S30 - A Deep Dive into the Top Clears

General Discussion
1

Hello!

With the close of Season 30, Rage and I are excited to share these highlights!

WEEK 12 HIGHLIGHTS

  • Crusader No-Set (+2.3)
  • Wizard No-Set (+2,0)
  • Zunimassa (+0.3)

CLEARS VS PARAGON

  • Average paragon for a clear appearing in this final week is 3749. Maximum is 8504 (Vyr). Minimum is 1356 (Trag’Oul).

  • 6 builds had paragon between 0 and 2000, 23 builds had paragon between 2000 and 4000, 7 builds had paragon between 4000 and 6000, 3 builds had paragon between 6000 and 8000, and 3 builds had paragon 8000+

CLEAR TIMES

  • Average time for a clear appearing in this final week is 6:24. Maximum is 14:12 (Jade). Minimum is 1:55 (Rathma).

  • 6 builds had time between 10:00 and 15:00, 23 builds had time between 5:00 and 10:00, and 13 builds had times below 5:00.

WEEKLY GAINS & LOSSES THROUGH THE SEASON

  • Largest single week gain: +15.7 tiers, by DH No Set in week 2.

  • Largest single week loss: -1.0 tier, by DH No Set in week 3, and Shadow in week 4.

  • Largest gain since week 1: +23.2 tiers, by IK.

  • Largest loss since week 1: -0.4 tiers, by Shadow. (Shadow and H90 were the two sets that finished week 12 with an adjusted clear lower than they had in week 1).

  • “Unstoppable Force”: Set with most changes of peak clear: No Set Wiz, 9 changes.

  • “Immovable Object”: Set with least changes of peak clear: Helltooth, 0 changes.

CLASS COMPARISONS

  • The strongest overall class was Necromancer, hitting an average of 158.9.

  • The weakest overall class was WD, with an average of 156.2.

  • The class with the largest spread (difference between its strongest and weakest set) was Necro, at 7.2 tiers.

  • The classes with the smallest spread were Barb and WD, at 4.1 tiers.

  • 6 sets finished the season below their peak in an earlier week: H90 (-0.2), AoV (-0.7), Shadow (-1.0), Trag’Oul (-0.1), WD No Set (-0.6), and Vyr (-0.1).

  • The difference between rank 1 Rathma and rank 2 Wiz No Set (2.6 tiers) is the same as the difference between Rank 2 Wiz No Set and rank 17 Monkey King. That’s a significant outlier. No other two builds next to each other in rank order (i.e. rank 7 and rank 8, or rank 25 and rank 26, etc) are separated by more than 0.7 tier. And the vast majority (33 sets) are separated from the set below them by 0.2 tiers or less.

S30 VERSUS PREVIOUS SEASONS

  • Overall, sets this season considerably outperformed those from earlier seasons. The average across all sets this season was 157.6, compared to 154.5 in season 28.

  • Much of this gain of 3.1 tiers in S30 as compared to S28 came to the weaker builds. S30’s strongest build (Rathma) is only +1.3 tiers vs S28’s strongest build (Tal). But S30’s rank 21 build, AoV, is +3.6 tiers vs S28’s rank 21 build, Zunimassa. And S30’s weakest build, Seeker, is +4.3 tiers vs S28’s weakest build, Roland.

  • The set with the largest lead of the rank 1 player over the rank 2 player is Masquerade (3.6 tiers). This is a considerable outlier. The next highest lead of a rank 1 player over the rank 2 is 2.0 tiers (No Set Wizard) and the average across all sets is just 0.6 tiers. So this Masquerade player has a lead 6 times higher than average! In fact, the total lead of all rank 1 players over their rank 2’s (i.e. you just add them up) is 25.5, meaning that this one Masquerade player accounts for 14% of this total amount all by himself.

PLAYER & REGION HIGHLIGHTS

  • 4 sets had multiple players tied at rank 1: Invoker, Monkey King, Inarius, and Typhon.

