Ladder Play
On the largest Shoddy Battle servers, including Smogon University, there is a mode of play known as ladder play. Ladder play attempts to rank players in a fair manner that is immune to fraud, and it also provides a speedy and efficient matchmaking service. If you enjoy standard battles, ladder play may be for you.
§ How to Play
Playing a ladder match is very easy. Click the Find tab and then the Ladder tab. Load your team of choice and then press Find Match. Within thirty seconds, a match should open automatically.
§ Nature of Ladder Play
The first precaution taken in ladder play is that all matches are against random opponents. It is impossible to arrange a ladder match against an opponent of your choice. Moreover, the system attempts to match you to the player with the conservative rating estimate nearest yours. This results in matches that are enjoyable for both players, and prevents battling against a desired weak opponent.
Secondly, all ladder matches use the same rule set, which is server-specific.
§ Ratings
All ratings and the leader board are updated in real-time as matches complete. You can view a given player's rating by right-clicking his or her name in the client and choosing "View Record". A rating consists of several statistics that may initially seem confusing.
Conservative Rating Estimate: This is an estimate of your true rating that assumes you are as bad as you could possibly be, given your play history. As you play more matches, the uncertainty in your rating will decrease and the conservative rating estimate will approach your true rating. For the mathematically inclined, the formula for this statistic is given by:
conservative rating estimate = μ - 4ϕ
where μ is the mean rating and ϕ is the standard deviation.
Rating: This is the range that comprises one standard deviation on either side of your mean rating. In other words, there is a 72% chance that your true rating falls within this range. Your rating is increased by winning matches, and the width of the range is decreased simply by playing more matches, regardless of whether you win.
Volatility: The volatility factor is a measure of how consistently you play. If your rating does not change much each day then your volatility will be low, whereas if you improve or worsen rapidly your volatility will be high.
§ FAQ
What happens if I don't play for a few days?
Each day you don't play, the uncertainty (standard deviation) in your rating changes according to
ϕ′ = √(ϕ2 + σ2)
where ϕ is the deviation and σ is the volatility.
Basically, if you don't play, the width of your rating range will increase, which has the effect of decreasing your conservative rating estimate.
Where can I learn more about the Glicko2 rating system?
Consider reading this paper by Professor Glickman, designer of the system.
How can I tell how good I am relative to another player?
This table gives you the chance of beating somebody with a conservative rating estimate Δestimate points lower than yours:
| Δestimate | % |
|---|---|
| 10 | 51.44 |
| 20 | 52.88 |
| 30 | 54.31 |
| 40 | 55.73 |
| 50 | 57.15 |
| 60 | 58.55 |
| 70 | 59.94 |
| 80 | 61.31 |
| 90 | 62.67 |
| 100 | 64.01 |
| 110 | 65.32 |
| 120 | 66.61 |
| 130 | 67.88 |
| 140 | 69.12 |
| 150 | 70.34 |
| 160 | 71.53 |
| 170 | 72.68 |
| 180 | 73.81 |
| 190 | 74.91 |
| 200 | 75.97 |
| 210 | 77.01 |
| 220 | 78.01 |
| 230 | 78.98 |
| 240 | 79.92 |
| 250 | 80.83 |
| 260 | 81.71 |
| 270 | 82.55 |
| 280 | 83.37 |
| 290 | 84.15 |
| 300 | 84.9 |
| 310 | 85.63 |
| 320 | 86.32 |
| 330 | 86.98 |
| 340 | 87.62 |
| 350 | 88.23 |
| 360 | 88.82 |
| 370 | 89.38 |
| 380 | 89.91 |
| 390 | 90.42 |
| 400 | 90.91 |
| 410 | 91.37 |
| 420 | 91.82 |
| 430 | 92.24 |
| 440 | 92.64 |
| 450 | 93.02 |
| 460 | 93.39 |
| 470 | 93.74 |
| 480 | 94.06 |
| 490 | 94.38 |
| 500 | 94.68 |
| 510 | 94.96 |
| > 520 | > 95 |