This has been a long time coming. I started running my fatigue test teams and documenting them with this thread back in 2013. This last year I also posted fatigue info from a Mike Marshall theme league and a TWISL theme focused on pitcher fatigue.

The following data set includes everything in this sheet: Which includes 19,135 games, 277 player seasons, and 57,532 IP across 60+ teams.

The following table is the list of all of the pitchers contained in this data:

Pitchers		Times Used		Pitchers		Times Used		Pitchers		Times Used
1997 Greg Maddux		33		1910 Jack Pfeister		1		1992 Greg Maddux		1
1996 Kevin Brown		29		1914 Rankin Johnson		1		1994 Brett Saberhagen		1
2003 Jason Schmidt		26		1915 Babe Ruth		1		1995 Greg Maddux		1
2002 Derek Lowe		25		1915 Jim Scott		1		2000 Pedro Martinez		1
2011 Roy Halladay		25		1919 Slim Sallee		1		2002 Ramon Ortiz		1
1974 Mike Marshall		24		1920 George Mogridge		1		2005 Johan Santana		1
1996 Greg Maddux		23		1920 Slim Sallee		1		2008 Tim Wakefield		1
2010 Cliff Lee		23		1922 Babe Adams		1		2009 Tim Lincecum		1
1902 Bill Bernhard		4		1923 Bill Harris		1		2011 Clayton Kershaw		1
1909 Addie Joss		4		1930 Les Sweetland		1		2011 Cliff Lee		1
1888 Silver King		2		1933 Hi Bell		1		2011 Doug Fister		1
1972 Steve Carlton		2		1946 Tommy Hughes		1		2013 AJ Griffin		1
2002 Pedro Martinez		2		1956 Herb Score		1		2015 Gerrit Cole		1
2019 Jacob deGrom		2		1963 Larry Bearnarth		1		2016 Clayton Kershaw		1
1887 Ed Seward		1		1971 John Cumberland		1		2018 Jacob deGrom		1
1890 Sadie McMahon		1		1972 Nolan Ryan		1		2018 Max Sherzer		1
1892 Alex Jones		1		1972 Wilbur Wood		1		2019 Charlie Morton		1
1899 Frank Bates		1		1976 Woodie Fryman		1		2019 Gerrit Cole		1
1901 Roscoe Miller		1		1980 Jerry Reuss		1		2019 Mike Clevinger		1
1903 Bill Bernhard		1		1984 Dwight Gooden		1		2019 Mike Soroka		1
1906 Addie Joss		1		1985 Orel Hershiser		1		2019 Tyler Glasnow		1
1907 Jake Weimar		1		1989 Nolan Ryan		1		2019 Zack Greinke		1
1908 Irv Young		1

As can be seen from the above, I didn't exclusively use any particular type or quality of pitcher, though I did focus the tests on certain pitcher seasons as they fully encapsulated what I was looking for. There is a wide range of quality in the above, but most importantly, these pitcher seasons are captured throughout the fatigue data points, so as bad as you can imagine 1899 Frank Bates at 100%, imagine him at 75-80%. This full set not only includes a mix of pitchers and pitcher types, but also includes a mix of ballparks and defenses.

This full data set also includes the Mike Marshall and TWISL fatigue test data. As such, the 99-100% category does include one pitcher who did not also pitch with the intention of being fatigued (1974 Mike Marshall - 20 seasons). Those 20 seasons are the only ones that also did not pitch fatigued and the details specific to them can be seen in the link above.

While I did use Wrigley, AFC, Coors, Baker Bowl, etc... as part of the test, I was trying to prove out controlling for variables within the decision tree (OAV & HRs primarily), so most of the tests were performed in Petco with A++ range defense, though a few were also centered around FLD% instead of range or not focused on defense at all. Similarly, I primarily focused on pitchers with low BB/9 and HR/9 to mitigate the effects of fatigue and good OAV such that the park and defense would essentially cancel out the effects of fatigue, if not even still allow them to perform better than RL despite fatigue.

