'Marvel Puzzle Quest' User Statistics (Heroes & Abilities)

Spoit

atomzed wrote:

Phantron wrote:

atomzed wrote:

Phantron wrote:

I'm surprised Star Spangled Avenger is that high for 3*.

It makes sense. Because once you get it going, you can cast it again in the same game.

Lets say 10 people use SSA one game each. That would mean that you probably get 25 times of SSA (asuming 2.5x per game) in the frequency count.

The same 10 people use CoTS in one game each. That would mean only 10times of COTS in the frequency bc y tinykitty

That'd mean no character is used roughly 2-3X as often as Captain America, which seems surprising to me. I wonder how the PvP/PvE breakdown goes. I can definitely see him being #1 in usage in PvE by the defined metrics but PvP would be quite surprising.

Hmm... when I saw usage percentage I was thinking along the line of

Ssa % = Total ability usage count / total number of Steve Rogers games.

Such a percentage count will mean that the total number of games are accounted for EACH character separately.

But you have a point, if they count the percentage as
Ssa % = Total ability usage count / total number of games of ALL characters

Then your question makes perfect sense.

Icex, could we have a response to this question? How are the stats calculated?

That'd mean that patch's healing is only triggered in an infinitesimally small number of games where he's used, which I don't buy

Unknown

Dartmaster01 wrote:

This just confirms most of the player base is in one and two star land. Also, possibly explains some of their decision making.

Of course. And this is pretty sad.

It reminds me of Hasbro's policies with Magic the Gathering : they targetted the "Walmart" consumer and not the Pro Tour players. They also don't give a damn about collectors.

simonsez

Phantron wrote:

If it's just out of that character's games that'd mean World Rupture is used one out of roughly 200 games that involves Sentry

It's not possible they're computing it that way. Just look at the top percents. There's no way in hell that Thor's red gets used only 1 in 30 matches that include him. And if they did only consider matches that included that character, you'd see Prehistoric Arms at 100%

Phumade

These stats are great and I really appreciate being able to consider the implications. For the forum readers, the sorting by power level is very help/insightful and it would be useful to know what powers the AI triggers most often in their respective power levels (specifically, 2*/3*/4*). Some very surprising results that call for some interpretation!

Top 3* abilities
Hulk (Indestructable): Thunderous Clap - 5.19%
CaptainAmerica (Rogers): Star Spangled Avenger - 4.14%
Hulk (Indestructable): Anger - 3.65%
Wolverine (Patch): Healing Factor - 3.55%

I have a hard time believing that Hulk/Spidy powers are among the 4 most activated 3* powers. For a couple of reasons: 1. I thought spiderman was a fairly rare 3* char. I've been playing for 7 seasons and his covers are among the most infrequently dropped/awarded. So its surprising for me to Spider sense as a commonly used power. Same thing for for Hulk's anger vs Thunder clap! In every hulk match I've participated in Anger fires twice as often as T-Clap! Maybe they consolidated all the anger CD tiles into 1 instance per turn? i.e. whenever I fire World rupture at least 8 Anger CD tiles appear! Does that mean Anger fired 1 time or 8 times? (This can be the confusing part of statistics). In addition, I felt that Hulk is a fairly rare/infrequent character. I have a somewhat hard time believing that more people use Hulk vs Patch/Daken/Thor. But that could be my own biases! LOL

Bottom 3* abilities
Colossus (Classic): Colossal Punch - 0.19%
Beast (Classic): Medical Marvel - 0.18%
Loki (Dark Reign): Trickery - 0.14%
Colossus (Classic): Immovable Object - 0.14%
Colossus (Classic): Fastball Special - 0.08%

These aren't all that surprising. To me these stats imply that Colossus/Beast are uncommon characters that aren't played often. The more interesting result is Loki. His covers are common and every 3* player has had plenty of opportunities to cover max him in 7 seasons. This implies that trickery is too expensive to fire (which it is) or that people aren't effectively teaming him up with patches or using him defensively vs Daken. Hypothetically, if you reduced trickery to 2AP would he now become the hard counter to Daken/Blade? Interesting question to consider.

