This last week, there was trouble at the World Baseball Classic surrounding the tiebreakers used to separate three teams in group play, resulting in Mexico’s controversial elimination from the tournament. The tiebreaker invoked was based on runs scored against per innings pitched, including partial innings. The interpretation of “partial innings” (and changing communications about and calculations based on that interpretation) impacted Mexico’s strategic decisions at the end of a game they won, and left them believing that they had advanced to the next tiebreaker game when in fact they had not.
Mexico is filing a protest, and the issue is proving to be a challenge to the WBC’s public credibility. But given the nature of these tiebreakers, the WBC is lucky that their tournament didn’t cause a complete breakdown of what baseball even is. Sports that feature group/pool play, tournaments, seasons, and seeded playoffs are all surrounded by their respective metagame structures. When designing those metagames, you create hierarchies of importance for different victories – the championship is more important than the match.
You want your metagame structure designed so that lesser victories always contribute to the larger victories. The metagame shouldn’t change the base game. If, say, you wanted to add a layer of metagame achievements to your multiplayer shooter, and you layer on an achievement that rewards players for earning 100 knife kills, guess what? Your metagame just turned your shooter into a stabber.
Anyhow, I just wanted to explore the dark underbelly of the WBC’s tiebreakers and construct a scenario to see just how badly I could screw up the game of baseball, and I think I did a pretty okay job.
Pool play consists of four teams in each pool (we’ll call our teams A/B/C/D), each playing each other once. The top two teams advance.
We don’t need to worry about Team D. They win the pool with 3 wins. Their results don’t impact tiebreakers at all, so we will only look at the three games between the other three teams. Teams A/B/C each need to win exactly one of these three games to invoke the three-way tiebreaker.
Teams A and C each have one win. In the final pool game, Team B needs a win over Team C to stay alive and trigger the three-way tiebreaker. The top two remaining teams between A/B/C after a three-way tiebreaker will then compete in a single tiebreaker game while the third team is eliminated.
With a B win, A/B/C will each have a 1-1 head-to-head-to-head record, so the tiebreaker will fall to the next step, which is “fewest runs allowed divided by the number of innings (including partial innings) played in defense in the games in that Round between the Teams tied”. Runs against over innings pitched. RA/IP.
Going into the final pool game, Team B needs to win outright to ensure entry into the three-way tiebreaker and stave off elimination.
Here’s where we stand in our scenario going in to the top of the 9th inning of the final game, with home Team B leading away Team C by a score of 1-0. Teams highlighted in green won that game (or will need to win that game to trigger the tiebreaker). The current RA/IP for each team is listed in the upper-right:
Weirdness #1: Elimination with Win, Survival with Extension
If Team B gives up no runs in the top of the 9th, they immediately win, but they are eliminated because they have the worst RA/IP of the three teams. They won’t have a bottom of the 9th or extra innings to work with to impact the tiebreaker.
Team B is therefore incentivized to extend the game (which they also still need to win) to improve their RA/IP. Team B needs to allow Team C to score runs in order to stay alive.
Team B wants to give up exactly one run in the top of the 9th to tie the game. They could afford to give up more, but anything further that they give up they will need to score back in the bottom of the 9th. I think that the smartest way to attempt to give up exactly one run (to a team that is incentivized to receive it) is:
- Play the defensive inning as normal until you’ve recorded two outs or acquired one baserunner.
- If you’ve recorded two outs before acquiring a baserunner, intentionally walk the next batter.
- Balk repeatedly until your baserunner scores.
- Complete the inning to the best of your ability to prevent further runs.
Is Team C cool with this? Team C probably doesn’t have any problem with Team B giving them a free run and will choose to take it. First of all, that free run would give them a better chance to win this game outright, which would send them past the pool stage without the need to win a head-to-head tiebreaker game. The only way this would backfire for them is if they lost this game in a very specific way, which I’ll discuss in a later section. This is the low risk and high reward option, and almost certainly what Team C would choose to do.
Weirdness #2: Refusal to Score Until the 13th Inning
Team B, having extended the game by giving up a free run, has two options to advance to the tiebreaker game: Either win and have a better RA/IP than Team A, or win and have a better RA/IP than Team C. If Team B plans on beating Team A in RA/IP, goose egg innings are more important than game-ending runs. The fastest way for Team B to best Team A in this manner in our scenario would be to throw scoreless innings in the 10th through 13th innings before winning.
