All the Minion Stats You Need to Know After Mean Streets of Gadgetzan

All the Minion Stats You Need to Know After Mean Streets of Gadgetzan

I’m a sucker for numbers, as stats can allow one to make a better decision in most avenues of life. Thus, when the remaining Mean Streets of Gadgetzan (MSG) cards were released today, I was able to update my master spreadsheet of cards. Given the ample amount of randomness in the game, I thought it would be a good idea to update various numbers on Hearthstone minions, given the new cards are officially part of the Standard rotation.


  • Learn about how the average cost, attack, and health of Hearthstone minions change when MSG is released.
  • Revisit the “evolution table,” which shows minion stat attack/health changes when moving up and down a mana cost. This is relevant given Evolve, Master of Evolution, and Devolve are standard cards.
  • Learn what the average drops are from portals and other special summons.
  • Look at other random stats.

Overall Minion Averages After MSG

Mode Cost Attack Health
Standard 4.19 3.5 4.04
MSG Std 4.17 3.47 4.02
Wild 4.14 3.47 4.05

The introduction of MSG only decreased the average minion cost, attack, and health by a small amount.

Set Cost Attack Health
Evergreen 3.99 3.28 3.85
Naxx 3.4 2.6 4.28
GvG 4.15 3.72 4.18
BRM 4.96 4.3 4.7
TGT 4.26 3.83 4.09
LoE 3.77 3.24 3.97
WotoG 4.64 3.77 4.43
Kara 3.82 2.94 3.59
MSG 4.08 3.31 3.89

Overall, the average minion set of MSG averages 4 and is close to a 3.5/4 statline. This isn’t hard to believe, given the number of anti-tempo minions around in this set with the Grimy Goons and Jade Lotus minions. With the dropout of Blackrock Mountain, Old Gods will represent the beefiest set of minions. MSG minions are overall on the small side.

Evolve/Devolve Table (Standard)

Org Cost Org Attack Org Health Trs Cost Trs Attack Trs Health ∆ Attack ∆ Health % Taunt % Charge
0 1.00 1.00 1 1.29 1.58 0.29 0.58 6.45% 1.61%
1 1.23 1.64 2 2.02 2.46 0.79 0.82 6.90% 1.15%
2 2.02 2.46 3 2.60 3.08 0.58 0.62 6.96% 1.74%
3 2.60 3.08 4 3.22 4.06 0.62 0.98 7.48% 1.87%
4 3.22 4.06 5 4.04 4.77 0.82 0.71 11.11% 3.70%
5 4.04 4.77 6 4.78 5.07 0.74 0.30 11.94% 2.99%
6 4.78 5.07 7 5.70 5.88 0.92 0.81 18.18% 0.00%
7 5.70 5.88 8 6.69 7.56 0.99 1.68 18.75% 12.50%
8 6.69 7.56 9 6.71 8.14 0.02 0.58 9.52% 9.52%
9 6.71 8.14 10 8.36 8.36 1.65 0.22 0.00% 0.00%
10 8.36 8.36 n/a n/a n/a n/a n/a n/a n/a

This table shows how the average statline of minions changes across mana cost. This is useful for determining stat changes from cards like Evolve, Devolve and Master of Evolution. The average attack and health numbers are across all Standard Play minions, with the 4 basic Shaman totems added to the 1-mana pool. The largest transformation benefit occurs from a 7-mana minion to an 8-mana minion. Taunt has a decent change of hitting from 6 to 7 and 7 to 8. The best chance to get Charge minions are from 7 to 8. The transformed minion stat values are also valuable for determining what minions you get out of various portals. For example, the average Firelands Portal summon will create a 4/5 minion on average.

Evolve/Devolve Table (Arena)

Org Cost Org Attack Org Health Trs Cost Trs Attack Trs Health ∆ Attack ∆ Health % Taunt % Charge
0 0.67 1.33 1 1.30 1.61 0.63 0.28 5.71% 1.43%
1 1.22 1.65 2 1.96 2.44 0.74 0.79 7.83% 1.74%
2 1.97 2.41 3 2.58 3.16 0.61 0.75 8.15% 2.22%
3 2.58 3.17 4 3.21 4.09 0.63 0.92 7.03% 2.34%
4 3.21 4.08 5 4.11 4.75 0.90 0.67 10.28% 2.80%
5 4.10 4.75 6 4.88 5.26 0.78 0.51 10.13% 2.53%
6 4.78 5.18 7 5.90 6.08 1.12 0.90 15.38% 0.00%
7 5.95 6.08 8 6.55 7.60 0.60 1.52 15.00% 10.00%
8 6.53 7.53 9 6.91 8.04 0.38 0.51 8.70% 8.70%
9 6.91 8.04 10 8.36 8.36 1.45 0.32 0.00% 0.00%
10 8.60 8.60 n/a n/a n/a n/a n/a n/a n/a

This table is the same as the table before this, but deals with the arena. The original minion includes the Wild set of minions, excluding all undraftable cards, like C’Thun cards and balanced out cards. The transformed minion includes any minion from the Wild set, as those cards can all appear in the arena through a summon.

