---What's with the Ladder?  A mini-guide by Ataraxia---

*Why are my points going down when I win?
*Why does someone with less wins / more losses / a worse record than me have a higher rating than me?
*What is the 'Bonus Pool?'
*Is the ladder going to get fixed?

---Intro---
Currently, we use the ACRE ranking system to calculate a given player's rank on a specific ladder.  One of the features of ACRE is that each player you defeat / lose against affects your rank differently.  Generally, defeating a higher-ranked player will earn you a massive amount of points, while defeating a lower ranked player won't affect your overall rank nearly as much.  The converse is same for losing battles.  Losing to a lower ranked player will result in a massive drop, while losing to a higher ranked player will barely affect you.

This feature is a beneficial one for a game like Pokemon.  Heck, it's more or less built into the cartridge games.  Defeating a level 2 Pidgey with a level 80 Tyranitar won't even increase your exp bar by a full percent. Yet, defeating a level 80 Tyranitar with a level 2 Pidgey would result in the Pidgey leveling up several times.  Being able to calculate the ranks of players involved in a battle, and using that to determine the points won/lost, was a major point of adopting the ACRE system.

However, the system was designed with the idea that people of generally similar ranks would be matched together.  It prefers to assume that people with, say, a rank of 1350, would be playing people that are near that rank.  Someone as low as 800 wouldn't be considered, as wouldn't an 1800.  Plenty of games that use ACRE designed their games to work around this mechanic.  However, it's not entirely compatible with the way we match people (first two people in go against each other, repeat).  We don't check for the rank of players, so it's possible for a 1000 to play a 2000 in a ranked match, and that is just an example of where things begin to get weird.

---Why are my points going down when I win?---
First off, if this were a betting game, anyone would bet on a rank 2000 beating a rank 1000.  It's the expected outcome.  In fact, it's not even a surprising one.  Few people would want to watch a level 80 Tyranitar fight a level 2 Pidgey.  Heck, most trainers just run away, as the battle itself would be a waste of time and PP.  ACRE thinks that way too.  So, when the 2000 beats the 1000, it's not going to reward them with anything, or do much to punish the 1000.  However, if the difference in ranks is high enough, ACRE would actually expect that the 2000 player should've done more than just simply win against them.  Just as we consider it a waste of effort to fight that Pidgey, ACRE considered that victory a waste of effort.  So, it actually ends up docking a few points from the winner.  Remember, ACRE isn't Pokemon-specific.  Many games have the ability to make that logic work, in such ways as killing multiple enemies / earning in-game bonuses / etc.  Pokemon doesn't have such mechanisms (and arguably, doesn't need them).  So, you end up losing a few points for that.

---Why does someone with less wins / more losses / a worse record than me have a higher rating than me?---

Let's go back to that Tyranitar / Pidgey battle.  Suppose, against all odds, the Pidgey won.  Now, in-game, we know that the Pidgey would level up a ton of times, get a ton of new moves, and probably evolve into Pidgeotto.  ACRE rewards winners with points, which we use to represent your rank here.  So, if a rank 1000 player beats a rank 2000 player, that 2000 player will take a massive drop to their rank, and the rank 1000 will get a massive increase in theirs.  That victory will be worth more to that player than any victory against the other rank 1000's they'd be matched up against (from a statistical standpoint).  If a relatively new player were to do nothing but win against high-ranked players, they would in turn gain an excessively high rank, even if they only did a small handful of battles.  So, with a bit of luck, it's entirely possible that someone tops the ladder with a tiny record, due to their past opponents.

---What is the Bonus Pool?--- (Why did the player that I beat get more points than me?)

In addition to having your score vary on the ranks of players that you played against, there is also another factor that affects your rank here: Decay.  Over time, as you're not playing, your rank will slowly go down.     It's completely normal, and to be expected.  Hypothetically, during that time that you're not playing, the players around you are all getting progressively better while you remain stagnant.  This is where the Bonus Pool comes in.  It gives those players back some of the points that have decayed over time.  This decay tends to happen fairly fast, so it's entirely possible to earn something like 2 points in a battle that you started immediately after a previous one, if that battle goes on for long enough.  Let's say that you just got done with a battle, and you get into another one.  That player's rank is close enough to yours, and you beat them.  So, you get a couple points for winning, and 1-2 from the bonus pool.  They lose a few points for losing, and gain 90 bonus pool points, so they actually end up leaving with more points than you do.  It's nothing they did: rather, it happened that way because it was likely their first battle that day.

---Is the ladder going to get fixed?---

Short Answer: Yes

Long Answer: Eventually.  Right now, there is nothing wrong with ACRE itself.  It's doing the job as it's expected to.  People are being shoved around the ranking system in accordance to how well/poorly they perform (regarding wins/losses/draws).  After all, those are the only three things that we care about,  There's no objective difference between a 6-0 blowout and a 6-5 luck finish with a pokemon at 2% remaining HP.  They are both wins, and are both recorded equally as such.  However, ACRE works best with a large userbase and some sort of in-game matching system that pits people of similar ranks against each other.  Though we consistently maintain a respectably-sized userbase, it's nothing like the thousands of concurrent users that make ACRE (or most other ranking systems) run optimally.  And as for similar-rank matching, we currently don't do that, as it's much more efficient to match the first two people in a queue together.  Think about an rank 1800 LC person in a queue, waiting for someone around his rank to show up.  They would likely never get in a match if we forced similar ranks to queue.

Simply put, we don't provide the optimal environment for ACRE.  It can still work, of course.  The people at the top of the ladder definitely showed a sensible amount of skill to get there (bar some potential trickery), and the people at the middle on bottom also displayed skill respective to their current locations.  The main issue is that, because ACRE isn't working at its best, it tends to lead to silly results, such as a 17/0/0 person ranking higher than a 40/0/0, or a person losing points even after winning a battle.  I haven't been clued in as to what the actual fix will be, but there is definitely one in the pipeline somewhere.