  • The class most dominated by a single player was Barb, which had 4 sets with 엔류 (Enryu) at the top.

  • DH, WD, and Wizard all had a different player claiming the crown for each set (i.e. six different players).

  • The single most dominant player overall was LebronJames, who took (or shared) the crown for 6 different builds (9 if you also count Kev, his hardcore account)

  • US and EU regions both had strong showings, with rank 1’s in 18 sets each. KR was significantly behind, with just 6.

OUR S30 PREDICTIONS

  • On average our pre-season predictions missed by about 2 tiers. We predicted Barb the best, missing by 1.5 tiers on average, while we predicted WD and Monk the worst, missing by 2.6 tiers.

  • Our biggest overestimate for a build was Jade Harvester, which underperformed our prediction by 6.6 tiers. Our biggest underestimate was Masquerade, which overperformed our prediction by 5.2 tiers.

  • We predicted 10 sets within 1.0 tier, 11 within 2.0 tiers, 12 within 3.0 tiers, and 9 at 3+ tiers.

  • Most sets had a fairly close relationship between their adjusted clear and their relative position on the basic (not adjusted) leaderboard. The average build ended up about 5 ranks higher or lower on the base leaderboard than it was on the adjusted board. 15 sets were within 2 ranks between both boards. 6 sets had a very large disparity, 10 or more ranks, with the largest going to Masquerade, which was 20 ranks lower on the base leaderboard (rank 28) than it was on the adjusted clear board (rank 8).

MORE INSIGHTS

In the Week 12 worksheet, we give recognition to the players behind the clears, we include a column showing the difference between their lead vs rank #2 – congratulations to the winners and those who contributed to raising the bar so high.

https://docs.google.com/spreadsheets/d/13Is-yaMsbIPVJKVkREuumXnhHD7GVpUHZlwYp2fWccs/edit#gid=1056583206

On the Current Ranking worksheet, we provide additional data showing the average peak clear by class, spreads, and how far off our predictions missed the S30 results.

https://docs.google.com/spreadsheets/d/13Is-yaMsbIPVJKVkREuumXnhHD7GVpUHZlwYp2fWccs/edit#gid=896614890

BARBARIAN

Set Clear Paragon Time Paragon Adj Time Adj Total Adj Adj Clear +/- vs last week
H90 150 3816 3:50 1.2 8.7 9.9 159.9 0.0
IK 150 3164 7:08 1.9 4.7 6.6 156.6 0.0
Wastes 150 2892 8:38 2.3 3.5 5.8 155.8 0.0
MOTE 150 2886 6:00 2.3 5.8 8.1 158.1 0.0
Raekor 150 3344 6:16 1.7 5.6 7.3 157.3 0.0
No Set 150 2991 5:07 2.2 6.9 9.1 159.1 0.0

CRUSADER

Set Clear Paragon Time Paragon Adj Time Adj Total Adj Adj Clear +/- vs last week
AoV 150 7851 3:14 -2.2 9.8 7.6 157.6 -0.7
Seeker 150 3144 11:00 2 2 4 154.0 0.0
Roland 150 3349 8:15 1.7 3.8 5.5 155.5 0.0
Invoker 150 2446 6:13 2.9 5.6 8.5 158.5 0.0
Akkhan 150 2031 6:43 3.5 5.1 8.6 158.6 0.0
No Set 150 8336 1:57 -2.5 13 10.5 160.5 2.3

DEMON HUNTER

Set Clear Paragon Time Paragon Adj Time Adj Total Adj Adj Clear +/- vs last week
GoD 150 3338 6:29 1.7 5.3 7 157.0 0.0
Natalya 150 4368 3:19 0.6 9.6 10.2 160.2 0.0
Unhallowed 150 3044 8:36 2.1 3.5 5.6 155.6 0.0
Shadow 150 3093 7:55 2 4.1 6.1 156.1 0.0
Marauder 150 2888 5:39 2.3 6.2 8.5 158.5 0.0
No Set 150 2853 11:15 2.3 1.8 4.1 154.1 0.0

