The Best Heroes-A Statistical Community Effort

Other Games SotM Game Engine
61

The stats arent put together yet. We don't have nearly enough data (only something like 30 or 40 games atm)

 

You can look at the list of games submitted and get an inkling of what you see - 

 

https://docs.google.com/spreadsheet/ccc?key=0AqClP82b_XgZdFVreEhkOWJYcmlvcDNMbWppejh5aGc#gid=0

 

is all the submitted games to date. Gives you some idea, but no hard data. 46 (as of this posting) games is such a small slice of data to even be close to useless. :P

 

Its going to be months before we get any signifigant data. But its worthwhile to build too!

62

hey guys

 

I just want to say about halfway through the month we have about 75 games. Thats pretty awesome! If this keeps up we can easily see around 125-150 games for the month, which while not a definitive statistical pool, ist still a pretty awesome one for our first month.

 

This stuff is great, encourage your friends, reference this sticky, get people posting games! :)

63

I was wondering, can we get this stickied on other forums as well on this site? I feel like a lot of proficient gamers hang out on this one, but perhaps we would get a broader pool on the other threads?

64

I was actually considering seeing if we could get a 'Awesome Stuff For Sentinels Fans!" thread stickied in theGeneral, with links to the Wiki, this thread, Spiffs stuff, anything else we can come up with that should go in there...

 

65

I don't visit the Rules and Gameplay forum enough, clearly. This is an awesome idea! Hopefully youll get a nice big data set to do some interesting analysis. Do we have any bond fide statisticians on the board? Or anyone who knows one? Some advice on how best to organise the data would be helpful, to make analysis easier down the road.

On the topic of analysis, I'd be really interested in more complex analysis that could be done to check certain combinations of heroes or other combinations of factors. For instance, how often does Fanatic win when paired with Legacy compared to without him? How much worse is the Chairman in Rook City compared to Insula Primalis? Or how often does [REDACTED] win in [REDACTED] games? (often, I expect). How does the presence or absence of healing heroes affect games against plague rat? Or how important is it to have [REDACTED] in games against [REDACTED]? (pretty important).

Also, knowing how often people lose to villains of particular difficulties will be very interesting when developing further villains - how often does >G want a level 4 villain to win?

Answering most of these questions with a usefully small confidence interval will require many hundreds or even thousands of data points. Sounds like fun!

66

i would love to do which 'Teams' win most often, but given the number of combinations of heroes (yeaaah. thats a lot. if i remember my  math right its... 15x14x13+15x14x13x12+15x14x13x12x11 ... but i was always bad at math) without a computer crunching the data i don't know if that would be even possible to do. 

 

The advantage is that allll the data will always be stored in the file. And I believe ive set it up well enough that we can crunch new information from old data eventually, if we choose to. Like I made sure we're asking about team order, but havent figured out a way to accurately figure out how that effects heroes (like is Team Leader Tachyon better in first slot or last slot? When is it best to put Argent Adept?) 

I'm also not a statitician :P

 

But at least by the end of the month we should be looking at 150-200 usuable entries in one month. Thats awesome and a good starting point. Put 3 in myself last night!

67

Just put in my first contributions. Great work on this, guys.

68

After this weekend ill be compiling the first months data. We have about 150 entries at the moment (most of them from Sound of Trees lol!) and we'll take a look on what that means. We'll loose a few due to having Playtesting stuff in them before it was pretty much finalized, but still a good start. Not nearly enough to create truely accurate data, but it is a -start-.

 

So far I'm surprised to see that about 25% of games are a loss. That is surpising to me given that in my playgroup we only loose about 10%

69

I did not enter my last games - I am jobless and play at least three games a day, if I keep entering everything there will be too big a bias.

 

Edit : I made a little analysis of the data some times ago, and many results were quite surprising. Barn Blade, especially, seemed more dangerous than we are used to think - maybe because some entries are from newer teams ?

70

Go ahead and post all your games.  We want raw data now.  A little more bias towards the SoundofTrees playstyle won't ruin anything.  I'm sure we'll get a lot of interesting people logging their plays here, and things will even out.  I'm storing up a bunch of post-playtest games to post once the known incarnations of the new guys are confirmed as accurate.

On another note, I have seen people talking about how they were doing something particular (like sprucing up his mechanics to have fun with Gloomweaver), but had to stop because they want to post their games.  I say just have the fun you want to have, and don't post the game.  If this results in fewer games against the official Gloomweaver, so be it.

71

Just to say that I'm totally stoked about this initiative, I'm a stat freak.  As a matter of fact, once I become more familiar with the game I may be able to help you compile some of these statistics.  I will let you know (i am receiveing the game soon), that I will totally contribute to this although the majority of my submissions (likely between 66% - 75%) will be solo plays.

