In the latest Collective Bargaining Agreement between Major League baseball, there were arrangements made to adjust each team’s schedules based on player feedback. Not only were start times adjusted, but additional off days were built into the schedule. These discussions got me curious about whether a home team’s performance was affected by the day of the week the game is played at home.
I’ve been having a lot of fun lately playing around with the Retrosheet* database I set up and figured this would be a good project for it to take a swing at. Just a warning to readers: not only is this article a bit number heavy, but the words “winning percentage” lack synonyms, and appear frequently.
With that in mind, I gathered a data set from 2010 to 2016, to have a sizable sample from before the current Collective Bargaining Agreement went into effect. Under each day of the week is the +/- winning percentage compared to that team’s overall home winning percentage. It has a gradient color code where the more green a cell is, the better a team did relative to their normal home winning percentage. Conversely, the more red a cell is, the worse they were on that specific day. The last column shows the variance in the values with smaller numbers meaning a team performed similarly regardless of day of the week. The more orange the number, the higher the variance. The Total row shows the average difference in winning percentage by day of the week and the average variance.
On average, teams league-wide seemed to win the most during their weekend home series of Friday through Sunday, performing poorer during midweek home games. Saturday was the best day, at an average of 0.012 winning percentage points among the 30 teams. The worst day of the week for a home team to play on was Thursday, costing a team 0.014 percentage points.
The team that seemed to benefit the most on a given day was the Chicago White Sox on Mondays, who normally had a .508 winning percentage at home but increased that by a whopping 0.184 on Mondays with an overall winning percentage of .692 on Mondays. Meanwhile, the team which did worst on a specific day at home was the San Diego Padres, who saw their overall home winning percentage of .517 tumble by .148 to a mere .368 winning percentage on Thursdays.
The stablest team at home, regardless of the day of the week, was the St. Louis Cardinals, who varied just .055 regardless of the day of the week, from a high of .021 on Friday to a deficit of .034 on Wednesdays. They also sported the overall highest home winning percentage of .608 of the thirty teams. The White Sox had the highest swing, going from that high of .692 on Mondays down to .450 on Thursdays.
I also decided to chop up the week between weekends, limited to Saturday and Sunday, and compare them to weekdays. The chart below takes those groupings and averages the winning percentage among two days for weekends and five days for weekdays.
With the data set limited to just Saturdays and Sundays, teams across the league tend to do .011 percentage points better on weekends and -.006 worse on the weekdays.
The best weekend warriors are the Los Angeles Dodgers, who increase their home winning percentage by .067 on Saturdays and Sundays but take a horrid nose dive during the week, punctuated by a -.133 on Wednesdays for an overall average of -.032 during the week. On the flipside, the Chicago Cubs do very well during the week, outperforming their 0.511 winning percentage by an average of .031 points during the week. However, they sport the worst Sundays at -0.089 in the league, dragging down their overall weekend numbers to worst in the league on weekends.
It appears that some general trends indicate that weekday games are harder for home teams than weekend games. However, how it affects each individual team varies wildly, with some teams impacted on certain days more than others. It will be interesting if, after the changes made to scheduling from the new Collective Bargaining Agreement, the issues will be mitigated for 2017 and beyond.
*The information used here was obtained free of charge from and is copyrighted by Retrosheet. Interested parties may contact Retrosheet at 20 Sunset Rd., Newark, DE 19711.