Need More Complete Games

Baby Ruth · 07-10-2024, 04:34 PM

In my 1901 season I'm expecting to see 99% (99+%?) of games pitched by a single pitcher, but the teams seem to make liberal use of relievers (at least for closing). Is there a setting that might be slightly off for this historical league that's causing it to not better represent what was happening in earlier baseball?

Garlon · 07-10-2024, 06:18 PM

The 1901 season had about 87% complete games and this is in the file the game uses for reference. However, that is 1913 CG but only 1094 team wins that season. There were more CG than wins. A couple considerations here is that you are then asking OOTP to value a CG more than trying to win by putting in a PH in the 8th inning. Another consideration is that what is far more important than CG are the distribution of innings between SP and RP. If you have the game go all out to target CG than you will end up with way more innings going to SP than relief pitchers.

The game weighs the IPouts per GS if a given season and the CG rate by default to calculate the SP stamina each season. It is possible to adjust those weights in the engine file and have the game only target CG rate but I advise against doing that.

Consider that a CG is all or nothing too. If you pitch 26 outs that is not a complete game. So you send up putting undo emphasis on a couple of outs at the end of the game for a CG stat when winning is more important.

Baby Ruth · 07-10-2024, 07:34 PM

But wasn't it different in 1901? Or did you mean to type 1901 when you included the 87% complete games number? Do we actually have any data for 1901 on this?

thehef · 07-10-2024, 07:38 PM

I checked a few 1901 historical tests that I had semi-recently done - for other reasons - and found that MLB CG% is in the 60%-70% range.

Garlon makes a good point in that we wouldn't want OOTP AI to put CG's above strategy to win, but I'm wondering what it was in real life - that is not reflected in OOTP - that resulted in more CG's while at the same time not sacrificing trying to win...

Also wondering if there's something - other than editing the engine file - that can be done to at least tweak those CG's upward a bit, even if they wouldn't approach the 87% from real life. For example, the "hook" setting (League Settings > Stats & AI, under General Strategic Tendencies) for 1901 MLB is +1, which is ever-so-slightly slower than the default or neither-fast-nor-slow/middle setting. So in a 1901 OOTP MLB where CG's are already at 65-70%, might a tweak to +3 or +4 bring that % up to maybe the high 70's or low 80's? If so, might there be some unintended consequences to this?

uruguru · 07-10-2024, 07:42 PM

I would just set the league-wide hook for SP as slow as possible. That combined with high pitcher staminas should give you a lot more CG.

Baby Ruth · 07-10-2024, 08:53 PM

I would think that the real world complete game % in the 20th century would have been more than 60-70%, but maybe my impression of the stamina of those workhorses like Cy Young is inflated. Maybe my concept is completely unrealistic and it's possibly just the idea of seeing Saves in that time period that's making me feel something is off. I'm not overly familiar with settings in OOTP, but I'm not opposed to making tweaks for reality purposes. I guess a good question would be to ask what the OOTP deadball veterans use as settings. Wasn't it normal even as late as the 1970s for pitchers to pitch a lot of complete games?

snepp · 07-10-2024, 09:07 PM

https://www.baseball-reference.com/l...ndard_pitching

2218 games started
1913 complete games

A shade over 86%

Baby Ruth · 07-10-2024, 09:24 PM

Quote:

Originally Posted by snepp

https://www.baseball-reference.com/l...ndard_pitching

2218 games started
1913 complete games

A shade over 86%

Interesting. So then maybe I'll just let the game run the way it is and see what happens.

Garlon · 07-10-2024, 09:34 PM

I meant 1901 in that earlier post. I mistyped it as 1991. We have the CG percentage in the era_stats file for every season and the game does use it.

Most likely what is different between real and OOTP is that teams were fine with losing while their pitcher threw a CG. There is no way around the fact that there far more CG than wins.

snepp · 07-10-2024, 09:37 PM

Quote:

Originally Posted by Baby Ruth

Interesting. So then maybe I'll just let the game run the way it is and see what happens.

Also of note, CG% makes a dramatic nosedive throughout the decade. By 1910 the percentage is just 62%.

thehef · 07-10-2024, 09:53 PM

Quote:

Originally Posted by Garlon

Most likely what is different between real and OOTP is that teams were fine with losing while their pitcher threw a CG. There is no way around the fact that there far more CG than wins.

Giving it more thought, I think it was a combination of things:

a) Mindset: You had 8 positions players and a few starting pitchers. All were expected to go the distance, generally speaking.

b) Small rosters: Reserve position players were simply backups. Many pitchers also played in the field, when needed, and at the plate there wasn't much difference between the two. IOW, many pitchers hit as well as many of the position guys on the bench.

c) Small pitching staffs: There weren't many - if any - relievers on a roster. And when there were, they generally weren't seen as better options than the guy who started the game, and who - again - probably hits as well as anyone on the bench who might pinch-hit for him.

I mean all of the above are intertwined... And it's probably not easy to tell OOTP AI to think that way... Although there are enough settings options that we can probably get close!

Garlon · 07-10-2024, 10:50 PM

Regarding pitchers hitting there actually was a substantial difference. In 1901, position players batted .279 and pitchers batted .200.

thehef · 07-11-2024, 02:31 AM

Quote:

Originally Posted by Garlon

Regarding pitchers hitting there actually was a substantial difference. In 1901, position players batted .279 and pitchers batted .200.

