Rating distribution DF season 3 midseason 3v3 percentiles

I have noticed a lot of people talking about percentages and cr. Saying things like “1800 is top 5%”, “No it is top 10%” etc. This got me thinking, what are the numbers and percentiles?

WARNING: This post is my ATTEMPT to make sense of the data. I may have made mistakes and that is ok. This project is meant to give a educated estimate to the percentiles.


  • After much consideration and consultation with rank 1 streamers I used the number 66 as the base for the calculations. This is, at the time of writing, the number of eligible players for rank 1 on NA 3v3. ie 66 players make up the top 0.1% of the NA ladder. This is displayed on checkpvp. Those that frequently check the ladder may notice that 75 characters are eligible for r1. Remember that blizzard recently changed the rules for r1 and hence I came to the conclusion to use 66.

Factors to consider:

  • The data was collected on 01.09.2024 (US). These numbers will change very quickly and are likely different from when you read this post.

  • Sample size: Due to limitations in sample size I was only able to find all players with atleast 1200cr in 3v3. Hence the percentiles will not be completely accurate as I cannot find people below 1200cr. This begs the question who should be included? I believe that having 1200cr is a good starting point, it eliminates people who dont participate in 3s. I also predict the number of people between 1-1199cr is small and shouldnt drastically effect the accuracy of the published results. Obviously we are not considering people with 0cr either as this would more than the entire 3s ladder.

  • This post is not intended to give my opinion as to where the percentiles SHOULD fall across the ladder (I do that in other posts). I am only posting this to show my findings to those interested.
    *The 3v3 ladder is more competitive than others. People who queue 3s are considered more sweaty than other pvpers in this season.

The inverse percentile list (3dp):

  • 2700 → 0.054%
  • 2600 → 0.200%
  • 2500 → 0.539%
  • 2400 → 1.272%
  • 2300 → 2.199%
  • 2200 → 3.966%
  • 2100 → 6.999%
  • 2000 → 11.164%
  • 1900 → 17.057%
  • 1800 → 25.823%
  • 1700 → 37.702%
  • 1600 → 54.268%
  • 1500 → 73.173%
  • 1400 → 85.019%
  • 1300 → 93.260%
  • 1200 → 100.000%

The percentage of players within each 100mmr bracket is as follows (2dp):

  • 27+ → 0.05%
  • 26-27 ->0.14%
  • 25-26 ->0.34%
  • 24-25 ->0.73%
  • 23-24 ->0.90%
  • 22-23 ->1.77%
  • 21-22 ->3.03%
  • 20-21 ->4.17%
  • 19-2k ->5.89%
  • 18-19 ->8.77%
  • 17-18 ->11.88%
  • 16-17 ->16.57%
  • 15-16 ->18.90%
  • 14-15 ->11.85%
  • 13-14 ->8.24%
  • 12-13 ->6.74%


*There seems to be a cluster of data around 14->1600 (expected). A steady decrease follows with a sharp drop off at 2500cr.

*No normal distribution in graph form (I couldn’t post here as I have no image-posting rights :frowning: ).

Resources: check-pvp.fr


  1. What do you think about this data?
  2. Did I make any mistakes? Please feel free to politely correct me.
  3. Did you expect the data to look like this?
  4. Should Blizzard release offical data like this in a blue post?
  5. Do you want to see any other rankings like this?
  6. How do you expect the data to change?
  7. Nerf sub?

Thanks for reading. Have a lovely day and good luck in your seasonal goals!


Neat post! I didn’t check the math, but it looks accurate from a cursory glance over. Also shows the normal distribution centered right around 1500 which makes perfect sense!


Thanks mate! This is one defs one way to kill time during RSS queues.


Woulda been neat to see a before and after the emergency mmr injection.


Cmon blizz release a post with data like this! When was that anyway?


3 weeks ago

1 Like

Thats right. I did read this at the time. I havnt been playing much recently, how much did they buff it by? Anyone know?

I’m not sure exactly, but it was SIGNIFICANT. At any rate, based off of the data you’ve collected, it seems like inflation should stop now, and was possibly even slightly overbuffed. Seeing as gladiator historically has been the top 0.5%, that would put gladiator at just above 2500 with the current inflation.

Based off of that, if we were to go back to % based titles, we’d actually need more DEflation to have them match up to historically compared equivalents.

1 Like

The entire system needs a rework. League based mmr is the way to go.


It’s tough because I honestly believe that at least 20% of the population in pvp is probably better than 97% of the pvp population was back in legion when there were percent based titles. It might not be that dramatic but I got rival back in legion and man I had absolutely NO IDEA what was going on at all. It’s pretty apparent people play the game for rewards. If we transitioned back to a % based that was anywhere near as exclusive as it used to be, I think it’d just be a kiss of death for pvp altogether. You’d have a huge group of people that have been getting elite, duelist and all the way down that would suddenly not be able to. Most aren’t going to “get good” to get back to where they were. They’ll just quit. It’s not like me suddenly being unable to get duelist over night is a problem I need to solve. It’s just not that important. Continuing to dump time into something that sent me backwards “just cuz” seems worthless. Others may disagree but I have a feeling I’m in the majority here.

