The Rating Spreadsheet (9/17/15: Link has changed!)
tl;dr: Here’s a rating spreadsheet that probably won’t mean a lot to you, but I’ll use it as a tool to advise on godfests and relative monster quality.
Rating PAD monsters is extremely difficult. The game changes on a weekly basis. Each change has repercussions on existing monsters. On top of that, it’s nigh impossible to have meaningful firsthand experience using all of them as subs and leads. This meant I had to rely on secondhand knowledge to affirm my assessments which has its own series of problems. I was about to give up my reviews, but some part of me wanted to continue.
To prevent myself from drowning in a sea of updates, I needed to minimize my scope while being as flexible as possible. To identify where I wanted to limit things, I needed to identify what I wanted to use these ratings for. I found I wasn’t really interested in the minutiae of each rating nor did I want to write in detail about them. All I wanted was a simple scale so I could advise on the quality of a given godfest. Accuracy didn’t really matter as long as it gave a general sense of monster quality. I also wanted to make sure I didn’t need to readjust ratings due to power creep.
I decided on a simple rating system that graded each monster on a arbitrary numeric scale. To sort these ratings I created a rudimentary grading scale based on the standard distribution. Each monster is assigned a letter grade based on their position in the bell curve. This means I won’t have to re-scale ratings manually if there is power creep; I just need to give the stronger monsters a higher rating than the current highest. To increase flexibility, each criterion can be assigned a weight, allowing for distributions based on different emphases. To be as consistent as possible, each criterion is governed by a concise set of rules making it easier to assign a rating.
How Monsters Are Rated
For the most part, I want to keep the specifics of the rating criteria obfuscated because there really isn’t a systematic way to accurately appraise a given monster. From a high level, this first iteration will simply judge a monster’s power as a lead and sub in late- to endgame dungeons with an emphasis on the endgame. What’s the difference between the two? Well, to put it simply, you can clear late-game dungeon with suboptimal teams, but an endgame team has to meet a much stricter set of requirements. It’s the difference between the team that can clear Hera-Is vs. Hera Rush. I initially only factored in the endgame, but the average player struggles in the late-game, thus some weight is given to late game usefulness. For example, FA Lucifer is a very good sub in the late-game and almost unusable in the endgame, but his late-game power should still be a factor in the overall rating.
You can take a look at the spreadsheet itself to glean some of the methodology I use if you really want to, but unless you’re just curious I don’t think there’s a lot to learn. This is just a tool that still needs some translation to be useful.
The grading scale is based on the standard deviation of a given rating for a monster or group of monsters using a normal distribution. I’ve adjusted it to my own likings. I’m sure you’ve been exposed — or subjected — to something similar sometime in your life, but if you want to brush up check out the Wikipedia entry.
- S – 2σ or greater
- A – 1σ to 2σ
- B – -1σ to 1σ
- C – -2σ to -1σ
- D – -2σ or lower
By binning monsters into letter grades, you can get a feel of “about how powerful they are” without having to go through the trouble of being accurate enough to create a precise hierarchy.
Based on this grading scale, most monsters will end up in the B range. This is where I truly believe most monsters lay, something intentionally set up by GungHo. If the amount of exceptional monsters at a given time is small, this will give advanced players something to chase. Conversely, bad monsters must exist to make the good ones stand out, but you can’t have too many or players would quickly get fed up with the REM. If most players have monsters near the average GungHo can control the pace through which they advance through the game, enticing them with the release of newer, more exceptional monsters to get them past any walls they may hit.
I do believe the system works. I started the spreadsheet well before the Indian 1 and Egypt 1 awoken uevos were released, but after I factored them in their pantheon grades shot up. This also caused the dominant Heroes pantheon to come back down to earth as the distribution had shifted. We’ll see how it plays out in the future though, and make adjustments as we go.