Hour of day

How much the win rate of a hero changes at different times of the day, compared with his win rate over the whole day (ranked All Pick, STRATZ, the games of all the world together). The hours are UTC: the game has no single time zone, so "morning" is morning in Greenwich, evening in the east of Asia.

#HeroWR at this timeWR overallGames
1Skywrath Mage
+4.9%
51.4%46.5%354
2Winter Wyvern
+3.1%
57.5%54.4%861
3Mirana
+2.9%
48.6%45.7%751
4Leshrac
+2.5%
52.2%49.6%918
5Juggernaut
+2.4%
48.5%46.0%712
6Kez
+2.2%
49.1%46.9%715
7Pugna
+2.2%
46.9%44.7%386
8Drow Ranger
+2.2%
45.6%43.5%436
9Void Spirit
+1.9%
51.5%49.6%1.1K
10Rubick
+1.9%
48.4%46.4%1.2K
11Bounty Hunter
+1.9%
45.1%43.2%528
12Lone Druid
+1.7%
46.9%45.2%974
13Sniper
+1.6%
47.2%45.6%2.2K
14Phantom Lancer
+1.5%
49.2%47.7%504
15Broodmother
+1.3%
49.3%48.0%1.5K
16Anti-Mage
+1.3%
49.8%48.5%1.1K
17Elder Titan
+1.3%
49.8%48.5%2.2K
18Alchemist
+1.1%
45.3%44.2%2.3K
19Enchantress
+1.1%
50.9%49.8%788
20Zeus
+1.0%
45.5%44.5%650
21Abaddon
+0.9%
50.6%49.7%3K
22Huskar
+0.9%
44.6%43.7%3.7K
23Lion
+0.8%
45.6%44.8%1.1K
24Nyx Assassin
+0.8%
51.4%50.6%2.5K
25Centaur Warrunner
+0.7%
50.0%49.2%17.2K
26Visage
+0.7%
54.7%54.0%5K
27Vengeful Spirit
+0.7%
52.5%51.8%3.1K
28Bloodseeker
+0.7%
51.6%50.9%1.3K
29Beastmaster
+0.6%
46.3%45.7%7.4K
30Monkey King
+0.5%
45.3%44.8%1.3K
31Venomancer
+0.4%
44.2%43.8%792
32Spirit Breaker
+0.4%
50.6%50.2%13.5K
33Tiny
+0.4%
43.5%43.1%5.1K
34Tidehunter
+0.3%
49.7%49.4%44K
35Wraith King
+0.3%
55.3%55.0%35.5K
36Ursa
+0.3%
46.4%46.1%1.9K
37Lina
+0.3%
43.9%43.6%597
38Sand King
+0.2%
49.5%49.2%9.7K
39Dragon Knight
+0.2%
52.5%52.3%42.4K
40Phoenix
+0.2%
51.5%51.3%7K
41Clockwerk
+0.2%
47.0%46.8%1.6K
42Primal Beast
+0.2%
49.4%49.2%14.1K
43Legion Commander
+0.2%
53.2%53.0%63.6K
44Windranger
+0.2%
50.0%49.8%8K
45Bristleback
+0.2%
48.8%48.6%22.6K
46Outworld Destroyer
+0.1%
47.9%47.8%1K
47Lycan
+0.1%
49.7%49.6%6.2K
48Axe
+0.1%
50.6%50.5%69.2K
49Witch Doctor
+0.0%
44.6%44.5%397
50Omniknight
−0.0%
50.7%50.7%4.1K
51Slardar
−0.0%
49.0%49.0%21.4K
52Dawnbreaker
−0.1%
51.9%52.0%43.3K
53Undying
−0.1%
47.8%47.9%10.8K
54Necrophos
−0.1%
50.6%50.7%22.5K
55Lifestealer
−0.1%
46.1%46.2%3.5K
56Slark
−0.2%
51.6%51.8%2.4K
57Treant Protector
−0.2%
40.7%40.9%469
58Faceless Void
−0.2%
46.5%46.7%809
59Dark Seer
−0.3%
49.8%50.1%16.2K
60Night Stalker
−0.3%
51.7%51.9%36.2K
61Pudge
−0.3%
52.3%52.6%29.2K
62Shadow Fiend
−0.3%
40.8%41.1%644
63Doom
−0.3%
47.6%47.9%21.1K
64Batrider
−0.3%
42.3%42.6%887
65Earthshaker
−0.4%
48.9%49.2%27.2K
66Razor
−0.4%
50.6%51.0%10.7K
67Weaver
−0.4%
47.3%47.7%1.5K
68Kunkka
−0.4%
49.0%49.4%11.6K
69Pangolier
−0.4%
44.3%44.7%4.6K
70Ogre Magi
−0.4%
48.7%49.1%15.5K
71Clinkz
−0.4%
45.8%46.3%537
72Enigma
−0.5%
53.3%53.8%12.2K
73Mars
−0.5%
46.7%47.2%16.2K
74Riki
−0.5%
53.4%53.9%1.5K
75Techies
−0.5%
46.2%46.7%1.2K
76Underlord
−0.6%
47.2%47.8%13.3K
77Magnus
−0.6%
48.4%49.0%27.1K
78Viper
−0.7%
47.1%47.8%9.4K
79Timbersaw
−0.7%
44.8%45.5%23.8K
80Tusk
−0.7%
46.0%46.7%2.1K
81Medusa
−0.7%
44.0%44.8%343
82Largo
−0.7%
46.6%47.3%3.5K
83Jakiro
−0.8%
43.3%44.1%390
84Gyrocopter
−0.8%
43.9%44.7%483
85Death Prophet
−0.9%
47.2%48.1%7.7K
86Io
−0.9%
48.9%49.7%804
87Marci
−0.9%
47.6%48.5%3.8K
88Brewmaster
−0.9%
50.2%51.0%11.2K
89Invoker
−0.9%
45.9%46.8%732
90Chaos Knight
−1.0%
49.8%50.8%11.8K
91Snapfire
−1.4%
46.2%47.6%2.9K
92Keeper of the Light
−1.5%
43.5%45.0%545
93Luna
−1.6%
43.7%45.3%442
94Nature's Prophet
−1.6%
39.8%41.4%1.9K
95Sven
−1.7%
45.3%47.0%2K
96Ember Spirit
−1.7%
44.4%46.1%468
97Earth Spirit
−1.8%
47.2%49.0%3.1K
98Phantom Assassin
−2.0%
44.5%46.6%743
99Silencer
−2.4%
45.6%48.0%608
100Spectre
−2.9%
48.3%51.2%789
101Troll Warlord
−3.6%
46.5%50.1%602
102Hoodwink
−3.8%
37.8%41.6%468
103Queen of Pain
−3.9%
42.7%46.5%898

WR delta = the hero's win rate in these hours minus his win rate over all hours. STRATZ gives the games by the hour only for the last day and a half, so we keep the hours ourselves and the sample grows day by day: 21.5 days of hours are collected so far (the target is 30). Heroes with fewer than 300 games in these hours are left out. The differences are small (a point or two), so until the sample grows they are only a tendency.