Sunday, May 13, 2007

Fair Competition II: On the Field

“A playoff would settle things on the field”

vs

“Any team can beat any other team on any given day”

Ahh… the ‘on the field’ argument. Always a fun one. And seemingly tricky as hell to deal with. But it’s not really, mainly because we’ve already dealt with very similar arguments – it’s yet another example of each side trying to poach on the other’s territory and use the other’s weapons with both sides ending up proving each other’s points. The first thing that will help us is to realize what on the field means. When you look at how the phrase is being used, it’s the same thing as ‘on that day’. As in, on that day better. And as we’ve been saying, for clear communication purposes we need to keep the betters separate – the objective on that day better is not the same as the subjective seasonally better. But besides keeping the game and season grammatically separate, do we have to keep them separate when evaluating performance too? The pro-playoff side says no, the anti-playoff side says yes.

This difference is all in how each side looks at the relationship between the game and the season. The pro-playoff side prefers to work with games, since they’re objective, and will always default back to them in the event of a conflict, while the anti-playoff side prefers to work with the subjective season as a whole and will always default back to the big picture in the event of a conflict. Based on these tendencies and combined with their overall goals, we can see that the pro-playoff side believes that in some circumstances you can expand a single game out into subjective territory so that it has seasonal meaning. The anti-playoff side believes the only way a single game can cross the boundary into seasonal meaning is when it’s taken as a whole with all of the other games. Stick with me – here’s the examples.

The main argument that the anti-playoff side uses to build up this boundary wall between the objective game and the subjective season is the upset. For instance, if 1-3 Stanford beats 4-0 USC, does that mean that the Cardinal are better than the Trojans? Emphatically yes and no. Yes, they were certainly on that day better, but just as certainly not seasonally better, before or after that game. Anybody who combines the betters or crosses that objective-subjective boundary to say either that 1) the Trojans were better on that day or 2) that the Cardinal were seasonally better, is either a huge homer or delusional (or both, as is usually the case). The first statement, that the Trojans were better on that day, is defeated by the scoreboard. Quick aside – some people might try to argue that Team C played better than Team A even though Team C lost, but that’s a subjective trap we don’t need to fall into. Stripped down to it's barest competitive and objective parts, the team that scores the most points was better. Sure there's lucky bounces, bad ref calls, yadda yadda. In the end, at its very simplest, when the wins and losses depends only on points, so too must our assessment of who was better. Without that, competition is meaningless. So we can be absolutely certain that Stanford was better on that day. The second statement, that the Cardinal were seasonally better, is defeated by common sense. (Imagine that.) USC went 11-2 that season, while Stanford was 4-8. Even after the game in question, USC went to 4-1 while Stanford was 2-3. Even though determining who was better over the course of the season is subjective, it’s not completely irrational – we can be just as absolutely certain that USC was seasonally better. This example shows that the boundary wall between the objective game and the subjective season definitely exists. It’s easy to see when you’re dealing with two teams that have very different win-loss records, and both sides agree that the wall is needed in these cases. But is the boundary still impenetrable when we’re dealing with two or more teams with similar records? Again, the anti-playoff side says yes, while the pro-playoff side says well, not always.

Here is where we get to what I like to call the Betters Loop. We mentioned this one a bit earlier – Texas beat Oklahoma beat Texas Tech beat Texas. There’s no objective or common sense way to determine which of these teams was seasonally better since all of them went 11-1 in the regular season. Any attempt to use a single objective game to determine seasonal superiority throws you onto a merry-go-round, and the only way to get off is to get subjective. The Big12 stopped the merry-go-round by using the BCS rankings, which we all agree are subjective. You could also stop it using margin of victory, or strength of schedule, or any number of other statistics, but those too are subjective judgment calls – not everybody sees those the same way. If we’re dealing with the only objective tools we have, game results, we’re stuck. Even if we drop all three of those head-to-head games, we’re still stuck because all of them are now 10-0, seasonally equal from an objective standpoint. (Incidentally, the way to get out of many Betters Loops, such as Mississippi beat Florida beat Georgia beat LSU beat South Carolina beat Mississippi, is to use the boundary wall – yes, 7-6 South Carolina was better than 9-4 Mississippi on that day, but common sense and the objective win-loss records show that they weren’t seasonally better. Loop broken.)