MONK

Set Clear Paragon Time Paragon Adj Time Adj Total Adj Adj Clear +/- vs last week
PoJ 150 3512 3:46 1.5 8.8 10.3 160.3 0.0
Inna 150 1599 10:24 4.2 2.3 6.5 156.5 0.0
Uliana 150 3182 5:02 1.9 7 8.9 158.9 0.0
Monkey King 150 5132 3:56 -0.1 8.5 8.4 158.4 0.0
Raiment 150 4587 6:41 0.4 5.1 5.5 155.5 0.0
No Set 150 4593 3:47 0.4 8.8 9.2 159.2 0.0

NECROMANCER

Set Clear Paragon Time Paragon Adj Time Adj Total Adj Adj Clear +/- vs last week
Masquerade 150 1544 6:45 4.3 5.1 9.4 159.4 0.0
Pestilence 150 3474 7:05 1.6 4.8 6.4 156.4 0.0
Inarius 150 2931 5:27 2.2 6.4 8.6 158.6 0.0
Trag’oul 150 1356 8:42 4.7 3.5 8.2 158.2 0.0
Rathma 150 4423 1:55 0.5 13.1 13.6 163.6 0.0
No Set 150 3468 6:01 1.6 5.8 7.4 157.4 0.0

WITCH DOCTOR

Set Clear Paragon Time Paragon Adj Time Adj Total Adj Adj Clear +/- vs last week
Mundunugu 150 4028 6:05 1 5.7 6.7 156.7 0.0
Arachyr 150 2479 10:37 2.8 2.2 5 155.0 0.0
Zunimassa 150 8485 2:40 -2.6 11 8.4 158.4 0.3
Helltooth 150 2078 8:46 3.4 3.4 6.8 156.8 0.0
Jade Harvester 150 1748 14:12 4 0.3 4.3 154.3 0.0
No Set 150 6716 4:44 -1.4 7.3 5.9 155.9 -0.6

WIZARD

Set Clear Paragon Time Paragon Adj Time Adj Total Adj Adj Clear +/- vs last week
Typhon 150 6761 3:39 -1.4 9 7.6 157.6 0.1
Tal Rasha 150 1858 6:59 3.8 4.9 8.7 158.7 0.0
Delsere 150 1959 10:46 3.6 2.1 5.7 155.7 0.0
Vyr 150 8504 5:12 -2.6 6.7 4.1 154.1 -0.1
Firebird 150 4018 3:34 1 9.1 10.1 160.1 0.0
No Set 150 3518 3:21 1.5 9.5 11 161.0 2.0

Please note that all clears listed on this sheet are normalized at 5000 paragon. The adjusted clear system is explained here:

https://docs.google.com/document/d/1DVc3ph3LRRpYZ8WPSsJ62pv0Vj5Iw0YguVxiRmrcVh8/

13 Likes
2

Nice to see only a difference of 7 GR’s between the top adjusted Wiz no set 161.0 and the bottom adjusted clear Seeker 154.0. Looks like they hit decent balance relative to the balance issues of the past. Still a ways to go for reasonable balance, but looks like that isn’t going to happen.

Curious DMKT if you and rage are going to put something like this together for next season (maybe not on a weekly basis but a few times during the season. It would be interesting to see how much difference there is with balance losing the Shards and adding the open cube.

3

@Alexismad - there was a semblance of balance just because of SoulShards (mainly Ring of Fire that was “boosting” the weaker sets).

Most of the sets are still very undertuned and IMHO there is a big discrepancy in their performance even within classes.

1 Like
4

This is so cool, I love stats.

5

Exactly because they balanced the sets around Shards and Altar. This is why I am curious to see how out of whack the balance will be for S31 with free cube slots instead of the shards.

TBH, they needed to do a balance for each of the seasonal themes going forward, but they aren’t going to that. Balancing around the seasonal themes does throw non-season out of whack, but they have already shown they do not give a :poop: about that game mode.

1 Like
6

It was a nice map.

video

4 Likes