I streamlined my old 4-week, 10-team Finals idea

[The Educator]

This is well thought out and creative system

You've touched on the underlying principles that you are emphasising - e.g. improving incentive for finishing first, avoiding repeat games etc.

It is important to be explicit about these as an entry point for any chance to have a sane informed discussion on this topic given that people don't actually get the complex dynamics (including the impact it has on the regular season) to start off with so what chance to do they have of relating these to the set of desire principles?

Given the prevelant ignorance, you end up with morons shouting down anyone who doesnt agree with them - asserting that their set of priority principles are ordained by god and throwing around very basic probabilities and confusing it with real analysis

My issue with your system is it flattens the payoff too much between 4th and 9th (and to a lesser extent 10th). This waters down the implications of matches for teams that are heading for that part of the ladder in the last third of the season.

The advantage of the AFL's final 10 over this model is that it creates more "cliffs" that ensure teams are likely to be playing for material outcomes late into the season. (i.e. 4th to 5th, 6th to 7th). It also puts 7th to 10th 4 wins rather than 3 wins from a GF berth

Another factor counting against your system is competitive balance. 1st versus 10th will generally mean a lopsided final - replacing a far more likely closer match in 1st versus 4th.

The reason we shifted from the mcyntire final 8 (during the 90s) to the current final 8 system was due to the importance of competitive balance and to a lessor extent clarity of outcome. That system obviously did reward 1st and 2nd more than the current system does but it turned out that having 4 better balanced first week games with clear outcomes for both participating teams were playing for was superiour.

Your system is somewhat of a hybrid of the essence of the two final 8 systems.

 

[steve-o]

It is important to be explicit about these as an entry point for any chance to have a sane informed discussion on this topic given that people don't actually get the complex dynamics (including the impact it has on the regular season) to start off with so what chance to do they have of relating these to the set of desire principles?

Given the prevalent ignorance, you end up with morons shouting down anyone who doesn't agree with them - asserting that their set of priority principles are ordained by god and throwing around very basic probabilities and confusing it with real analysis.

I agree with you, but took a different approach: instead of inundating everyone with information out of the gate, keep it simple and clear up any arguments or confusion in the comment section. I'm looking for feedback, it makes sense to stay active in the thread.

My issue with your system is it flattens the payoff too much between 4th and 9th (and to a lesser extent 10th). This waters down the implications of matches for teams that are heading for that part of the ladder in the last third of the season.

The advantage of the AFL's final 10 over this model is that it creates more "cliffs" that ensure teams are likely to be playing for material outcomes late into the season. (i.e. 4th to 5th, 6th to 7th). It also puts 7th to 10th 4 wins rather than 3 wins from a GF berth.

I understand the "cliffs" idea - the current format has them every two seedings - but two things:

• I consider Week 3 rest disadvantage as a contributing element in those "cliffs." You really don't want to play a team coming off of a bye. I'd love to guarantee a discrepancy in Week 3 (and it works in a 12-team format), but that's even more flattening between 3rd and 12th.
• I prefer the "cliffs" to be syncopated if they have to be in pairs: the 1st seed should be signficantly better than the 2nd. The current format incentivizes being in the Top 2, but not the Top 1.

The analysis on these incentives is much easier without the double-chance, which is why I suggested straight-bracket Top 7 as well. Top 11 isn't bad, Top 13 is probably the best, but neither of those are palatable numbers for most fans until/unless the league gets really big.

Another factor counting against your system is competitive balance. 1st versus 10th will generally mean a lopsided final - replacing a far more likely closer match in 1st versus 4th.

The reason we shifted from the mcyntire final 8 (during the 90s) to the current final 8 system was due to the importance of competitive balance and to a lessor extent clarity of outcome. That system obviously did reward 1st and 2nd more than the current system does but it turned out that having 4 better balanced first week games with clear outcomes for both participating teams were playing for was superiour.

Your system is somewhat of a hybrid of the essence of the two final 8 systems.

I'm definitely in the camp of clarity-of-outcome, it's why the NRL stopped using that system after a while. That said, I'm not concerned with clarity of opponent - it'd be nice to know at least where you're going next after a final; I think this can be done with scheduling.

On competitive balance...well, the 1st vs. 10th game here MAKES the format work. In most finals/playoff formats, the lowest seeded teams aren't really expected to make a strong push for the Championship, their primary function there is to be a little bit of live-target cannon fodder for the top teams. Unless you're looking at 4-5 teams in a finals, there's going to be some fodder.

---

I don't disagree with anything you've said, I think we just disagree on what's more important when considering a new format. Thanks for the feedback, it's helpful.