11/22/11

Guest Post: - REALIGNMENT & BALANCE - by Hobie

REALIGNMENT & BALANCE
By Hobie
For some reason I’m a scheduling buff.  It’s an avocation, not a vocation.  My wife is a HS Assistant Principal and we spend weeks during the summer scheduling her 500+ students in a grid organized to avoid conflicts in student preferences, optimize teacher loading and accommodate room availability.  Some decades ago I was the local Little League Commissioner and managed to work out a balanced schedule which included inter-league games with three neighboring leagues with varying numbers of teams. 
Balanced is the operative word here.  If W-L records are to determine reward, then those in the same divisional competition should face ideally the exact same schedule in terms of opponents, Home & Away, etc.  That’s not always possible, but it should be the goal anyway.
In the 8-team leagues of my youth, each club was host for 11 & guest for another 11 games against their seven competitors.  The 7x22 yielded the magic 154 game season, not the other way around.  The initial expansion to 10-team leagues gave us the now standard 162 while maintaining a still perfect balance of a 9x18 format.  Subsequent expansion and interleague play has grossly, IMO, upset the symmetry of this most symmetric game.
The recent re-alignment into equally populated divisions, however, gives an opportunity to restore a semblance of balance.  It does require an interleague contest (or some odd-number of IL contests) on any given gameday which is just some mathematical fact we’ll have to live with.  And assuming the 162 game envelope is sacrosanct perfection will allude us, but I have an idea.
My preference would not have been to kick the Astros into the AL, but to optimize geographically and still hold out the prospective of a “local” WS (subway in NY, Loop in Chicago, BART in the Bay Area, Beltway in DC, etc).  This is a personal fit of nostalgia sparked by an entire pre-adolescence consumed by local October baseball in NY  Maybe an Astros-Rangers WS doesn’t have the same appeal to Texans.  Anyway, my division organization  (and note the vertical cross-league pairings) would be:
AL EAST: Yankees—BoSox—Pirates—Orioles—Rays
NL EAST: Mets—Braves—Phillies—Nationals—Marlins
AL CENTRAL: ChiSox—Twins—Indians—Blue Jays—Royals
NL CENTRAL: Cubs—Brewers—Reds—Tigers—Cards
AL WEST: Angels—Athletics—D’Backs—Mariners—Rangers
NL WEST: Dodgers—Giants—Padres—Rockies—Astros
The subject at hand is schedule however, and although I have a geographic weighting component, it is optimizing divisional balance that I am aiming for.  This scheme then is independent of specific divisional membership. 
I think of a season spanning 180 days in sixty 3-day blocks of DATES.  A block could contain a 3-game set or a 2-game set with an off-day.  For a 162 game schedule, 18 of the blocks would house 2-game sets for each team (& 18 off-days) with the remaining 42 filled with 3 game sets. 
Divisional Play:  19 games vs. the four intradivisional opponents yields 76 games.  They would be played in five 3-game sets and two 2-games sets.  In year 1, the Mets would get three 3-game home sets (9G) and a pair of 3-game and another pair of 2-game sets (10G) in say, Philadelphia (HFA Phillies, 10-9).  The next year that would flip.  The Mets would have HFA vs. Braves & Marlins in year 1, Phillies HFA over Mets & Nats, Nats HFA over Mets & Marlins, Marlins over Braves & Phillies and those would flip in year 2.  We have something like that now, except the presently the number of intradivisional games are not even equal.  So far 20 3G sets & 8 2G sets per team.

“Rival” Interleague Play:  Each NLE team would play each of the ALE teams two 2G sets, 1 home, 1 away. Likewise for NLC/ALC and NLW/ALW. That’s 20 inter-league games (20 more common opponent games for division members) filling the remaining ten 2G sets.

Common Interdivision Play: In alternating years the NLE would visit all five NLC teams while hosting all five NLW teams for 3G sets.  That’s another ten 3G sets and another 30 common opponent games.  So far, 30 3G sets & 18 2G sets (126 common opponent games). 

The remaining 36 games (12 3G sets) would be spread over 3 opponents in each of the “other” NL & AL divisions in “supplemental” interdivision and interleague play  Commonality becomes a victim as none of the NLE teams would face the same three NLC teams (at home since they would visit all the NLC teams in year 1).  But each team the Mets face would also play two other NLE rivals, and in a cycle of 5 years, every NLE team will have faced every NLC (& NLW) team four times—two home, two away.  Likewise for NLW opponents.
Similarly the NLE would visit three ALW cities and host three ALC teams in a rotation that would insure meeting every non-East AL team three times in a five year cycle.  In ten years, the Mets will have faced Seattle 6 times, 3 home and 3 away.
Summarizing, that would be:
·  76 games against common division rivals (19x4)
·  48 games against other league teams (24 each against “other” in-league divisions)—30 games in common home & away.
·  20 games against (common) rival division in other league
·  18 games against other league non-rival divisions.


Upside is my beloved symmetry: 126 common opponents for all division members and another 12 in common for any two division members.  And with 16 sets (10 2G and 6 3G) of interleague play for each of a league’s 15 teams, there’s more than enough to insure at least one interleague contest on any playing date.
The downside is it’s not perfectly balanced and there are too many interleague games for my taste—a price I don’t see how to avoid with odd-numbered leagues.
OK, Bud.  The ball’s in your court.


1 comment:

Mack Ade said...

Hobie:

This makes too much sense.

I have a funny feeling that the team chosen for realignment was the one that bitched the least when approached with the idea.