I'll mostly let the data speak for itself. My thoughts from the TWISL thread linked above are also borne out here, so I won't rehash my full thoughts on the data or what it means, but will just state I think this demonstrates quite well that fatigue can be mitigated to make it almost a non-factor down to the 75-80% range. Most of all, Enjoy!

[Edited out a note about 1994 Brett Saberhagen from the TWISL as I did not include his stats in this sheet.]

Games	Avg PC	AVG %	IP	BFP	PC	W	L	SV	R	ER	H	HR	SO	BB	WP	ERA	OAV	WHIP	HR/9	BB/9
690	81.04	45	2896.36	16602	55921	36	354	24	4962	4495	5966	986	801	1961	43	13.97	0.407	2.74	3.06	6.09	0-59%
1380	86.82	66	7270.69	36115	119817	226	693	146	8039	7205	11364	1609	2497	3228	79	8.92	0.343	2.01	1.99	4.00	60-69%
1983	60.08	72	8486.58	36578	119145	425	512	58	4867	4220	9465	1012	3837	1915	57	4.48	0.271	1.34	1.07	2.03	70-74%
1470	60.57	77	6350.59	27059	89045	266	379	35	3141	2759	6768	561	3058	1337	48	3.91	0.262	1.28	0.80	1.89	75-79%
2120	57.33	82	8649.19	36804	121543	409	486	44	4166	3562	8784	654	4034	2084	86	3.71	0.253	1.26	0.68	2.17	80-84%
2266	53.89	87	8976.11	36936	122121	410	482	74	3504	2970	8176	457	4679	1960	70	2.98	0.233	1.13	0.46	1.97	85-90%
2361	56.29	92	9462.86	39837	132899	515	462	68	4040	3438	8880	507	5382	2399	177	3.27	0.238	1.19	0.48	2.28	90-94%
1479	51.63	96	5439.57	22718	76361	384	238	72	2168	1869	4938	297	3433	1364	49	3.09	0.232	1.16	0.49	2.26	95-98%
5566	70.88	100	28071.09	118108	394494	1947	1160	320	11925	10323	27098	1579	17902	6390	250	3.31	0.243	1.19	0.51	2.05	99-100%
19135	64.28	86	57531.95	252649	836852	2671	3606	521	34887	30518	64341	6083	27721	16248	609	4.77	0.272	1.40	0.95	2.54	Total

Just so it's clear, the fatigue range in the rightmost column is the pitchers fatigue % at the start of each game recorded. The fatigue % in the 3rd column from the left is the average fatigue % within that subgroup for at the start of each game recorded. All games can be seen in the google sheet link in the first post.

If I’m seeing this data correctly, the biggest change between 70-75% and 100% is HR9 and OAV, but the OAV dropping .30-.40 points isn’t as dramatic as I’d have thought.

BB/9 numbers are virtually unchanged through fatigue based on this data

It also appears that 85-90% is the money spot lol

Very interesting. In my own subjective recollections, I see competing teams use fatigue pitchers successfully at times, but when that occurs with mine, the wheels come off much more often than not. I suppose someone would counter that this is because my pitchers are near the edge of effectiveness in the first place.

QUESTION, J4M:
In old communications with WIS staff, they speak of fatigue kicking off as if it's decided by a "dice roll" when the fatigued-by-percentage pitcher first takes the mound. Have you see this kind of hard Fatigue:ON vs. Fatigue:OFF effect, as you can tell?

I think that may have been the case prior to the "fatigue strategy" and removing the random factor of fatigue to giving a more consistent downgrading of performance to help mitigate people utilizing that strategy. From what I can tell, it appears that fatigue is always on, and a very strict linear adjustment to the decision tree factors (OAV, BB/9, HR/9). So, in short, you start with good, low numbers in those categories (say .240, 1.00, 0.50) to start. Run them down to 80% and those all get multiplied by 1.2, so that line becomes .288, 1.20, 0.60. The OAV and HR rate can both be manipulated with park/defense, and the BB rate is low enough to start that the % increase doesn't really impact the raw totals. Petco alone brings that OAV to .261, range on defense can then take away enough hits to completely mitigate the effect of fatigue there. Similarly, Petco can take that HR/9 back to a .46, ant your pitcher at 80% fatigue is looking at a .238/1.20/0.46 OAV, BB/9, HR/9 slash, which is essentially where they started for a neutral park.