This also means refusing to put runs on the board in the bottom halves of the 9th through 12th innings, to keep extending the game as needed. If Team C was happy to accept Team B’s gift run in the top of the 9th, as we are assuming here, that means they’d also be happy to accept four innings of Team B refusing to score, should Team B choose to go that route.
Weirdness #3: Walk-off Grand Slam or Bust
However, there is still another route to the tiebreaker game for Team B – they could play to have a better RA/IP than Team C. This is achievable in our scenario if Team B beats Team C by four or more runs. By the time you reach the 9th inning of any game, the only possible way that the home team could win by four runs is with a walk-off grand slam. Winning by one, two, or three equals instant elimination.
This possibility means that Team B is actually playing real baseball in the bottom halves of the remaining innings. All baserunners will advance as far as they possibly can along the base path without heading home until they fill the bases, and they will only make a move towards home plate if a grand slam has been hit (or if runs were scored in the top of the inning).
Weirdness #4: Nega-Baseball
There is a non-zero chance that Team C would choose to play it safe and refuse the free run at the top of the 9th. They might try to score zero runs in the top of the ninth, immediately losing the game, but sending themselves to the tiebreaker game with Team A and eliminating Team B in the process. This may be a fine choice if Team B is a known behemoth that you want to strike down when given the chance, or if Team A is fluky cupcake team that just saw their entire pitching roster go down with injuries. That leaves us with a defense trying to allow runs and an offense refusing to score runs.
What happens, then, when two teams engage in this brand of nega-baseball? What happens if the goals are switched, so to speak? (For related reading, see here.)
It’s hard to know exactly who comes out on top here, since both sides have plenty of tools to aid their cause. Pitchers have walks, balks, and hit-by-pitches to work with to earn and advance runners. Batters/runners, on the other hand, have tons of ways they can intentionally get themselves out.
A pitcher could try to throw four balls, but a batter could still swing at every one, so intentional walks are out of the question for the pitcher. His only remaining option? Throwin’ beaners!
I don’t really know the eccentricities of the rulebook, but for the batter, one apparent surefire way to earn an intentional out (that I pulled from here) would be if the batter “steps from one batter’s box to the other when the pitcher is ready to pitch.” That seems pretty foolproof, and it prevents a pitcher from spending more than the slightest amount of time being ready to pitch.
If both teams adopt these strategies, baseball becomes target practice: The pitcher, immediately after he receives the ball, hurls it at the batter who he himself is attempting to leap across home plate before the pitch is thrown. Batters will try to earn three intentional outs before one of them gets hit and balked home. I think this is what cricket is. There’s a pretty clear batter advantage here, but who’s to say how the timing of these things will work.
The WBC declares that it will be played under the same official rules of the MLB. The MLB, in an attempt to improve the pace of play, is doing away with the need to throw four pitches for an intentional walk. Starting this season, a manager can indicate his intent to walk and it will be automatically awarded.
Is the WBC using the 2016 rulebook? I don’t know. If/when the WBC adopts this, the nega-baseball advantage will shift heavily to the pitcher, I’d think. On the other hand, you can’t force a balked-in batter to touch home plate, so who knows. Baseball.exe has stopped working.
Weirdness #5: Intentional Mercy Rule
The WBC additionally institutes a mercy rule in its games. If a team leads by 15 at the end of an inning after the 5th inning, or if they lead by 10 after the end of an inning after the 7th inning, the game immediately ends in victory. This year in the WBC, there was precedent for a ‘walk-off’ mercy rule invocation, where the game ended in the middle of a team’s inning when their lead hit 10.
Given what I’ve laid out above outlining why a team might need to extend a game, it stands to reason that the opposite is also true – a team may have a pool play reason to shorten the length of the game with an intentional loss. Say that in the final pool game, the away team is already trailing 9-0 in the top of the 7th. The away team has one win under their belt, and their opponent needs the win to enter the three-team tiebreaker (of teams with 1-2 records). The away team may find it preferable to choose to score nothing in the top of the inning in order to invoke the mercy rule during the bottom of the 7th. They might do this to avoid giving their opponent more opportunity to improve their RA/IP, or to avoid taking additional risk squandering their own.
The complete construction of such a scenario is left as an exercise for the reader.