	SS
Carey, Max	1921	CF
McGee, Willie	1990	LF
Marte, Ketel	2019	RF

Player	SN	Pos	G	E	+	-	Fld%
Frisch, Frankie	1925	2B	159	45	26	0	0.954
Templeton, Garry	1980	SS	141	51	38	0	0.942
Carey, Max	1921	CF	162	20	40	0	0.957

Thunders team ranked 3rd worst with 196 errors and 26 - plays. Accounting for their 134 + plays, they were -88 on defense.

Thunders team ERA ranked 2nd in the league behind sford at 3.09 in safeco (his 104 wins were also 2nd best), but had an R/9 of 3.85. Taking into account the park factor (.88), you can estimate his neutral park R/9 to be 4.37 R/9.

Posted by chargingryno on 1/24/2022 11:21:00 PM (view original):
Here's an example of what j4m is talking about and doctors, I think answers your questions on if it would take down errors - and is also something I'd be down to test again in another TWISL league.

League #: MLB131914

My team: Making You Earn A++ V.19
sford's team: Petco Ten Two

Before I break this down, let it be known I am not saying anything is wrong with sfords team (they won 111 regular season games, and is facing off against my team in the WS as we speak), but am rather just highlighting some of the things J4M mentions above.

The two teams had very similar offensive/defensive makeups:

MYEA++- Chargingryno			PTT -sford
Clements, Jack	1891	C	Schang, Wally	1921	C
Connor, Roger	1894	1B	Connor, Roger	1894	1B
Kelly, King	1887	2B	Kelly, King	1887	2B
Baker, Frank	1910	3B	Joyce, Bill	1897	3B
Templeton, Garry	1978	SS	Bancroft, Dave	1923	SS
Carey, Max	1921	CF	Carey, Max	1921	CF
Smith, Reggie	1969	LF	Chapman, Ben	1936	LF
Seymour, Cy	1904	RF	McCreery, Tom	1899	RF

Both teams featured primarily deadball hitters with high range. The only difference in the makeup of our teams were the pitching staffs. My team features all modern pitchers; versus sfords featuring all deadballers. This isn't the main point of this, but I'll also notate the defensive differences between our common players...Max Carey and King Kelly primarily:

Making You Earn A++ - Chargingryno									Petco Ten Two - sford
Player	SN	Pos	G	E	+	-	Fld%		Player	SN	Pos	G	E	+	-	Fld%
Kelly, King	1887	2B	162	24	24	0	0.972		Kelly, King	1887	2B	147	39	24	0	0.957
Baker, Frank	1910	3B	162	30	22	1	0.949		Joyce, Bill	1897	3B	152	46	16	0	0.922
Templeton, Garry	1978	SS	162	49	27	0	0.943		Bancroft, Dave	1923	SS	143	61	21	0	0.932
Carey, Max	1921	CF	162	8	50	0	0.982		Carey, Max	1921	CF	151	19	32	0	0.959

In each position listed, you see a pretty clear difference in errors committed, despite having similar (or identical) players.

Defensively, sfords team was the worst in the league with 252 errors and 18 - plays. Accounting for their 141 + plays, they were -129 on defense.
My team ranked just below league average in errors with 169 and 12 - plays. Accounting for my 183 + plays, we were +2 on defense (best in the league)

sfords team ERA was 2.38 playing in petco, but had an R/9 of 3.23. Taking into account the park factor (.82), you can estimate his neutral park R/9 to be 3.93 R/9.

My teams ERA was 3.80 playing in Hilltop, and had an R/9 of 4.48. Taking into account the park factor (1.14), you can estimate my neutral park R/9 to be 3.93 R/9.

Salary for sfords pitching $40.2m
Salary for my teams pitching $39.4m

Sford is one of the best owners on this site, and he knows how to build a good team. But after all was said and done, $ for $, his team that led the league with a 2.38 era had the same neutral R/9 as my team that had a 3.80 era.

This is great stuff thanks for sharing it!!
Could you explain how you derived the 'neutral park' R/9? I'm trying to understand it better

For neutral park R/9, I took the total runs allowed, and divided it by the park that was used (in examples above it was .82 for Petco, and 1.14 for hilltop). That gives you the “neutral park runs allowed”, and I then divided that by the innings pitched and multiplied by 9 to get the R/9 total.

That's using the Park Factor, which describes the run environment. 1.00 is neutral. Below 1.00 has lower run scoring environment, above 1.00 has a higher run scoring environment. The other effects (1B, 2B, 3B, HR) play into the overall environment, but are not themselves impacted by it. So you can easily compare run production across parks by dividing runs by the factor.

I have an almost identical team that just started in petco - I’ll update how this team finishes up

Posted by just4me on 1/25/2022 1:07:00 AM (view original):
That's using the Park Factor, which describes the run environment. 1.00 is neutral. Below 1.00 has lower run scoring environment, above 1.00 has a higher run scoring environment. The other effects (1B, 2B, 3B, HR) play into the overall environment, but are not themselves impacted by it. So you can easily compare run production across parks by dividing runs by the factor.

This ^^^

Posted by chargingryno on 1/25/2022 1:05:00 AM (view original):
For neutral park R/9, I took the total runs allowed, and divided it by the park that was used (in examples above it was .82 for Petco, and 1.14 for hilltop). That gives you the “neutral park runs allowed”, and I then divided that by the innings pitched and multiplied by 9 to get the R/9 total.

I guess I'm missing something. They only are playing half their games in their home park, why are you dividing their total runs scored by their home park factor?

Posted by akira_hokuto on 1/25/2022 11:01:00 AM (view original):

Posted by chargingryno on 1/25/2022 1:05:00 AM (view original):
For neutral park R/9, I took the total runs allowed, and divided it by the park that was used (in examples above it was .82 for Petco, and 1.14 for hilltop). That gives you the “neutral park runs allowed”, and I then divided that by the innings pitched and multiplied by 9 to get the R/9 total.

I guess I'm missing something. They only are playing half their games in their home park, why are you dividing their total runs scored by their home park factor?

That's a fair point.

Trying to calculate this now - looks like advanced stats don't show runs at home/away splits, just ER.