Tuesday, May 26, 2009

Tweaks to make the NHL even better

The evolution of the NHL's rules since the lockout has undoubtedly improved game play in what is already the best professional sports league on the planet. Below, I will describe a few tweaks that could make the game even better. I'm going to skip the draft order, since I've covered that in the past.

Starting with scheduling, to keep with this string of posts, a mild realignment putting Vancouver in the Pacific and moving Dallas to Northwest (renamed to Mountain or something suitable) would make it more likely that an extra Canadian team would make the playoffs. The NHL is obviously a North American League, but there's no doubt that stronger Canadian franchises are good for the league as a whole. Consolidating the six Canadian teams into two divisions just serves to have them beat each other up all year; even as late as '92-'93, only the Patrick Division lacked a Canadian team. Then again, back then 16/21 teams made the playoffs, so maybe that's a topic for another day.

More urgent than realignment is scheduling. The league recognized the need to have every team play every other team at least once each season, which is fantastic. However, to make up the three extra interconference games necessary to fill out an 82-game schedule, bonus games were added arbitrarily. For example, this past season, LA played New Jersey, Washington, and Philadelphia twice. Nobody knows why, although I suppose getting to see Bruce Boudreau twice was fun for the former Monarchs, and with Dean Lombardi and Ron Hextall in the front office, an extra trip to Philly is nothing to complain about. So, as it stands, teams play 6 games against 4 division rivals (24); 4 games against 10 conference opponents (40), and 15+3 (18) games against teams from the opposite conference. This is pretty good, but those three interconference bonus matchups should be against teams who finished in the same place the previous season, a move facilitated by there being an even number of teams in each division (at least until Jim Basille gets the expansion Hamilton Blackberries skating). While we're only talking about three games, there could still be a mild compression effect.

The league is already considering an 84-game schedule that allows each team to visit every city every year: 6 games against division rivals (24), 3 games against 10 conference foes (30), and 2 games against interconference teams (30). The extra home game in the intraconference games can rotate annually but track by division—i.e., Pacific teams would visit Northwest teams twice but host them only once; Northwest visits Central twice but hosts only once; Central visits Pacific twice but hosts only once. Rock paper scissors, flips to srossics repap kcor every other year.

Should the league find a way to expand to 32 teams, I would hope that the league would consider restoring the classic division names and realigning to 4 divisions of 8 teams each. In this scenario, the 4x7 division (28)/3x8 conference (24)/2x16 interconference (32) plan for an 84-game schedule would be best. It would be critical following expansion for each market and arena to get a visit from every team every year, even if the travel costs eat away at the margins. Furthermore, rivalries are born in the playoffs, not redundant regular season division matchups.

OK, on to game play.

Eliminating the two-line pass has helped counter the neutral-zone trap that had, along with extensive clutching and grabbing, crippled game flow. The problem with the NZT is that it works: New Jersey won three Stanley Cups between 1995 and 2003 while playing a borderline unwatchable brand of hockey. Martin Brodeur is a genius, but, there are only so many 2-1 games dominated by possession changes in the neutral zone that a fan can watch before those hours are better spent elsewhere.

The trapezoid restricting goalies playing the puck is a mixed bag: freeing up the corners allows teams with less skilled forwards to focus on gaining the zone through dump-and-chase and relying on their forecheck to establish an offensive cycle. However, goalies skilled at making the long outlet stretch passes are not as likely to be able to take advantage of the elimination of the two-line pass rule. The fact that teams on the penalty kill are allowed to clear without being whistled for icing is a little strange, actually; they suddenly have a tactical advantage allowing a full-length dump for a line change. However, I'm too much of a purist to deny home fans the chance to cheer their PK units for quality clears that frustrate the opposition. The legality of goalies contacting the puck outside the trapezoid is logically related to icing, and should be connected as follows...
Proposal: allow a goalie to play the puck outside the trapezoid and behind the goal line if a) his team is on the power play, thus reducing the time killed by a clear and marginally increasing the number of scoring chances each power play opportunity should yield; b) if the opposing team has just iced the puck, such that a team could choose to touch up for an offensive zone draw with the offenders prevented from changing skaters OR keep momentum going with a stretch pass that might prevent the opponents from changing lines anyway.

Automatic offensive zone faceoffs following penalties are terrific for improving the chances that the attacking team will get their power play set up early and often. However, minor penalties that are interrupted by an intermission between periods unfairly dilute that advantage.
Proposal: if a period starts with one team on the power play, drop the first puck in the attacking zone. Center ice faceoffs look great on TV but most broadcasts miss them anyway while they're talking about...nothing in particular.

Requiring a full zone clear on a delayed penalty rather than a possession change or freeze has been covered by PuckDaddy, but I'm in favor of it.

When a puck is shot or deflected out of play in the offensive zone, the faceoff stays in if it was last touched by the defenders or the crossbar/post. Great. Why not keep the faceoff in the zone following any shot, regardless of who touched it last? There is absolutely no circumstance in which an attacking team would deliberately shoot the puck out of play; keeping the possession alive is paramount. If the man doing the dirty work in front of the net is taking crosschecks from a defenseman just so that he can set up a screen and get a deflection, and he gets a little too high or opens the blade up a little too much and puts the puck into the protective net, why should the faceoff come out of the zone? The defending team is already perfectly happy to accept the freeze and change lines; their gaining further advantage with the faceoff coming outside is unnecessarily punitive.
Proposal: faceoff stays in the zone following any shot attempt that results in a puck out of play. Reward shooting.