Best/Worst Outcomes

Mana Cost Best Outcome Worst Outcome
1 Injured Kvaldir (2/4) Elven Archer (1/1)
2 Dirty Rat (2/6/taunt), Milhouse Manastorm (4/4) Novice Engineer (1/1)
3 Injured Blademaster (4/7) Blubber Baron (1/1)
4 A 4 mana 7/7 Faceless Shambler (1/1/taunt)
5 Earth Elemental (7/8/taunt), Leeroy Jenkins (6/2/charge), Doomguard (5/7/charge) Bomb Squad (2/2)
6 Savannah Highmane (6/5/DR), Sylvanas (5/5/DR), Illidan (7/5/effect) Big Time Racketeer (1/1)
7 Archmage Antonidas (5/7/effect), Bog Creeper (6/8/taunt) Acidmaw (4/2)
8 Tirion Fordring (6/6/taunt/effect), Ragnaros (8/8/dmg) Boogeymonster (6/7)
9 Ysera (4/12/effect) Blade of C’Thun (4/4), Majordomo (9/7/bad effect)
10 Deathwing (12/12) N’Zoth (5/7)

Here is a rough table of the best and worst outcomes. While there is a concern than evolving a 5 into a 6 or devolving a 7 to a 6 will make a 1/1 Big-Time Racketeer, it is also an equal chance for a Savannah Highmane or Sylvanas.

Average Demon

Cost Attack Health
Standard Demon 4.04 3.7 4.96
Wild Demon 4.19 3.78 4.93

Ever since the banishment of Mal’Ganis to Wild Play, demons have been mostly ignored in Hearthstone class design. MSG brings about some new demon support with an interesting card in Kabal Trafficker, which adds a random demon each turn. Krul the Unshackled is another big interesting demon in the new set.

Average Murloc

Cost Attack Health
Standard Murloc 2.33 1.94 2.06
Wild Murloc 2.48 2 2.29

4 new Murlocs are in this new set, and most interesting one is Finja. These minion averages only apply for the Murloc Knight Inspires, and overall the new Murlocs just further dilute the best Murloc summons of Murloc Knight and Murloc Warleader.

Average Beast

Class Cost Attack Health
Hunter 3.64 3.16 3.53
Druid 3.43 3.25 3.4
Warrior 3.37 3.09 3.4
Rogue 3.31 3.06 3.31
All Other Classes 3.38 3.09 3.38

Tomb Spider is one of the top 3 arena commons picked on average, so I took a look at what the average beast stats are across classes. Here, you’re looking for a useful effect over raw stats, but interesting that the Hunter beasts are actually heavier than Druid ones.

Average Legendary

Cost Attack Health
Pre-MSG Std 6.43 5.21 5.83
MSG Std 6.2 4.95 5.55

Finally, this table shows the average stats for legendary minions in the standard pool before MSG and after MSG. As you can see, the average legendary minion is cheaper and has a bit less attack and health after the release. This confirms my eye test of noticing smaller legendaries made in this set, and the changing Team 5 philosophy of making less impactful neutral legendaries.