Now we get into some of the nitty-gritty differences between the two sides. The anti-playoff side will say that the Betters Loop shows that there’s never any way to use a single objective game to determine who’s seasonally better. Not so fast, the pro-playoff side counters. This is where we have to bring the overall goals back in. The anti-playoff side’s goal is a best champion, so no, a single objective game can’t help determine who’s seasonally better. But it can help determine a clear-cut champion, which is the pro-playoff side’s goal. Let’s go to the example.

The pro-playoff side will often support their position by highlighting times when things were settled on the field, and I’m sure you remember one of their key examples. From the BCS controversy of 2000, “Well if both Florida State and Miami only have one loss, and Miami beat Florida State, then obviously Miami was the seasonally better team – they settled it on the field”. There’s no loop there – we have a two teams with similar records who are, for all intents and purposes, extremely close to being seasonally equal. Therefore, the pro-playoff side defaults to the objective, using the game between them as the tiebreaker. Miami won, so they get the edge.

But the anti-playoff side has a few counterarguments that they use here, all of which allow them to throw the pro-playoff side’s penchant for objectivity back in their face. The first is that Florida State won 11 games while Miami only won 10 – objectively, the edge can just as easily go to Florida State. Simple, and somewhat effective, but not as good as their next barrage. Another tack the anti-playoff side will take is saying that the pro-playoff side is incorrect in making the assumption that both teams were seasonally equal and that their head-to-head game needs to be the tiebreaker. For instance, without that head-to-head game, Miami was 9-1 while Florida State was 11-0. In that situation it’s hard to say that the Hurricanes weren’t seasonally better, so there was no tie between them for that head-to-head game to break. So that head-to-head game, while relevant to their seasonal success, cannot be used on its own as the single determining factor as to which has the better season. Had they been equal without that head-to-head game, with both teams at 11-0, such as in 2007 when Ohio State and Michigan met up, then that head-to-head game would’ve broken the tie and one would have been “proven” seasonally better, as Ohio State was. But that wasn’t the case. So basically if the pro-playoff side is going to argue that we have to use the objective head-to-head game to break a seasonally equal “tie” between these two teams, then they're arguing in favor of subjectivity, since they're saying that 9-1 is seasonally equal to 11-0.

It’s a solid argument, but now it’s the pro-playoff side’s turn to be snarky and use the anti-playoff side’s penchant for subjectivity to their advantage – they’ll argue that since “seasonally better” is subjective anyways, why couldn’t Miami and Florida State have been on equal footing before that head-to-head game is added in? They played different schedules, had different margins of victory, and it’s only one loss – teams with fewer losses are jumped by teams with more losses in the rankings all the time. So why can’t we say that Miami and Florida State were close enough in seasonal accomplishments to use the game between them as a tiebreaker? As usual, both sides have good arguments, and the “right” answer depends on which side you’re on and which definition of champion you’re going for. Since the pro-playoff side is more concerned with having a clear-cut champion, they’re going to default to a head-to-head outcome when two teams are close enough to seasonally equal – they don’t like playing in those subjective waters, so they go with the scoreboard. The anti-playoff side is more concerned with having the best champion, so they’re perfectly at ease with subjectivity and refuse to elevate one game above the others.

So (finally) back to our original arguments, all is kosher if by saying a playoff would ‘settle things on the field’ you mean that a playoff would ‘determine an objective, clear-cut champion’. It would. But if by saying a playoff would ‘settle things on the field’ you mean that it would automatically ‘determine the best champion’, it wouldn’t. We’ve just realized this in our discussion of what happens when each side attempts to claim the other’s champion with a team that’s not dominant. We’ve seen how quickly the on that day better argument breaks down when we’re dealing with a subjective, seasonal goal. At the same time, the anti-playoff side’s retort of “any team can win on any given day”, while true for them, is just as unrealistic when dealing with an objective, clear-cut champion.

Speaking of how much games should matter…

Fair Competition I: Switching Sides <> Fair Competition III: Making Meaning

No comments: