Unbalanced Schedule Working... sort of"Conference" play and a modest proposal to revise Baseball's standingsBy The Crank It's Memorial Day weekend, the traditional start of summer. Tradition has it a team in first place on the fourth of July will go on to win, and reality has often reflected this rule of thumb. With holes in the standings dug or not dug, a team can still turn it around in the five weeks between our two great summer holidays. As I write this, 11 of 16 NL teams are at or above .500 and two more are just a game under .500. Two of those teams under .500 are in the hapless NL West, and still have reasonable shots at winning that division (relatively). Only the homeless Expos are buried more than ten games out just shy of the one-third mark of the season. In the AL, it's a bit more divided into haves and have-nots, but not too much so. Tampa Bay, Kansas City, and a little surprisingly, Seattle have dug early graves for themselves (that's not a crack about how old the Mariners are). Things look a little grim for Baltimore and Toronto chasing the Yanks and Red Sox on general principle, but they're not totally out of it. We have pennant races shaping up in all six divisions: Sox-Yanks in the AL East, ChiSox- Twinkies in the AL Central, Texas-Oakland-Anaheim in the West; Phish-Phillies in the NL East (with miracle workers in Atlanta not to be counted out just yet), the entire NL Central seemingly competitive this year, and realistically, any of the teams in the NL West given the number of games they have to play against one another and the general weaknesses of the teams there. There has been some complaining about the "unfairness" of the unbalanced schedule in determining the wild card, and no doubt that will continue to be true. One league has two fewer teams than the other, which inherently gives any given AL team a better chance to make the playoffs as a division champion than any given NL team. That's the part that's really unfair, most especially to the fans of NL Central teams, the division being crowded with six teams. Compared to the AL West, where a team starts the year with a one in four chance (all things being equal) of winning, an NL Central team starts the year with a 17% chance of being a division winner - a third less chance from a mathematical perspective. What "makes up" for this in a somewhat randomized way is the unbalanced schedule and the wild card, but not the way you might think. The wild card, of course, may be as good as the second-best team in the league, and by definition will be the second-best team in a division. With teams playing twice as much against specific division opponents as against non-division league opponents, that gives an advantage in the wildcard race to clubs that are in divisions with only two good clubs, and more to teams with only two good clubs in a bigger division. I don't think we've had enough years under these rules and configurations to make the case statistically, but think about it. Last year's wild cards were the Red Sox and Marlins, both terrific clubs, and they came from divisions with only two solid clubs. ( Don't tell me about how the Phillies were in it in September: they exceeded their Pythagorean expectations as it was.) If a division has three good clubs, all three teams will have fewer chances to get wins against weak opponents and thus hurt one another's chances at the wild card. Adding to this confusion is interleague play, which is completely unbalanced and nearly random in terms of relative strength of schedule. I don't know if the scheduling problem can be fixed without going to 32 teams and 4-team divisions: there's no mathematical way of "balancing" a schedule with an unbalanced intradivisional component otherwise. And as much as frequent readers of this column will be surprised to hear it, I think the Selig regime has fairly legitimate marketing reasons for maintaining the imbalanced interleague schedule. If we start to analyze the standings at this point, with the schedule having oddities like some season series being over already and others not even having started, it's even weirder -- except within the divisions, where every team has started, but not finished, its season series with any other club in the same division. Let's take a look at how some of the teams are doing and where they've won their games, and what that might tell us about where these next five weeks might go. In the AL East, the Sox and Yanks have been Cat and Mouse, Hatfield and McCoy, Seabiscuit and War Admiral close for much of the season. The Yanks, starting out 1-6 against the Sox, are only 8-8 against East opponents, while the Red Sox are 22-9. The Yanks have made up the deficit by going 14-7 against the West. The Red Sox have played only eight games against the West, going 3-5, including being swept by the New Kids from Texas. With this and one or two other exceptions, the current divisional standings closely mirror the records of clubs against divisional opponents. In other words, if this were, say, NCAA Conference play, and the "Conference Championship" were determined just by in- conference play, the standings would be very similar to what they are now, just a third of the way into the season. This basically suggests that the "unbalanced" schedule, the purpose of which was to make divisional play a little more intense and important, has worked -- but only to a point. Because the divisions aren't composed of equal numbers of teams, teams in different divisions play divisional opponents an unequal number of times compared to teams in other