That's the presumption that the vast majority of the test data above was run on, and the performances seem to bear out. Especially with the number of individual performances that set new PH bests ('96 Brown, '96 Maddux, '11 Halladay (twice), among others). Though my favorite might just be the '72 Carlton who put up the following slashes in his two uses:
Real Life: 360 IP, 1.97 ERA, 0.99 WHIP, .206 OAV, 0.44 HR/9
# stats: 396 IP, 2.10 ERA, 1.03 WHIP, .214 OAV, 0.42 HR/9
Attempt 1: 471.33 IP, 3.13 ERA, 1.14 WHIP, .224 OAV, 0.65 HR/9 (OL)
Attempt 2: 459.67 IP, 3.15 ERA, 1.24 WHIP, .238 OAV, 0.57 HR/9 (CL)

If you control the variables you can, you can essentially buy down their $/IP by keeping their performance roughly static while increasing their IP.

I think the reason why anecdotally, it can feel like “fatigued pitchers always do worse for me”, is because in the majority of those situations, those teams weren’t drafted with fatigue in mind. You put a 80% pitcher on the mound with a bunch of B-/C+ fielders behind him, and then add onto it a semi hitters friendly park (or even a pitchers park for that matter), the results likely won’t be very good.

however, like just4me says above, if you draft your team with the goal of mitigating specific things like HR9/oav, and draft defense and park accordingly, then you can manipulate it a little bit to your favor.

Super interesting! Thanks for sharing

Posted by chargingryno on 3/14/2021 12:43:00 PM (view original):
I think the reason why anecdotally, it can feel like “fatigued pitchers always do worse for me”, is because in the majority of those situations, those teams weren’t drafted with fatigue in mind. You put a 80% pitcher on the mound with a bunch of B-/C+ fielders behind him, and then add onto it a semi hitters friendly park (or even a pitchers park for that matter), the results likely won’t be very good.

however, like just4me says above, if you draft your team with the goal of mitigating specific things like HR9/oav, and draft defense and park accordingly, then you can manipulate it a little bit to your favor.

Also - there's a big difference between "I drafted this team on purpose to pitch the whole year at 80%" and "I ended up with everyone in the 80s but we're surviving" and even "We started with everyone in the 80s and then went into a death spiral"

It's tough to draw the right line where your staff is consistently fatigued but not at risk of collapse.

Posted by jfranco77 on 3/16/2021 11:36:00 AM (view original):

Posted by chargingryno on 3/14/2021 12:43:00 PM (view original):
I think the reason why anecdotally, it can feel like “fatigued pitchers always do worse for me”, is because in the majority of those situations, those teams weren’t drafted with fatigue in mind. You put a 80% pitcher on the mound with a bunch of B-/C+ fielders behind him, and then add onto it a semi hitters friendly park (or even a pitchers park for that matter), the results likely won’t be very good.

however, like just4me says above, if you draft your team with the goal of mitigating specific things like HR9/oav, and draft defense and park accordingly, then you can manipulate it a little bit to your favor.

Also - there's a big difference between "I drafted this team on purpose to pitch the whole year at 80%" and "I ended up with everyone in the 80s but we're surviving" and even "We started with everyone in the 80s and then went into a death spiral"

It's tough to draw the right line where your staff is consistently fatigued but not at risk of collapse.

Correct. I have two OL teams; one I spent $28m on pitching (1250ip) and drafted 6 Jimmie Foxx seasons and stuck them in Coors. They’re pitching at 70-90% and giving up 8-15 runs a game and are 20-32. The other I spent $37m on pitching (1230ip) and intentionally drafted so my starters would pitch at 70-80% majority of the season, drafted A+ range fielders and stuck them in Hilltop. They’re giving up 3-4 runs a game and are 83-41.

Thank you for sharing, J4M!

Thank you, Just. And charging. I get pitcher "issues" with 99% and very good defense. WIS has a cheat-code working against me.

BTW, 4 spots left in the current Champs League join. Then there's the one after that....

Bump

Bump for savoybg