Top 4* abilities
Wolverine (X-Force): X-Force - 24.29%
Wolverine (X-Force): Surgical Strike - 21.17%
Wolverine (X-Force): Recovery - 12.79%
Thor (Goddess): Smite - 10.85%
Thor (Goddess): Power Surge - 8.92%

Somewhat boring stats to consider in light of the fact that there are only 5 4* characters of which 2 are virtually unsuable (IW is badly underpowered, Dino is a trophy) That said the appearance of Thor really just means people played the hell out of the LR to get double ISO and or Nick Fury is a rarely used character. I do find it amusing that Dino Disco dance doesn't show up on stats. In everymatch I participate, Dino purple fires every other round! lol

Top 5 Passives
Black Widow (Original): Espionage - 17.1%
Daken (DA Wolverine): Pheromone Rage - 12.34%
Daken (DA Wolverine): Healing & Heat - 10.9%
Storm (Classic): Raging Tempest - 7.77%
Spider-Man (Classic): Spider Sense - 7.35%

Bottom 5 Passives
Falcon (Mighty Avengers): Inspiration - 2.32%
Doctor Octopus (Octavius): Monologue - 1.77%
Falcon (Mighty Avengers): Redwing - 1.7%
Deadpool (It's Me!): Life of the Party - 0.39%
Captain Marvel (Modern): Energy Absorption - 0.32%

The passives are very interesting to consider because of a few factors. All things being equal green passives should trigger as often as purple as often as black. I do agree that some colors will be more popular than others if only because players are actively trying to collect as color to trigger another power. I normally wouldn't expect to see as many purple matches as the stats indicate. I thing that really stands out is spider sense. popular passives should belong to the 2* characters. There are significantly more 2* teams than 3* and within 3* chars, spider sense should be far behind LDaken Healing/heat or pheromone rage! This seems counter intuitive and perhaps is a result of how Ice is collecting/consolidating stats? I would have assumed that there are way more Falcon covers in the wild than Spiderman covers and or Falcon is used at a far higher rate than Spiderman!

OH well, this was all speculative interpreation and I really do enjoy seeing these types of stats. Good Work Dave and Ice!!

Phumade

In light of of this new found sense of openness and full disclosure. Would you consider disclosing how MMR is calculated and implemented? I think that is the one algorithm, that every mid tier player really wants to know/understand.

I.E. How much do I really have to tank. Does it need to be active suicides / retreats. How many matches, before I need to tank again. How much/little effect does tanking in simulator affect pvp tanking.

And while I'm thinking about it. How exactly does the scaling algorithm work? What is the breakdown between personal/community scaling. How much does my personal roster affect scaling. Am I better off hammering the nodes early in PVE in the hopes that I can scale out the other players before the rubber band kicks in?

Might as well include rubberbanding in that that discussion as well!

Thanks for the improved communication!

IGN Kung Phu Panda

GuntherBlobel

It would have been nice to see the usage for every power. I think we'd be surprised at how infrequently a bunch of really awesome powers are used.

The WR usage doesn't surprise me. I almost never use it besides for that one time I went for 1300 pts. I used it exactly 3 times every 3 hours and then stopped playing until the end of the event. Isn't that how it's generally used?

Moral

Cover availability is bound to factor in. And Steve was a prize quite often.

Consider also the number of matches you play for a given event:

For a routine PvP, 36 matches to hit 800, then hopefully another 12 matches to 1200+ for a 2.5 day event .

PvE, most subs seem to have about 10 repeatable nodes. Players looking to place well are likely aiming for 40 matches a day.

I'm really thinking PvE sees more matches played overall by a significant amount.

Moon Roach

Obviously the game phones home, otherwise this information wouldn't be available. I wonder what level of detail is retained.

At the minimum we must have the basic "World Rupture fired at 16:30 UTC November 1 2014". At the other extreme we could have something like "lvl 166 5 cover World Rupture fired at 16:30 UTC November 1 2014 by Fred64666 at IP address 127.0.0.1 (oh look, it's Vatican City) in match against Wilma32333 in Quasar tournament followed by lvl 166 5 cover Sacrifice and lvl 166 5 cover Intimidation. Game over, victory to Fred64666 gaining {checks against server to find how many, oh only 9} points". Though possibly there could be much more sent.

Could someone at D3P comment on how much the game does send back?

:edit: extended the paranoia somewhat.

esoxnepa

mags1587 wrote:

Yelena is in the top 5 team ups used by players? Really? Would not have called that.

Likely, because it is available to them. All through prologue and many of the PvE missions feature her.

