[UNAC S3] Pre-Season Discussion

Orion said:

So the tiebreaker follows its own little ruleset and will be nothing like the other 99% of the tournament?

I don't see what's so bad about playing full maps until the thing is decided. When you're scoring by set wins your likelihood of going past a third map is quite low. Low enough, I think, that it's practically a non-issue. I assume someone that has put some thought into this would realize this (which means I assume you realized this), so that makes me question why tiebreaker maps aren't played like they have been in the past. What's the motivation for this change? Fairness and time constraints are all I can think of right now.

The stuff in spoilers is theoretical questioning and reasoning. You can skip that if you just care about my suggestion. If you have a question about my suggestion, read the stuff in the spoilers. You might find your answer, or at least figure out what questions to ask. And yes, it's really, really long. Sorry.

Spoiler

So, what's unfair about the way tiebreakers have been handled in the past? It seems like the consensus now is that one team got to pick the map, which gave them 2 map picks out of a total of 3. I agree that this is questionable, and I think that fixtures for tiebreaker maps are fine. Limiting the available pool of maps isn't a new idea (it goes back to NASTe season 2, where the tiebreaker map had to be from a different category than those previously played), and ensuring that maps are symmetrical is certainly a way to nullify any kind of inherent map advantage one team may have. Faction swapping was an attempt at solving this problem, but it's naturally prone to more ties if both teams can equally exploit the inherent map advantage on their "turn" for that spawn, so I'm OK with symmetrical maps and no swaps as well, to an extent. I think swapping is an interesting mechanic in regards to faction variety, because it gets teams to play things outside of their comfort zone and try new approaches. It can keep the metagame going as teams develop new strategies and counters to accommodate for factions they're not used to playing. For this reason, I think swapping is still a good concept and should not be wholly abandoned in tiebreakers, and it's impossible when factions are forced to be mirrored. For this reason, I am also against forced mirroring of factions in the tiebreaker.

This also means that you must have a match long enough to accommodate at least one swap, so that both teams can play both factions. This implies that an entire map is played, 2 sets by your reckoning, and allowing mirrored factions means there is no advantage in getting first faction pick. Along with maps being predetermined, this wholly negates any unfairness inherent in the picking system.

As for time constraints, I understand this is a very real concern and that matches in Warband are already considerably longer than in similar games. However, a single set is a relatively short period of time to resolve a match that is already quite close, especially considering the potential and decisiveness of snowballing. Removing round bonus gold somewhat alleviates this, but combat gold is still a very real concern because it still gives an advantage to the winning team. However, this line of reasoning runs counter to the fact that changing either/both of these deviates from the way all other matches are played throughout the tournament, and this sort of deviation raises questions about the validity of the tiebreaker as a whole. If it doesn't follow the same rules as everything else, then how is it an accurate representation of how two teams perform in the match? It's akin to having tiebreakers be resolved by a duel. The rules of the competition change, so the outcome of the tiebreaker is arguably not an accurate representation of the capabilities of the teams involved. For this reason, I think it is best to stick as close as possible to the standard match rules, and not have anything drastically different from the first 2 maps.

Which, obviously, still leaves the problem of time unresolved. In order to stay true to the standard match rules, you would play an entire map with a swap in the middle, and what if you're still tied after that? Well, regarding the problem of time required to play another whole map, you could be more lenient in regards to how many players are required per team during tiebreakers. Allow some players to leave and force the opposing team to match players. Your standard matches require 7 players to start, and I think a minimum of 5 during tiebreakers is still enough to emphasize teamwork over individual skill. This would allow 2 players per team to leave if necessary, and the likelihood of more having to leave is low. Personally, I think this is a lax and easy-going solution to adopt, because I believe such problems can be largely avoided at the team level with proper scheduling, but I know that sometimes there just isn't enough time to go around and deadlines have to be met. So, such a rule would accommodate those situations if/when they arise.

In the event of a tie after a third full map, you could evaluate the teams by other metrics. One possibility is round wins during the map. If my understanding of your current scoring system is correct, you play sets in first-to-four and the team with the greatest number of set wins is the overall victor of the match. Because of this scoring system, there are only 4 possible score combinations that results in a tie of set wins and rounds out of 30 possible score combinations when dealing with only 2 sets (10 ways for a team to win, 10 to lose, 10 to tie). This gives you a chance to tie of ~13% based on statistical probability alone. If you weight the probability by team strength, then you get results similar to what we've had in previous tournaments, which is to say very, very few ties. So really, making special rules to guarantee a victor in a tiebreaker map is kind of pointless. The odds of reaching a tiebreaker map are quite small, while the odds of playing a second tiebreaker map are even smaller.

My advice, which is general advice when creating rules, is to write rules that acknowledge the possibility extraordinary circumstances, and provide the means with which they can be handled. For tiebreakers, in the unlikely event of their occurrence, have one map predetermined by fixture. I recommend simply shuffling the list (guaranteeing that the same one isn't on top twice in a row) and saying "if you tie, play through the tiebreaker map list in order until the tie is broken." You can say "but what if they keep getting tied? they'll play forever!" Well, yeah, that's the absolute worst case scenario, isn't it? Be realistic. When has anything even approached that? I can't recall a single tournament match that went past 3 maps. It's just so unlikely to happen that it isn't statistically significant. However, with such a rule in place (play through the map list) you have the means with which to handle it should it happen.

Of course, let's be realistic here. If you tie on your first tiebreaker map, that implies you've already been playing for 2 hours or more. I would say, allow teams that tie on a tiebreaker map to reschedule to a later date, and give them a clean slate to play from. Argue all you want about how the players that show up may be different and it may change the outcome, fine, but the tournament measures team success and not player success and all of those players are on the same team, and you can't reasonably expect a team to continue playing non-stop until a victor is decided.

Perhaps you could allow mirrored factions but not force teams to mirror. Not every clan has the same strengths and weaknesses, and forcing one team to play a faction they're weak on but their opponent isn't is arguably unfair. This would allow both teams to pick to their strengths, and I think what's most fair is two teams at their strongest rather than two teams with different talents using the same tools. I also like the concept of swapping for reasons of metagame advancement and forced variation in play. This implies a map long enough to accommodate a swap, so I think a full map should be played, and to resolve possible ties in tiebreaker maps, round wins could be a secondary metric.

Given the statistical probability of a set & round wins tie on a single map (~13.3%, unweighted by team skill), the chance of a full-blown tie is acceptably small. In the event of a full-blown tie, allow teams the option of rescheduling or continuing to play through tiebreaker maps until one team wins. You could also factor in overall match round wins as a third metric, if you want to reduce the likelihood of ties further, though you would still need some way to resolve total ties.

I'll also add that unless you remove all forms of earning gold (kills and round bonus), you won't negate snowballing. Gold from kills is significant (100+10% of gear cost, IIRC), and because of forced MOTF spawns at 2 minutes more rounds are resolved by kills than by flag captures because of forced confrontation.



And here's a novel idea, if you don't read the post, don't respond to the post. I know lots of people don't like walls of text, but this is a complicated issue. The important parts are outside of the spoiler, and the explanations for them are inside the spoiler, to save you the trouble of reading all of it. If you want an explanation for something, open the spoiler first. I don't mind writing this stuff, but I do mind repeating it.

Click to expand...