Overall, sure. But I'll bet if we compare the top starting pitchers - the ones getting the bulk of the complete games - with the position players on the bench, the difference in batting avg, etc., would be smaller. And that's really the comparison: Letting the starting pitcher hit, vs pinch-hitting for him.

thehef · 07-11-2024, 03:39 AM

Quote:

Originally Posted by thehef

Overall, sure. But I'll bet if we compare the top starting pitchers - the ones getting the bulk of the complete games - with the position players on the bench, the difference in batting avg, etc., would be smaller. And that's really the comparison: Letting the starting pitcher hit, vs pinch-hitting for him.

Was curious about this so I picked 8 of the 16 1901 teams at random, then added up the AB's & hits of their non-starter position players*, and then did the same for their top three pitchers**. Then totaled it all up and calculated batting averages:

Non-starter position players: .235
Top three starting pitchers: .220

*From retrosheet, these would generally be the guys listed below the 8 starters.
**From retrosheet, these would simply be the three pitchers with the most at-bats

07-10-2024, 04:34 PM	#1
Baby Ruth All Star Reserve Join Date: Mar 2021 Location: Reading Posts: 632 Infractions: 0/2 (3)	Need More Complete Games In my 1901 season I'm expecting to see 99% (99+%?) of games pitched by a single pitcher, but the teams seem to make liberal use of relievers (at least for closing). Is there a setting that might be slightly off for this historical league that's causing it to not better represent what was happening in earlier baseball?

07-10-2024, 06:18 PM	#2
Garlon Hall Of Famer Join Date: Jun 2004 Posts: 4,270	The 1901 season had about 87% complete games and this is in the file the game uses for reference. However, that is 1913 CG but only 1094 team wins that season. There were more CG than wins. A couple considerations here is that you are then asking OOTP to value a CG more than trying to win by putting in a PH in the 8th inning. Another consideration is that what is far more important than CG are the distribution of innings between SP and RP. If you have the game go all out to target CG than you will end up with way more innings going to SP than relief pitchers. The game weighs the IPouts per GS if a given season and the CG rate by default to calculate the SP stamina each season. It is possible to adjust those weights in the engine file and have the game only target CG rate but I advise against doing that. Consider that a CG is all or nothing too. If you pitch 26 outs that is not a complete game. So you send up putting undo emphasis on a couple of outs at the end of the game for a CG stat when winning is more important. Last edited by Garlon; 07-10-2024 at 09:30 PM.

07-10-2024, 07:34 PM	#3
Baby Ruth All Star Reserve Join Date: Mar 2021 Location: Reading Posts: 632 Infractions: 0/2 (3)	But wasn't it different in 1901? Or did you mean to type 1901 when you included the 87% complete games number? Do we actually have any data for 1901 on this?

07-10-2024, 07:38 PM	#4
thehef Hall Of Famer Join Date: Jun 2006 Posts: 5,345	I checked a few 1901 historical tests that I had semi-recently done - for other reasons - and found that MLB CG% is in the 60%-70% range. Garlon makes a good point in that we wouldn't want OOTP AI to put CG's above strategy to win, but I'm wondering what it was in real life - that is not reflected in OOTP - that resulted in more CG's while at the same time not sacrificing trying to win... Also wondering if there's something - other than editing the engine file - that can be done to at least tweak those CG's upward a bit, even if they wouldn't approach the 87% from real life. For example, the "hook" setting (League Settings > Stats & AI, under General Strategic Tendencies) for 1901 MLB is +1, which is ever-so-slightly slower than the default or neither-fast-nor-slow/middle setting. So in a 1901 OOTP MLB where CG's are already at 65-70%, might a tweak to +3 or +4 bring that % up to maybe the high 70's or low 80's? If so, might there be some unintended consequences to this?

07-10-2024, 07:42 PM	#5
uruguru All Star Starter Join Date: May 2022 Posts: 1,297	I would just set the league-wide hook for SP as slow as possible. That combined with high pitcher staminas should give you a lot more CG.

07-10-2024, 08:53 PM	#6
Baby Ruth All Star Reserve Join Date: Mar 2021 Location: Reading Posts: 632 Infractions: 0/2 (3)	I would think that the real world complete game % in the 20th century would have been more than 60-70%, but maybe my impression of the stamina of those workhorses like Cy Young is inflated. Maybe my concept is completely unrealistic and it's possibly just the idea of seeing Saves in that time period that's making me feel something is off. I'm not overly familiar with settings in OOTP, but I'm not opposed to making tweaks for reality purposes. I guess a good question would be to ask what the OOTP deadball veterans use as settings. Wasn't it normal even as late as the 1970s for pitchers to pitch a lot of complete games?

07-10-2024, 09:07 PM	#7
snepp All Star Starter Join Date: Jan 2008 Posts: 1,500	https://www.baseball-reference.com/l...ndard_pitching 2218 games started 1913 complete games A shade over 86%

07-10-2024, 09:34 PM	#9
Garlon Hall Of Famer Join Date: Jun 2004 Posts: 4,270	I meant 1901 in that earlier post. I mistyped it as 1991. We have the CG percentage in the era_stats file for every season and the game does use it. Most likely what is different between real and OOTP is that teams were fine with losing while their pitcher threw a CG. There is no way around the fact that there far more CG than wins.

07-10-2024, 10:50 PM	#12
Garlon Hall Of Famer Join Date: Jun 2004 Posts: 4,270	Regarding pitchers hitting there actually was a substantial difference. In 1901, position players batted .279 and pitchers batted .200.