Unknown

Phumade wrote:

Same thing for for Hulk's anger vs Thunder clap! In every hulk match I've participated in Anger fires twice as often as T-Clap! Maybe they consolidated all the anger CD tiles into 1 instance per turn? i.e. whenever I fire World rupture at least 8 Anger CD tiles appear! Does that mean Anger fired 1 time or 8 times? (This can be the confusing part of statistics). In addition, I felt that Hulk is a fairly rare/infrequent character. I have a somewhat hard time believing that more people use Hulk vs Patch/Daken/Thor. But that could be my own biases! LOL

A Hulk green cover has been awarded a few times as a node award from the Hulk PVE event. If you played those at all, it was super easy to get the Hulk green. It's likely that there's lots of people who have a few Hulk green covers and no Hulk black covers. For those folks, Hulk may also be one of the few 3* characters in their roster as well. I started playing early on, and Hulk was my first maxed out 3*. So his covers used to be more available. I've actually sold a few of his green covers.

Unknown

sarlok5 wrote:

Phumade wrote:

Same thing for for Hulk's anger vs Thunder clap! In every hulk match I've participated in Anger fires twice as often as T-Clap! Maybe they consolidated all the anger CD tiles into 1 instance per turn? i.e. whenever I fire World rupture at least 8 Anger CD tiles appear! Does that mean Anger fired 1 time or 8 times? (This can be the confusing part of statistics). In addition, I felt that Hulk is a fairly rare/infrequent character. I have a somewhat hard time believing that more people use Hulk vs Patch/Daken/Thor. But that could be my own biases! LOL

A Hulk green cover has been awarded a few times as a node award from the Hulk PVE event. If you played those at all, it was super easy to get the Hulk green. It's likely that there's lots of people who have a few Hulk green covers and no Hulk black covers. For those folks, Hulk may also be one of the few 3* characters in their roster as well. I started playing early on, and Hulk was my first maxed out 3*. So his covers used to be more available. I've actually sold a few of his green covers.

The loaner Hulk has Thunderous Clap but definitely not Anger or Smash.

Unknown

locked wrote:

Bottom 5 Characters by activated ability

Captain Marvel - 0.2%

That's right, Captain Marvel never procs! Lower the threshold on yellow

Nice stats, especially by rarity tiers, and thanks.

And fix dp black. It needs a boost, lower threshold, and regeneration each turn (for offense).

Unknown

And fix Sentry. Combine World Rupture and Sacrifice. Also lower the countdown to 1.

Unknown

I was just reminded of my statistics professor back in college. Who's laughing now you Excel freak?

Unknown

I love statistics. They're so much fun to use and abuse. Could we see how these compare to just abilities from players in the top 100 for Versus events (for simplicity, how about just the most recent Cap Marvel event). Those are the people earning the most 3* covers, and I suspect the picture would be very different.

Unknown

simonsez wrote:

Phantron wrote:

If it's just out of that character's games that'd mean World Rupture is used one out of roughly 200 games that involves Sentry

It's not possible they're computing it that way. Just look at the top percents. There's no way in hell that Thor's red gets used only 1 in 30 matches that include him. And if they did only consider matches that included that character, you'd see Prehistoric Arms at 100%

They stated in the original post they're computing all abilities used total. So if abilities were activated a billion times, Sentry's world rupture would have been activated about 9 million times. It's believable considering few players have Sentry, and more tend to have 2* Thor.

Overall, the stats just seem to confirm what we already know. People don't use bagman, Sentinel of Liberty is too expensive, Colossus and Beast are new character's, but somehow Beast might be more popular than Colossus? (probably the healing). Giving out Daken's Pheromone rage as a free cover on day 30 means that people will use him even without true healing. If you give people a free Green Hulk cover, they'll use it, especially if the rest of their roster is underleveled 1* and 2*. If you fight Yelena in PVE a lot, you'll get a lot of her team-ups.

I'm guessing:
1. Fake healing is more popular in the mainstream than on forums.
2. Low level players like playing hulk PVE because Hulk's high HP gives them a chance to win more games.

Chrono_Tata

I don't think Beast is more "popular" than Colossus. A Beast event was run relatively more recently than a Colossus event. Assuming the data was sampled over the last month or so, that explains why he has been used more.

Unknown

IceIX,

I appreciate the stats on usage, but I think a more illuminating cut of the data would be to show percentage of damage is associated with each power. I believe it would be more representative of the strength of each power. Hopefully you have that information and can provide.

Thanks.

Unknown

So many statistics I would love to see. One would be a breakdown of how commonly owned each character is, preferably with data regarding #of covers.

EG. -Sentry with completely made up numbers
13 0.4%
10-12 0.9%
7-9 1.5%
4-6 2.3
1-3 4.9
0 90%

Then to compare that to characters played and we can see who gets used adjusted for character rarity.