72

 

I have been watching the forum for quite a while but this is my first post. I only started to upload data since this week and so far I have uploaded 8 games if I'm not wrong and my winning rate is around 50%. I only play solo with 3 heroes and there are a lot of times I can tell I can win by having just one more hero. However I prefer the challenge of using 3 heroes, plus 4 characters do take up a lot of space.

 

I did go through all the data had been collected, and it is pretty obvious that the winning ratio of 3 heroes is much lower than that of 4 heroes. So I guess it is safe to say that it's not really fair to judge the rate  merely  from winning or loser while neglecting the number of players. Although the H mechanics was introduced to have a better balance among different number of players, that does not change the fact that the more heroes there are, the easier it will be.

73

yeah im excited to get into this  - probably will be transfering over the stuff to the 'data sheet' on tuesday afternoon, or weds morning (Have a new years eve party to go to, dont know when Ill be home lol)

 

As for the Playtest, I believe they have hit their final iteration and any changes that occur will be minor in effect that it -should- be fine for us to start including them now as of Jan. Unless something major shows up in the playtesting forms...

74

The First Statistics are POSTED!!!

 

 

https://docs.google.com/spreadsheet/ccc?key=0AqClP82b_XgZdGJKUVJRakJmZ3RMX2pFUjB3MEtBWHc#gid=0

 

 

(zelinke, if you dont mind addding that link to the first post please?)

 

Couple things

 

1) I gotta find a better way to do this. It took me nigh on 9 hours to input everything. Admittedly i was getting up and moving, and reorganizing the charts, setting up some auto populating fields so i wouldnt have to enter everything 3 times, and completely started over after about an hours worth of work cause i wanted a better way to do it... but still. Oy. )

2) So with that above thought - They are not currently sorted by most effecient or whatever - becaue apparently googledoc excell cannot sort a field once it has a formula in it and i used a lot of formulas to make it much easier for me to enter things.

3) I am going to work over the next few days on seeing if i can make this much quicker and easier for myself, so that form may change somewhat in the final days. I think I have a way to do it, but we'll see.

 

 

4) I threw out ALL the playtester data entries. The decks I think are 99% done now, so any future ones ill put in, but just to be safe, there is no playtester data in this release. So with that in mind there are only 170ish data points - meaning that this data is far from conlusive in anyway, so PLEASE remember that.

 

 

5) Chairman is the most dangerous with a only a 60% loss rate.

6) Baron Blade is surprisingly second with a 43% loss rate - i chalk this up to new players and baron being their first one. With more data this will probably smooth out.

7) Legacy Trounced his Nemisis with a 100% win rate (including Mad Bomber, and promo Legacies) over like 10 games

8) Fanatic did well to at 100% over Apostate with 2 games

9) Sadly for my feelings, Expatriette and Mr Fixer could not defeat their Nemisis with 0% win rate (1 game for Expat, and 4 games for Fixer)

10) Regular Ra did not ever play the Ennead, but Ra: Horus of the Two Horizons played them once and defeated them.

11) Megalopolis was the most played in enviorment, but the Ruins of Atlantis were the most dangerous with a 44.44% loss rate.

12) 4 Hero teams won nearly 80% of t he time, while 3 hero teams barely 54%. 5 hero teams (with only a 9 count of games) won 88%of the time.

13) The most played hero is Absolute Zero (with 43 games)

14) Spiffs Randomizer is the most popluar way to play, while Picking your team follows next.

 

 

Anyway. Thats some interesting stuff i saw.

 

Feel free to browse through it but please remember again, this is a small data pool and we have a long way to go before any conclusive information is presented.

75

If it makes you feel better, I logged a play last night with Mr. Fixer beating The Chairman.

 

Also, a big thank you for putting this together :D

76

When I think of the hundreds of games I've played that are not represented here ... I want to make sure I put all the future ones in.

77

I'm going to try to keep better track of all the games I have played and post them up as well. I tend to play a lot of solo games with only three heroes, that would probably increase the percentage of wins for that catergory at least if I were to tally that in there. So from now on I will carry a sheet of paper inside my game box so I can right down the stuff.

78

Not sure if this was mentioned, but I think tracking if the game was played solo or multi player might provide some good information.

Solo wise, I would suspect players would tend to win more - being able to strategize and coordinate better, etc.

79

I dunno. We coordinate pretty heavily in our multiplayer games. I would actually expect multiplayer games to have a higher win rate because you have more eyes on everything to catch stuff the one person might miss.

80

I just want to say that this is incredibly awesome. Keep up the great work!