The Best Heroes-A Statistical Community Effort

Other Games SotM Game Engine
81

Wow. With some deep research there is a lot that excell sheets can do!

 

If you go take a look at the 'raw data' of the form youll see a few additional sheets - live time updating! (sorta, mostly, up to a certain point then it seems i have to manually update the ranges it pulls data from)

 

https://docs.google.com/spreadsheet/ccc?key=0AqClP82b_XgZdFVreEhkOWJYcmlvcDNMbWppejh5aGc#gid=0 -- to take a look - see the new sheets on the bottom? 

 

 

Anyways, villain is done, still have to do Hero - but im going to have to go through all 187 games and add the incap field in manually, since i changed how it was done on the form - and its a good thing I did, cause it makes the method ive learned to gather the data much easier.

 

once all this is done, next month ill be able to just manually enter the data into the 'official statistics' one (the one that looks pretty with colors) and then can sort by greatest to least, makint it -much- clearer, and get some awesome looking graphs in there as well :)

82

another post on the update

 

I only have a few more forumals to work out (man all these paranthesis are driving me INSANE!) and then all the data can be pulled automatically from the sheet ...

 

and in the process ive already noticed some issues - like some how i missed a game here, or there, or input things wrong :P With 175ish entries its no surprise i made some mistakes! When I finish these formulas I may 'rerelease' the january update, in order to make things perfect :)

 

 

 

(if any of you are curious, here is the formulat to determine if a hero was in a 5 player game)

 

=if($A$1="is_calculation_on",ArrayFormula(SUMPRODUCT( ((first_hero=$A3)+(second_hero=$A3)+(third_hero=$A3)+(fourth_hero=$A3)+(fifth_hero=$A3))-( (first_hero=$A3)+(second_hero=$A3)+(third_hero=$A3)+(fourth_hero=$A3))*(fifth_hero="(None)") ) ),"off")

 

 

there - might- be a more effecient way to do that? but... im not a mathmatician, statician, or engineer :p

83

This formula seems to add up all of the times a hero was used and then remove the times in which that hero was used in a 4-hero game. What about 3-hero (or even 2-hero) games?

Suggestion: Use some intermediate cells. One possibility is to first calculate how many times a hero was used in a two-hero game, than use that result to help calculate how many times a hero was used in a three-hero game, then use that result to help calculate how many times a hero was used in a four-hero game, etc. Another possibility is for each game to calculate a cell that marks the game as to how many heroes were in it, and then calculate how many heroes were in a five-hero game by only adding up games that are already marked as five-hero games.

 

What would REALLY help you is if you could use functions. Apologies if this is something you know already, but what a function does is wrap up a set of calculations that you want to use more than once so you don't have to write it more than once. There are lots of heroes, so you are probably copying and pasting the formula you use to calculate the number of times a hero was used in a five hero game for each hero. If you could write the formula once and put in into the function, than you could just call that function for each hero, so if you ever needed to fix/change/upgrade the formula, you could just fix/change/upgrade the function. Welcome to the world of programming! This can be done in Microsoft Excel using Visual Basic for Applications.

84

im not sure if Gdocs Sheets can handle functions - as once a budding computer scientist (before i flunked physics 3 times and decided to  be a chef :P) I did indeed try to see if i could use functions, but was unable to find the process for doing so. Because for some reason attempting to download the gdoc corrupts the file every time, ive kept everything in gdoc for now Yet to find a way around this (it might be because im using gdoc regions, which are namesets in excell and may not translate properly. Or because im still using 2010 excell?) -- If you know how to use functions in Gdocs Sheets ill gladly take the advice for ease of updating later.

 

and i am not certain why it does not -need- the - 3 hero games (i remove the 2 hero games from the data sheet manually)  -- when I put in the - 3 hero game section it ends up giving me like 30 games. Which is far from correct. I /believe/ it is because of the way the data is laid out -- 

hero name, incap, heroname, incap, heroname, incap, heroname, incap, heroname, incap -- all in seperate cells-

 

so a 3 hero game looks like

 

heroname, incap, heroname, incap, heroname, incap, none, incap, none, incap

 

and a four looks like

 

heroname, incap, heroname, incap, heroname, incap, heroname, incap, none, incap

 

so in calculating the pair, and removing the 4 hero games from a the pool of 5 player games, it is automatically removing the 3 hero games as well.... that -shouldn-t be what its doing but everything ive double check and manually checked says its calculating the correct number. I don't know enough to tell you why