1 Like

I am 100% sure rhe people asking for %based rewards are banking on blizzard making the tiers for rewards similar or “easier” than what it is right now.

Their biggest excuse of “Im facing gladiators at 1600” is void and meaningless because that is one match while the gladiators skyrocket to their mmr tier while they are just sitting at 1600 getting smashed by bettee players quiclly climbing or fightin their equals and losing winning losing winning


I mean acting like gladiator players playing well below their skill level isn’t an issue is turning a blind eye to something that does actually have an impact imo. Go checkpvp any 10 randomly selected players that are currently above 2400 in basically any bracket and get an average of how many alts they have of the same class/classes they have equal experience on. Take that number multiplied by 10 placement games times the number of players that are above that rating and you can get an inkling of how many games lesser experienced players are getting roflstomped by high end ladder players. It’s not a non-issue. You take outliers like pikaboo and nahj who are constantly on stream doing viewer 2s and dumping on players who have no business queing into them and that’s literal thousands of matches every season where a guy who’s still learning the ropes is getting dunked on by a metaphorical MJ in his prime. I’ve played the game long enough to know it’s not a big deal for me personally and I’ll shrug it off. But for somebody even semi new to the pvp scene, it’s a recipe for them just not sticking with it. Heck, anecdotally I just qued my first shuffle of the season on my pres evoker for kicks the other day (~1700 mmr) and the first thing I saw was an enh sham on the s1 df glad mount. He went 5-1. There goes the experience of 5 other players in the lobby for however long they waited in que to be a part of that. Fun times.


You are absolutely correct.
Based off of this data, elite sets would be at about 2150 (which honestly isnt far off the 2200 they used to be).

I think more people would be upset by this than enjoy it.

It reminds me of when they brought back the mage tower because SO many people advocated bringing back “challenging content” and then there were THOUSANDS of complaints from the same people that said it was too hard and looked for a million excuses as to why they couldn’t do it.

Im all for percent based titles and clarity of rating between seasons.

I think most people just ask for that because they want rewards easier.

The sample size could be affecting a lot here. I see you chose 1200 as the cutoff for what “counts” towards calculations but blizzard obviously has there own system and it isn’t 1200. I used to know what it was but I believe it has changed. I think it was a mixture of a certain number of games played and minimum rating of around 1000 to be counted, but I could be mistaken.

Regardless, if it was what I said above, instead of 1200…that could equal MASSIVE changes to the percentages. Also, if .1% means 66 spots, couldn’t you simply multiply the number of spots times 5 for the real .5% cutoff?. This would mean top 330 spots which currently sits at about 2445 for 3v3.

Anyways, it’s important that relative difficulties in obtaining these titles have consistency and integrity. As it is right now, you have times (like the end of shadowlands S2) where gladiator literally ends up being acheivable at top 4%+. Then you have beginning of seasons like this one where Gladiator isn’t even obtainable by the top .1% for 6 weeks+. Those kind of variations wreak havoc down the rest of the ladder and have led to people having large misconceptions on what their “skill level” actually is.

When it is percentage based, people know where they stand and it is what it is. If you think you’re better than top 3% for example, then f’ing prove it and get your rating within top 3% or f off.

Lastly, if you really are worried about people quitting because titles are too hard to get you could change them to the following:

top 60% Initiate
top 50% Combatant I
top 45% Combatant II
top 35% Challenger I
top 30% Challenger II
top 20% Rival I
top 10% Rival II
top 5% Duelist
top 3% Elite

End of season reward:

top 1% Gladiator
top .25% Rank 1

This essentially makes each reward a bit easier to achieve than it historically was, and adds rewards for those historically left completely out at under top 35%.


So as I predicted the ratings changed dramatically. What I didnt predict is that the base number would change this much. It moved from 66 to 77 in a few hours.

Added some minor updates. Aditionally, what do you think causes more people to be added to rank 1. 24h ago it was 66 but now it is 76? I understand this can happen from more people playing in the bracket but surely not this quickly. Is there something I am missing?

basically. you have people asking for elite transmogs at 1200 rating now.

Do you even know who played during this time, cause that’s a statement to make ngl.

Elite transmog starts @ 1k w/bracers currently.

1 Like

I mean I’m not trying to say people back then were bad. There is just soooo much more information available now about how to increase your skill, weak auras are pseudo required, and I think a larger percentage of players that were on the lower half of the ladder have quit pvping than the upper half of the ladder since legion. You have a large part of the pvp population that has been increasing their skills in arena for years now. Idk maybe I’m crazy but I feel like your average 1800 player today would wreck the version of me from legion and I received rival for finishing around 2k. The skill floor has come up over time and despite a larger percentage of players receiving gladiator or duelist today, when you compare how good we all were at the game back then I think just as much or more time/effort needs to be invested to receive those titles as before.

Man, to go back to BC season 1 when there were legitimate back peddling and clicking gladiators lol. It’s true the average level of skill has gone up a bit over the years.