Quote:
Originally Posted by kq76
The #-# code is just the result of a formula comparing what a team is doing over 3 rows/series with the middle row/series being the considered for being moved or not.
|
I gave an example as a case of a 4-series set, but really it is the same 3-series logic. The AHHA is the AHH part where I would want the offday between AH or the HHA part where I would want the offday between HA.
The positions of the series in the weeks matter. Like you describe, there is an early week series that can be shifted and a late week series that is locked in to end on Sunday. So I do my check on early week series sandwiched by the previous and next weekend series I think pretty much same as you describe...
- HH_H - keep or make H_HH, does not matter
- HH_A - keep
- AH_H - make A_HH
- AH_A - keep or make A_HA, does not matter
The underscore represents the offday, either Monday if earlier or Thursday if later. For the "does not matter" cases I will try to shift more or not to try roughly to balance the offdays earlier versus later in the week. I only apply the logic for the home team for the early week series because my key criterion is minimizing offdays within homestands.
Something I resist doing but that can pop up is a situation where my logic says to make a wraparound series that stretches into Monday. MLB has that sometimes, but I have not investigated how often those happen. The classic "wraparound" definition is 4-game Fri-Mon...
Quote:
Originally Posted by kq76
I've found it pretty much necessary to schedule 27 weeks instead of 26 because of all the offdays and I figure it's probably a lot easier to move around series within 27 weeks than it would be 26.
|
Sounds like you are talking about doing 162 games in 54 half-week 3-game series, thus 27 weeks (not counting an All-Star Break). Some years back the labor agreement shifted MLB to having 52 half-weeks - it was 51 when I first started making schedules. So those now 6 "extra" games from the "extra" week are most conveniently spread into making 4-game series.
For 12 teams in two 6-team divisions, an obvious breakdown of 162 games is 18 games times five intradivision opponents and 12 games times six interdivision opponents. Instead of making all 3-game series though, a week can be trimmed by sacrificing some of the perfect mathematical balance to have 8H & 10A against one division opponent and 10H & 8A against another. Instead of six 3-game series, for those two opponents it is five series, with three of them being 4-game.
The 4-game series can block movement. One itself will not be moved. Except I feel like I should be more open to allowing Fri-Mon, which requires moving or having moved the early week series in that next week. In the usual early week cases there will be instances my logic says to move the early week series to Tue-Thu because of the home team layout, but it is blocked because of the away team already having a Thu-Sun series.
I am not a fan of 2-game series in building. Early on I used to do 2/2/3 weeks (Mon-Tue, Wed-Thu, Fri-Sun) more often. Smaller series can help counts come out more ideally. That 18 games I split 10/8 or 8/10 above could be kept 9/9 by having the 9 games be like 2/3/4. But pairs of 2-game series I would only put early week which limits flexibility. I already have more single series homestands that I wish, and those pairs can lead to cases of 2-game trips with no offday on either side.
I am more open to the occasional standalone 2-game series early week. The ideal may be 46/6 split of 3-game series and 4-game series. But maybe like 2023 MLB you need to have one opponent 2H & 2A. So to make those two short series you can make two more 4-game series - 2 * 2 + 42 * 3 + 8 * 4 = 162. I prefer to have 2-game series followed by a weekend 4-game series, in which case the logic for positioning the 2-game series is same as above. If a team then has only a 3-game weekend series I force it to be Tue-Wed & Fri-Sun to keep from having consecutive offdays.
The non-3-game series allow can be distributed unevenly to allow for teams having played different number of games at a given time, and more half-games in the GB standings column. I like that variety and not having everybody having played the same X games after every Sunday.
Quote:
Originally Posted by kq76
Yeah, I was thinking I might do that one day. I stuck with all Sunday games at 1:05 and all others at 7:05 as it didn't seem worth it at the time to alter them further. I did look at a couple of real schedules to see if I could spot a pattern for why some non-Sunday games started early and I remember it actually seemed somewhat random. I have to imagine I looked at whether the home team was leaving or not, but I'm sure I didn't consider whether the away team was playing the next day or not. Is that the only factor you've found?
|
Gametimes are a relatively new thing for me to tinker with - I had been happy making everything 7pm. I have not done real research (common theme), mostly just going off vibes. I know people like to see have weekend day games. And the getaway day thing feels right.
I made logic to allow random setting of gametimes. E.g., make it so every Sunday has 60% chance of 1pm, 25% of 4pm, 15% of 7pm. It is from random number generation case-by-case, so the distribution will not be perfectly those percentages.
For getaway days, if I turn on changing the gametime it can be "either" or "both" mode. Of course since the away team has to travel somewhere else, "either" mode would mean always shifting unless the away team is off the next day plus the home team is off or still at home. That seems too frequent, so I doubt I would use it much.
I am pretty sure MLB has some rules/guidelines for gametimes & offdays for teams traveling from west coast to east coast, like to keep from playing in Oakland one night then Boston the next night. Doing that in a build would add more difficulty, so nice just to assume that never know where teams are.