I seen a guy with almost 9 estimated wins but without any actual wins.
Comment has been collapsed.
My first win was well before EW was added to the site.
Comment has been collapsed.
A weteran! I respect and envy that. I remember how proud I was when we were developing some gaming site (it's now commercial one, gaminator.pl, but in the beginning it was just a three-man free-time project) and I was a proud owner of an account with id=3, where id=0 was for testing and 1 and 2 were programmers accounts :)
I think it was one of the happiest moments of my professional life. Even when the project and programmers was bought, and the designer (that's me) was bailed out. I was happy until I saw our innovative ideas (it was community page well before facebook, xfire, or steam community, with screenshot sharing and tools for scheduling multiplayer games) were replaced with stereotype commercial approach. Still it's decent gaming portal novadays.
Comment has been collapsed.
You don't have to be that veteran-y to have had a win before the EW stat was added. I linked it up at the forum and it looks like it was implemented about 4 month ago, so it's a pretty new stat. I started using this site in December 2011 and I wouldn't consider myself as a veteran. But that depends on your own opinion.
Comment has been collapsed.
My estimated wins is 10.43, and my actual wins is 4. :/
Comment has been collapsed.
Estimated wins 5.52, actual wins 2. I won my first game when I had 3 estimated wins.
I also won on my birthday a couple of days ago, only to be told by the guy that ran the the giveaway that he doesn't have the game to give me. I call that pretty unlucky :(
Comment has been collapsed.
GA entered: 1715
Gifts won: 2
Estimated wins: 4.11
When I won my first game however (Dungeon Defenders Collection), my estimated wins was <1, so I'd say I was fairly lucky then. Now...not so much. Not complaining though.
Comment has been collapsed.
More important than estimated wins in the quality of wins.
I would not mind a EW of 10 with an actual wins of 1, if that actual win was Bioshock Infinite.
Instead of 9 DLCs and a Fortix.
Not that I am complaining, I have had some very nice wins.
Comment has been collapsed.
I remember having like 7 wins with a EW of like 2 or 3.
For some reason I won most days for a week and a half even with minimal contrib and being in no groups (when I first joined). I probably had a EW of like .2 or far less for my first win, but I wasn't looking then.
Then, I went about 3 months with nothing. now I just won 4 games in as many days, and have a EW of 14 with 20 games won.
Comment has been collapsed.
Estimated wins wasn't a feature on the site when I won my first giveaway so I can't really tell you. It happened within a few hundred entries though.
Comment has been collapsed.
I saw random forum people with 1 estimated and 8 wins. Forgot who was he
Comment has been collapsed.
My friend signed up like 1 week ago and already got 3 games :S
Comment has been collapsed.
but hey its a group GA :D do you know when they will accept new members ? :D
Comment has been collapsed.
This seems to come up regularly or slight variations thereof, you can calculate what the likely highest and lowest estimated wins prior to an actual win would be given that we know there are currently about 330000 members on this site. The mathematical formula required is rather complex and you may need to ask a teacher if your still at school or a nerdy friend or whoever if you are not well versed in maths lore. The distribution curve for this problem requires a logarithmic graph, or a semi-log graph as some call it as only the x axis, which should indicate the expected wins to actual wins ratio needs to a logarithmic scale, the y axis is just a linear frequency indicating how many people have this expected wins to actual wins ratio. From there due to the large sample we have (the sample is the quantity of users on the site who have entered at least one giveaway) you can base your calculations entirely on averages (you can look these up I don't know them off by heart). Now you can fit this into a linear graph showing the z-scores on the x axis (also known as the standard deviations) and keep the frequency on the y axis. We know that the standard distribution ought to fit 68% of steamgift members within one standard deviation of a ratio of 1, therefore when you refer to your table of averages (I still cant be bothered to look it up for you) you will have the value of a standard deviation (find the members who have a ratio of more than 1, then take the closest 68% of them (closest to one of course)(this will be equal to 34% of the total users since we are basing it off averages) and now find the user in that 68% who has the highest ratio (refer to your table of averages for this again)) and you now have the value of one standard deviation (this also works using the same method on the lower side of 1). Now the bit that is really interesting for this thread, is that 68% of users will be within one standard deviation (they are the boring ones) but 95% of users will be within 2 standard deviations (also not that interesting) and finally 99.7% of users will be within 3 standard deviations (a spot of mental arithmetic tells me that will be approximately 900 users). So now apply the results of having calculated what 3 standard deviations are equal to both above and below the ration 1 and you will find the range within which 450 incredibly lucky people have won their first game and subsequently also the range within which the 450 unluckiest people have won their first win.