Elements of Hearthstone RNG: Pings

I made a pact... to clear big taunts for lethal
I made a pact… to clear big taunts for lethal
The most common type of RNG in Hearthstone comes in the form of a card having an ability that will cause damage or destroy other characters, without actually attacking on it’s own. Unlike a Mage Fireblast, these “pings” go random places and cannot be targeted.
  • Battlecries and other one-time effects have less RNG than continuous effects or triggered effects, by virtue of the effect happening once.
  • Abilities that hit minions only have less RNG than those that can hit face.
One time effect, limited minions affected
Stampeding Kodo – The kodo is the least random card in the ping/attack class. It has a one-time effect to kill something that has 0, 1, or 2 attack. The effect is most effective (and has zero randomness) when there is only 1 opposing minion that has 0-2 attack on the board. It is clearly ineffective when the opponent has stacked 7 small creatures on the board (14% for each minion). Against 3+ attack minions, the kodo is just a bad tempo 3/5.
One time effect, one random minion
Bomb Lobber/Flamecannon – Bomb Lobber and Flamecannon have the same exact effect of dealing 4 damage to a random minion. Like the Kodo, it works best when there is one minion on the board with 4 or less health.
Deadly Shot/Sabotage – Same effect, except with a much bigger swing, as it completely destroys a minion. These random hard removals work best against one big thing on the board. The 4 cards above have probabilities to hit minions from 100% to 14%.
 Void Crusher – The Void Crusher is unique and doesn’t technically have a one-time effect. However, the potential to destroy itself upon inspiring really limits how long it will stay on the board. Also the amount of RNG is potentially doubled depending on how many minions are on each side. A board with 2 minions on each side will result in 4 possible outcomes, for a 25% chance. A board with 7 minions on each side will result in 49 possible outcomes, for a 2% chance. The chance of destroying a particular minion on each side remains at 14%.
One time effect, 2 minions hit
Forked Lightning/Multi-Shot/Dark Bargain – Like above, the fewer minions on the opposing side, the better the chances, however there is a minimum of 2 minions. Using these spells on a stacked board of 7 minions can result in 27 possible outcomes, for a small 3.7% chance. However, the chance of hitting a particular minion on a stacked side is about 28.5%.
One time effect, one character hit
Flame Juggler /Fist of Jaraxxus – The Flame Juggler is considered one of the better 2-drops in the arena, as it has the chance to snipe an annoying 1 health minion, like Worgen Infiltrator. Well, unless it hits the face. Any pings that open up heroes to attack are automatically more random than those that just target minions, as an extra target is provided. Though these pings are more dangerous (closer to lethal), their reliability of clearing a minion (what you typically want) is reduced by a big margin.
One time effect, 2 characters hit
Misdirection – Misdirection is unique, though it fits the one time effect being a secret, but causes 2 characters to run into each other. The RNG hurts the side with more minions, causing the possibility of 2 minions killing each other off. The only way this card has 0 RNG is if your opponent has 1 minion on the board while you don’t have any. This will cause a 100% probability of your opponent hero getting hit with their minion. Same goes will opponent hero attacking their own minion instead.
Two time effect, one character hit
Dr. Boom – Dr. Boom is unique in that he has Shaman RNG (damage range) and is the only card that will specifically allow for 2 independent instances of one character being hit. Each Boom Bot that is destroyed will have a chance to hit any one character on your opponent’s side.
Turn-based effect, one character hit
Ragnaros the Firelord – Pretty much the same probabilities as the above class, except the ping occurring each turn will help increase the number of instances occurring. The concept of flooding the board to prevent Rag from getting a lethal shot demonstrates the basic understanding of probability. Flood the board and he has a 12.5% chance to go face and end things.
Demolisher – These two cards have the same effect, except that the Demolisher has slightly more RNG than Ragnaros. Because Ragnaros will hit after your turn, it means you can actively help reduce the number of minions on the board, in order to increase the probability of something getting hit. Demolisher hits at the start of your turn, which means, your opponent has a chance to set up a board to reduce the probability of something getting hit.
One time effect, multiple missiles
Arcane Missiles/Goblin Blastmage/Avenging Wrath – With more pings comes more RNG. The health of opposing minions matters more than ever with these. If a board with all 1-health minions is hit by Avenging Wrath, each dead minion will increase  the probability of the other minions are getting hit. A minion that is hit and dead will not get overkilled, and thus increase the probability of another getting hit. Another example, let’s say it’s turn 4, you have a Target Dummy and your opponent summons an Evil Heckler (5/4). You drop Blastmage. Because each ping has a 50% chance to hit the Heckler, the chance of the Blastmage killing the Heckler is a little over 6%. Let’s say same scenario, but instead of an Evil Heckler, your opponent plays Mini-Mage (4/1). The Blastmage still has 50% chance of hitting the minion, but there is an almost 94% chance the Mini-Mage will get killed.
Triggered effect, one character hit
Ship’s Cannon/Eydis Darkbane – Triggered ping minions are interesting because they can provide a good deal of RNG or none at all. Their effects are typically more useful in decks built around them, meaning they can do more damage in constructed. Ship’s Cannon and Eydis Darkbane can be picked in the arena, and due to the lack of synergy or an incompatible class, never have their effects activated at all. The Knife Juggler will always have it’s effect activated (as you’ll always have minions), while the Shadowboxer will always have it’s effect activated (so long as a Priest can spend mana to heal.) These cards have a little more RNG than the one-time effect minions, as there are more chances for their effects to hit. Ship’s Cannon and Eydis Darkbane cannot throw multiple shots in one turn, and instead have an independent event for each trigger.
Knife Juggler/Shadowboxer – While these 2 minions have the weaker ping effects, they are hybrids of one character hit and multiple missiles types. Knife Juggler will throw multiple knives in one instance if 2 minions are summoned at once, or if a card like Unleash the Hounds is played. Shadowboxer will throw multiple punches in one instance if Holy Nova is used, and multiple characters are healed. As described above, the health of opposing minions matters more in affecting the randomness for multiple missile cards.
Triggered effect, multiple missiles
Flamewaker – The Flamewaker is unique on the RNG list in that it is the only true trigger multiple ping. Cast a spell, and 2 pings will be sent at opposing characters. While the RNG of each trigger is lower than a Goblin Blastmage, the ability to cast multiple instances of pings will help increase the RNG.
Turn-based effect, one minion hit, either side, most minions effected
Fel Cannon – Fel Cannon finally brings about the concept of friendly fire. This essentially doubles the amount of RNG, as the amount of minions is potentially doubled. On a board state with 7 minions on either side, Fel Cannon has a 7.7% chance to ping any particular minion. Because Fel Cannon can’t hit itself, the probability of your minion getting hit (46%) is lower than that of your opponent’s getting hit (54%). The rub is that Fel Cannon will not hit any mechs, so if your opponent plays mechs, the Fel Cannon can backstab your own board.
Warrior RNG
Brawl – Oh hey, the Warrior is a roulette player sometimes. Brawl RNG is pretty simple, as the side that has more minions is more likely to have a minion live. Any opposing boards with the same number of minions on each side will be a 50% coin flip for who wins the brawl. Brawl is funny, as The Warrior typically doesn’t care if he wins the Brawl. Brawl is typically played as a board clear (who cares who dies, as long as there is one minion left) or to kill one big threat.
Bouncing Blade – The same concept with Brawl, except the cumulative health points of minions will help determine the probability here. A side with 2 Yetis (4/5) has a 67% chance to win a Bouncing Blade foray over a side with 1 Yeti. While cumulative health totals are important here, the appearance of a low-health minion is even more important, as Bouncing Blade ends once a minion dies. So, the side with a one health minion is a the greatest disadvantage of that minion dying. There is no real way to calculate the probability of this card, as every situation is different. Much situational RNG.
One-time effect, multiple missiles, any character hit
Mad Bomber/Madder Bomber – These goblins are used in promotional work for Hearthstone and happen to be the most RNG cards in the game. Sticks of dynamite are thrown haphazardly 3 or 6 times, at any character on the board. While they are typically used to clear a board, enraging your own minions can serve some utility as well. Let’s go through a specific case, ADWCTA’s Gahz’rilla Dream.
I do not own this. It was stolen off ADWCTA's YouTube. Source:
I do not own this. It was stolen off ADWCTA’s YouTube. Source:

