20182019 White Sox Experiment, Week 4: How Good Should This Team Be?

Expect a lot more celebrations, White Sox fans.

Current status: After 28 games, the 20182019 team is sitting in 3rd in the AL Central, with a 13-15 record. When I started this, I envisioned that the results would be better than either of the previous two seasons, but that isn't what has happened. Instead, the 28-game record of this results-optimized team is exactly the same as the 2019 team: 13-15.

There are a couple reasons I suspect are driving this:

• early-season weather affecting the schedule, resulting in a lot of days with only one game, and 
• the fact that we've only gotten the bad versions of both Reynaldo Lopez and Lucas Giolito. 

So, there's probably some upside going forward based on the fact that those two will be pitching better (probably). After all, it's the starting pitching that is making this team really struggle.
But what should we expect? Even though 2019 was better than 2018, both teams were bad. Still, based on each team's record and some assumptions on how games are scheduled, we can make an estimate.

How Good Should This Team Be?

In order to estimate where things should be, we need to look at what actually happened. The 2018 team was 62-100 (a .383 winning %); the 2019 record was 72-90 (.444). If we assume that the season is ~26 weeks long (182 days), and each team has to fit 162 games into those days, then each team will have a game on 89% of days.

Days with both teams having games: 144 (182 * .89 * .89)
Days with neither team having a game: 2 (182 * (1-.89) * (1-.89))
Days with one team having a game: 36

If we assume that the one-team days are randomly distributed, then on those 36 days the 20182019 White Sox should be expected to win at a .413 pace (the average between 2018 and 2019). This would result in 14.7 wins.

Obviously, the days with neither team playing will result in 0 wins.

Which leaves the days when both teams play. To figure out the days on which both teams win, we can just take the combination of their winning percentages (144 * .383 * .444). We would expect 22 dual-win days. Likewise, we can easily figure out how many dual-loss days to plan for: 44.

What we have left is 78 days where one of the two teams wins. Taken in total, here's how it breaks out:

Result	W	L	Pct
2 games, both wins	22	0	1.000
2 games, both losses	0	44	0.000
1 game	15	21	0.417
2 games, 1 team wins	78	0	1.000
Total	115	65	0.639
Scaled Total	104	58	0.642

That's right. According to math, we should expect this team to win 104 games. Given that we've started 13-15, that means we should expect a 91-43 stretch coming up. That'll play.