Whistles are bad. Visors are ubiquitous. Knocking the puck out of the air with a high stick is...what, a safety rule? But...there's already a penalty for high-sticking, right? So...if...I...use...enough...ellipses...
Proposal: don't add whistles and freeze play for something that is pretty much irrelevant. If someone fouls another player with a high stick while trying to play a puck out of the air, he's an idiot. Maybe it's not worth it to risk an eye to save a few whistles, but, it's a silly violation punished with a defensive zone faceoff.

Here's a fun one. Everyone loves penalty shots awarded during a game (shootouts are another story). Why not award the minor penalty if the penalty shot fails? If a player without great shootout moves gets a breakaway, why wouldn't a defender just take him down? Sure, it's a penalty shot, but it's one awarded to a player with average skill going up against a goalie with plenty of preparation. The PS fails, play continues, no power play. Not really much of an advantage to the offensive team. Success rates for penalty shots hover around 31%; for power plays, roughly 20%. Giving the minor after a missed penalty shot would bring the expected value of that PS/PP combo up to 0.5 from 0.3, and unless the league can come up with some way to make penalty shots more frequent, that increase would be a good thing. Along those lines, there's a difference between penalties taken during possession and positional battles and penalties that directly impede scoring chances...
Proposal: award a penalty shot for any penalty that directly prevents a shot on goal within a specified area on the ice (rectangle between the offensive zone faceoff dots and the goal line, or a trapezoid defined by the goal crease and faceoff dots), regardless of clear path to the net. Following any failed penalty shot, award the minor penalty that would have been called anyway. This will not only increase scoring, but also make it easier for referees to decide whether to call a minor or a penalty shot.

This post is getting longer than I'd originally intended so I'll sign off with a couple requests of the online and print media responsible for publishing box scores: try organizing the individual players according to the line combos that started the game. Even if they get changed up over the course of a game, it's a valuable aid for fans...at least those of us who can't be bothered to deconvolve the combos and pairings from the plus/minus and time on ice stats.

RIP Peter Zezel 1965-2009.

Thursday, May 21, 2009

MLB Interleague Play

This is the first of what will be a multi-part series of posts on scheduling in pro sports.

Baseball has an untapped source of excitement in interleague play. While players complain about visiting cities that are neither historical nor geographical rivals, the league correctly cites the increased attendance as a mark of success, as the novelty of seeing a team that doesn't normally visit your market is a significant motivation for casual fans--the sort that catch one or two games a year--to take the family to the park.

Mathematically, however, MLB is failing to make the most of the potential for interleague play to provide competitive enhancement commensurate with its entertainment value. The All-Star Game tie debacle and subsequent assignment of World Series home-field advantage to the winner of an exhibition game was a knee-jerk reaction and poor execution of logic. Furthermore, the sequence of expansions and realignments that left one league with 14 teams and the other 16 was absurd.

Currently, interleague play is combined into large schedule islands, with the matchups rotating year-to-year at the divisional level, with some exceptions made to preserve regional rivalries and balance out the number of games played (a necessary artifact of there being 14 AL and 16 NL teams).

If instead, interleague matchups were based on the previous year's final standings, two enormously positive consequences may result: first, marquee matchups, most likely including both a rematch of the previous year's World Series and a preview of the upcoming fall classic; second, natural compression of the standings, as weaker teams would have marginally weaker schedules and should theoretically last longer in the pennant races. The net result of such a set-up could be a season with no team boasting a 0.600 winning rate and no team at 0.400 or worse.

Realignment to three divisions in each league and the addition of the wild-card extended the pennant race intrigue deep into September: brilliant. However, the scheduling imbalances caused by having leagues with different numbers of teams and 4-, 5-, and 6-team divisions make things murky at best. The following steps would clear all of this up and add a ninth race to the eight playoff berths:

Step 1: Put the Brewers back in the AL Central.
Step 2: Put the Royals in the AL West.
Step 3: With 6 divisions of 5 teams each, scheduling 162 games is simple: 16 games against the other 4 teams in the division (64, a 5-game and a 3-game series in each park); 8 games against the other 10 teams in the league (80, a 4-game series in each park); 6 games against 3 teams from the other league (18, a 3-game series in each park): one from each division that finished in the same place as your team last year.
Step 4: Make the overall AL vs. NL interleague record decide WS home-field advantage, with the ASG as a tiebreaker. This will create a ninth race that will not be decided until as late as the playoff participants themselves are ironed out--and there can be a ticker on the standings that updates with each day's interleague contest.
Step 5: With 15 teams in each league, interleague play must be happening every night, thus potentially creating an additional point of interest in September, as the AL vs. NL composite record is yet undecided. Lost are the weird mid-summer interleague sideshows; gained is the "Tonight's Interleague Matchup" tagline on Sportscenter.