(im still trying to figure out exactly how ArrayFormula and SUMPRODUCT work, actually - why does finding the number of times a villain used give the same value with both -  =ArrayFormula(SUMPRODUCT((villains=$A3))) and =ArrayFormula(SUMPRODUCT((villains=$A3)*(villains=$A3))) --- shouldnt that expodentially give a value thats all wrong for the second formula? But it doesn't - they give the same value. Why?

it has been a long time since my programming days (10 years about) and trying to keep all the () straight in a single line is killer. Why can't excell (gdoc sheets in this case :P) have a multi line editor? (and if it does and you know where it is pleeeassee tell me?)

 

85

double post cause i think my post got too long and it stopped putting carriage returns in  (oor because i copied directly out of gdoc sheets it left some residue cell information somwhere in there and wouldnt process a return - silly things)

 

 

for instance you can see here(maybe)

 

 

 
Absolute Zero 51 35 68.63% 18 2 35.29% 11 9 16 9 12 6 9 8 3 3 17 7 28 22 6 6
 
 
the last 6 numbers are - 3hero games, 3 hero gamewins, 4 hero games, 4 hero game wins, 5 hero games, 5 hero games wins
so if you add them them up - 17+28+6 - you get 51 (the first number, or number of games played) and 7+22+6 = 35 (the second number, number of wins)
 
 
so... no idea why the formula that -seems- to need another subtraction of games isn't needing it.
 
 
 
the values if you are curious are (games played, win, win percent, times incapped, but still won, %incap, 1st Pos, win, 2nd pos, win, 3rd pos, win, 4th pos, win, 5th pos, win, 3 player team, win, 4player team, win, 5player team, win)

 

86

There is no way in the stats page to relate an alternate victory condition.  As of right now, we have to write them in the notes field.  I think this would be best solved by having different answers to the game won/lost section.  It seems to me like this would be easier to code, too, but hell if I know anything about that.  :slightly_smiling_face:

What was the result of the game?

  • Victory by destroying the villain
  • Victory by destroying Gloomweaver's Relics
  • Defeat by incapacitation
  • Defeat by Terralunar Impulsion Beam
  • Defeat by Voss's Minion Overrun
  • Defeat by Omnitron's devices after Omnitron's defeat
  • Defeat by [redacting] [redacted]
87
  • Defeat by Mars explosion.
  • Defeat by Silver Gulch [redacted].
88

Given the way that i am pulling the stats, that doesn't ... quite... work

 

but ill change the question on 

 

If you Lost, was it because of HP or an Alternate Loss Condition?Enter: Incapacitated if every hero was taken out, or a cards alternate loss - such as TerraLunar Beam, or Self Destruct Sequence

 

to something more easy to understand (a checkbox like that, probably)

89

It's not just the alternate loss conditions to worry about - it's any alternate win conditions, like Gloomweaver's relics.  We need to try to have a way to put that into the form proper, rather than just the notes field.

90

gloomweaver relic is on that new version of that question and it now says how did the game end rather the how you lost

91

Excuse me, but I am a little confused by these "most effective" numbers.  I would assume that means one hero from every game is considered the most effective from the game, but then they shouldbe averaging out in the high twenties, insteaf of the low seventies.  Is every hero except one considered "most effective" each game?  If so, why?

92

If I'm not mistaken, Most Effective is merely the percentage of winning games that the hero played in, not a title assigned to one character per game.  Legacy was in 36 games, 31 of which the heroes won.  If we define "effective" as (at least theoretically) contributing to a win, then he is 86.11% effective.  

93

Flamethrower is right. Without sitting down and watching every single game, and then coming up with some way to measure 'effectiveness' within the game, there is no way to determine who is most effective in any indivdiual game.

 

Rather that stat is simply - THis hero wins more than others - Legacy rarely ever is in a loosing game.  -- 83percentof the time (as of this post) if legacy isin the game,  the heroes win. You can draw from that (moreso once we have more data) that legacy is a large contributing force to wining the game 

 

of course with our current data #s its not perfect. Wraith:Price of freedom has won 100% of her games. Of course shes only been in 2 games XD  But we're easily pushing 200-300 games a month (we have over 130ish new games n the tracker already and its only half way through the month!)  and that just means that it will continue to get more and more accurate as time goes on. 