divisions. Teams do typically play about 19 games against divisional opponents, very close to the 22 each team played against league rivals in the old 154-game schedule of eight-team leagues. Tendencies do tend to work out, but as we'll see in the example of the 2003 standings, not evenly. I have a modest proposal for revising the Wild Card: instead of making it the team with the "second best" record, make it the team which, not having qualified for the playoffs, had the best "non-conference" overall record. This would adjust, in a fashion, for the unbalance of the divisions. It wouldn't exactly be level, since the "non-conference" schedule would still be wildly different from team to team, but at least it wouldn't give teams in the types of divisions that produce wildcards -- big divisions with only two good teams, according to my theory -- an undue advantage on top of that unbalanced "non- conference" schedule. We could go a bit further and disregard an "overall" record for divisions, and only count "conference" play for divisional winnners. By separating them, this would truly create two races for each team -- one for the division, one for the wild card. Looking at the 2003 standings, would that have changed anything? Here are the 2003 "conference" and "non-conference" standings Of interest, the Sox and Yanks had identical "non-conference" standings, and with the season split between the two, the differenece between first and second was purely how well they did against the Rays, Jays, and O's. In the AL Central, we see the most dramatic change in results. The Royals, who finished third with a late fade -- some games in September against "non-conference" opponents down the stretch -- would've won their "conference", Minnesota and Chicago each slipping a notch. Not to say it wouldn't have been a trhilling ending: with only a three-game spread, the last series of the season would've been a pennant winner. The Red Sox win the wild card by percentage, but barely over the Mariners. If one posited a truly balanced divisional line-up, the Mariners might've won here, without having to play Oakland and Anaheim so many times in a small division. If you find this a confusing statement, just look at the larger number of "non-conference" games played by AL West teams compared to NL Central teams - more than 30! In the NL East, the team that many argued really played the best all year, the Marlins, come out as division winners instead of runners-up. In the NL Central, Houston edges the Cubs by two games to be the division champions. Milwaukee and Cincy swap spots at the bottom of the division. In the West, where the Giants ran away with it, LA and Colorado end up tied for second and Arizona drops to fourth. Atlanta is the Wild Card winner with a 60-26 "non-conference" record. Given how well the Marlins did in the post-season relative to Atlanta's struggles, this seems an honest reflection of which was the better team in the division -- as does the record against common opponents in the balanced intradivisional schedule. If we look beyond this, an interesting phenomenon comes out. Fewer teams are competitive for the Wild Card, more teams are competitive for divisional titles. In the AL, "second place" for the wild card would be the Minnesota Twins, finishing four games behind Seattle, and the Red Sox finish a distant nine games out of it. In the NL, second-place for the wild card goes to LA, ten games back, and third place is Philadelphia, 13 games back. In the AL East, the Red Sox lose by a convincing six games, the same margin of their actual defeat. The AL Central would've been a three-way race down to the wire, as it was. And the AL West would've been a one-game photo finish, with the A's prevailing. In the NL, the East wouldn't've been close - Florida the runaway victors. Houston, the Cubs, and Cards would've been fighting it out the last weekend. The west would still have been a walk for first, but the four also-rans would've been fighting it out for position. With the unbalanced schedule having teams play divisional opponents nearly exclusively in September -- the Astros, for example, play 23 of 26 September games against divisional opponents -- these hypothetical final standings mask what would be the reality of fierce divisional contests. Under the "conference" and "non-conference" plan, nearly every team would've had a chance to win their "conference" with good September play. If the point of the playoff system and schedule is to keep more fans into their teams late -- having the possibility of winning a division crown into September would surely do the trick. Staying in touch with the wild card as is gives a guarantee of at least one good "race" at the end of the year. This system would probably produce four or five races each year, and in theory all six divisions would be in play. This might keep some teams that have had bad luck or bad starts interested in competing a little bit longer. This is a way of keeping the current playoff format and the current unbalanced schedule -- which MLB has its reasons for wanting to keep -- but would keep more teams in a genuine race. And I suspect it may produce results of playoff teams, conference victors, and certainly wildcard representatives that are more representative of deserving teams. "Finish" - Team - "Conference Record" - "Non-Conference Record"
Playoffs (based on overall record) note: Mariners might've been wild card if "intradivisional" schedule was truly balanced vis a vis the Central and East.
Marlins-Astros
Leave feedback on our message board. |