Comment has been collapsed.
As you would know if you had read it, the audience that the wall of text is intended for does not care much for formatting, correct grammar, punctuation, spelling and of course most importantly facts are far more important to them. So please, if you wish to try to find errors go right ahead, but make them worthy errors.
Comment has been collapsed.
You took the effort I was too lazy to do. I'm somewhat familiar with statistical methods, but didn't think through the calculations needed. I know it's computable but I was interested in real maximum values existing, not the probalities (as there is infinitely small probalility of infinitely large value). It's also easy, one well written SELECT would probably do it, if I would happen to have access to the site database ;)
But thank you for the analysis anyway. I used to stretch my mind with a little math too :)
Comment has been collapsed.
Yeah thats entirely right, statistics based on averages will never give you the values of the outliers. Its just that the chance that such outliers see this thread and reply to it are not necessarily very high, so it could give you an indication of what a non forum user might have actually managed. Who knows maybe somebody rage quit this site a while ago because they felt they would never win after reaching some phenomenal score like 10 expected wins.
Comment has been collapsed.
I'm not planning on raging. I'm planning on setting a record. D
I dunno if I'll make it, though. I figure I need 2500 entries and 10 EW without a single win. At least.
Comment has been collapsed.
I read your comment and didn't know what to think and then I clicked your profile and I was like jeeeeeeeeeeeeeeeeesus talk about unlucky wtf lol. 8 and a half is well beyond the estimated 4 that has been edited into the OP, and in regards to the mathematical side of it doubtlessly you would be considered and outlier.
Comment has been collapsed.
that sucks... although if it makes you feel any better i have seen people with 8000+ entries in giveaways without a win (this was before estimated wins)
Comment has been collapsed.
wow. I have 460 entries less than that and already won 7 things. But the other guy is right I've seen people with 8,000+ entries and no wins also so it could be worse ;)
Comment has been collapsed.
Estimated wins are also calculated based on entry amounts, as far as I can tell. jperry has 7 won and 6.93 estimated wins.
That said, private and puzzle giveaways give you a much higher chance of winning than a public giveaway. I have won 3 GA's with an EW of 1.00, but all of mine were pretty easy.
Comment has been collapsed.
Estimated Wins is a REALLY REALLY bad way of determining if you are really lucky/unlucky or just a little lucky/unlucky.
Estimated wins only lets you know if you are lucky or unlucky. Variance or Standard Deviation would let you know exactly how lucky/unlucky you are.
Comment has been collapsed.
I've got 2 wins for GAR and red faction dlc (I'm not complaining :D Free is free :D + I had a lot of fun with gar and red faction) and have 4.31 Estimated wins :D
Comment has been collapsed.
25 Comments - Last post 40 minutes ago by CelticBatman
76 Comments - Last post 1 hour ago by jonnysonny
65 Comments - Last post 2 hours ago by pb1
29 Comments - Last post 2 hours ago by lostsoul67
47,094 Comments - Last post 2 hours ago by paco7533
48 Comments - Last post 6 hours ago by FranckCastle
4 Comments - Last post 7 hours ago by Yamaraus
12 Comments - Last post 1 minute ago by AllTracTurbo
16,704 Comments - Last post 13 minutes ago by Operations
118 Comments - Last post 41 minutes ago by Naitas
16 Comments - Last post 50 minutes ago by windkit
117 Comments - Last post 1 hour ago by Yamaraus
640 Comments - Last post 1 hour ago by jonnysonny
7,932 Comments - Last post 1 hour ago by SergeD
Don't take me wrong, this is not another complain thread. I admit, it would be nice to win something, but that is not the reason I joined this site.
I'm just curious, what are the best and worst estimated wins for your first win?
I have puny rating of 1.22 (but with my bad luck I may go as far as 2.0 ;)), but I guess there must be people with 2.0, maybe even with 3.0, and some with 0.5 or maybe even 0.2.
So, who's got the best and the worst estimated wins when you won your first game?
EDIT: So right now we have:
The luckiest: Blommas, with 0.2 EW.
The least lucky: verksies, with circa 4.0 EW.
Comment has been collapsed.