 

 

But you also can't look at a hero just by that single stat - its ment to be a quick 'This is the most likely to be in a winning game' stat yes, but ther are other things to consider

 

like Legacy -also- has a 23% incap rate. Meaning in 23% of his games he was incapitated. But that doesnt add up to a 100 if you notice with the wins... because ~30%  of the time if he was incapitated, the heroes still won.  -- which is one of the better numbers in that category. Which all this does is prove with numbers and hard data that Legacy is one of the 'More Successful' heroes in the game (i refuse to say /better/ but there is clearly a track record here that needs to be noted.

94

The one statistic I can't see on the sheet, and the on I am most interested in, is of the wins how many occurred with at least one hero incapacitated? Almost all of the wins in my group are with all heroes up, we find that heroes are like dominos, once one falls it's hard to keep any standing.

95

hmm. Im not sure how to formulate that to pull information from the sheet, though I think I could. 

 

But you can also look at the Raw Data  at https://docs.google.com/spreadsheet/ccc?key=0AqClP82b_XgZdFVreEhkOWJYcmlvcDNMbWppejh5aGc#gid=0 

and do a quick scan yourself - ill try to see if i can't pull a formula out for that one - might not make it onto the official sheet but would be availabel to view elsewhere

96

I think I can be of assistance with that. If I download the raw data as an excel file I'm fairly certain I can write a formula which would calculate the percentage of games won with various heroes/positions incapped. The only thing I don't know is if Google Docs uses the same formula commands as excel.

97

it does and it doesnt - it does not have all the commands (other wise i would be using Functions instead of arrayFormula 's ... bleh.)

 

you can look at the format for the setup to see if you can try to write a formula, but again - a good chunk of the commands that excell has arent in gdocs yet

98

Having actually had some free time that I did not spend on playing Sentinels, I spent my free time making some quick formulas to analyse data about Sentinels!

 

Just some quick calculations on incapacitated heroes and the current results;

There are 48 wins with incapacitated hereoes (12.73%)

There are 203 wins with no incapacitated hereoes (53.85%)

There are 126 losses (no matter the status of the heroes) (33.42%)

 

Looks like 1 in 7 wins is a win with an incapacitated hero, higher than I would have thought. My personal stat is 3 incap wins out of 27 total wins (48 games total), so I'm at average for incap wins but letting the side down with almost 50% losses!

 

 

(I can provide the formulas I used if people want them for the spreadsheet, I wrote them on Excel)

99

I just want to say.

 

Vossh has now overwhelmed the earth with his Minions once.

 

And Baron Blade has crashed the moon into it /twice/

 

Awesome. Just. Awesome.

100

I've started keeping a note of our game stats so I can contribute here - Just uploaded our stats for this weekend :). I went home early 'cause of the snow and didn't want to risk that the trains might not be running later this evening (plus it'd be dark and even colder by then). So we didn't get round to fighting Blade, Rat, Gloomy, or the Ennead. We did the other eight, though. And finally beat the Matriarch after ages of failing to do so! And we did it in Atlantis (we choose our villain and environment combos via dice roll, so this is never our choice!), where she kicked our arses twice last week! Legacy's combo of Fortitude, Lead from the Front, and Next Evolution (Psychic immunity) meant we could basically ignore all penalties for killing birdies even before she flipped :D. Plus Haka had eaten most of her stuff by the latter half of the game - he had the Mask, both cohorts, and a load of other birdies all stuffed under Savage Mana. Bunker managed to get a few cards under his Omni-Cannon and in the last round he let rip with this (buffed by both Legacy and the Adept, as was Haka), dropping the Matriarch's health to the point where Haka could unleash his own assult with Savage Mana, finishing her off :D.

We had a similar victory earlier against Citizen Dawn, in which Haka had about nine(ish) cards under Savage Mana, plus was buffed by Legacy. He then used Haka of Battle and poured his entire hand (about 17 cards or something like that) into fuelling the damage Savage Mana was about to do. Just to make a big finish of things. Because Dawn had something like 2hp left. Ahh, now that was satisfying :D.