The premise here is that Gahz’rilla needed to be hit by 2 out of 6 Madder Bomber pings to grow to 24 attack and win the game. There are 8 characters on the board (Uther, Mad Bomber, 2 Silver Hands, Faerie Dragon, Youthful Brewmaster, Gazh’rilla, Rexxar).

Bomb 1 – Rexxar
Bomb 2 – Gahz’rilla
Bomb 3 – Mad Bomber
Bomb 4 – Faerie Dragon
Bomb 5 – Rexxar
Bomb 6 – Gahz’rilla
Gahz’rilla got hit 2 times. What were the chances of that happening?

(Disclaimer: This math could be very wrong)

8 characters, 6 flips of dynamite

8^6 = 262144 possible independent events

2 Gahz = GahzA, GahzB

_ _ _ _ _ _

GahzA = 6 spots; GahzB = 5 spots

_ _

Gahz A = 2 spots; GahzB = 1 spot

# of possible outcomes = (6 x 5) / (2 x 1) = 15 possible outcomes where Gahz is hit exactly 2 times

Probability = 15 / 262144 = 0.0000572%

The math is not completely correct, as if one of the opponent minions died, the chances of hitting Gahz increase for each minion death.

As seen the bombers make for extreme RNG, as there are so many different combinations/permutations to consider. But they also make for some great moments now and then.

  • This is the most common form of RNG in Hearthstone as of now.
  • Ping RNG is typically countered by playing more minions on the board to reduce the chances of particular minions getting hit.
  • Conversely, to make the RNG work in your favor, reduce the number of minions on the board to have a better chance of hitting the desired target.
  • Health of minions matters more for some types of this RNG.
  • Minions with trigger pings or turn-based pings allow for more events to occur.
  • Don’t try to do the math for Madder Bomber probability. It literally took me a whole day, and I’m not sure I got it right.