Your story was a bit long and the space translator got stuck.
However, if you have a similar story in the past, these are good references.
Comment has been collapsed.
That discussion is just the plain luck score, i.e. just the difference between expected and actual wins. I think it is more useful to discuss proportional luck.
For example, suppose person 1 has 10 expected and 14 actual wins, and person 2 has 100 expected and 104 actual wins. They both have the same plain luck score (+4), but person 1 has a much bigger proportional luck.
Comment has been collapsed.
Hmm ok I'll give it a go:
I joined SG in Jan end, So I can only count Feb as a whole. (did u take urs as an average of all months?)
Games I won in Feb: 4
Luck in Feb: 0.51
Expected wins = 4+0.51
wait something is weird
Edit:
Proportional luck= 14.6%
Thank you OP for breaking it down.
Looking at other users results...Holy Sh@t, Im one lucky Son of B@tch! (for NOW!!)
Comment has been collapsed.
The steamgifts page I linked, https://www.steamgifts.com/stats/personal/community also reports your overall luck, not just per month. Both statistics are useful but separate discussions. In this paragraph on that page, the last sentence is your overall luck (the example text is what it says for me):
Luck Per Month
See if you've been lucky or unlucky throughout the past year. The graph estimates the number of giveaways you should have won each month based on the giveaways you entered, and displays the difference compared to the number of gifts you actually won. Overall, you've been unlucky, and won 4 fewer gifts than estimated since you first joined.
The proportional luck too can be calculated either overall or per month, the formula is the same. In your case, for February you won 4 games and your luck was +0.51. So your expected wins in February were 4 - 0.51 = 3.49 and your proportional luck for February is 0.51 / 3.49 = 14.6%. You have won 14.6% extra gifts compared to what was expected.
Comment has been collapsed.
That's exactly right! And yes, that's not bad at all. :)
Comment has been collapsed.
+0,63%
That seems to be within reason for my 8 years and millions of giveaways entered.
Comment has been collapsed.
I'm at -0.02%, gonna have to re-subscribe to SG Premium I guess. 😿
Comment has been collapsed.
You could also sacrifice some chickens, worked for fine me.
Comment has been collapsed.
Just here to balance out the good results.
-12/20= -60% -11/21+11=-34%
Comment has been collapsed.
So.. Have won 134 games. Have won 4 more games. So expected wins 134-4 = 130. proportional luck 4/130 = 0.03 -> 3% (I think you have to point out that you move point by 2 decimal places to get percent. Almost missed that one myself)
So basically I've won 3% more then expected?
Comment has been collapsed.
Wow! Since you have a relatively large amount of wins, having that large (negative/positive) proportional luck is quite surprising.
Comment has been collapsed.
Oh yes, slight error there, your number should be -136 / (3837 + 136) = -3.4%. Still a fairly big negative luck with that many wins!
Comment has been collapsed.
Haha ;) Thankfully it's not a tautology like that. In your notation Expected Wins = x, not Luck / x.
Comment has been collapsed.
+2.82% in 3 years 5 months.
Since PRNG are "supposed" to create normal numbers, pluck tends to zero the larger the sample is. Indeed it's more significative looking at it as proportional to wins than just as a bare number outside any context, and I always thought the "luck per month" thingy was missing that info. However I doubt there'll be too many exceptions to proportional luck in users with a large number of wins, specially reaching about 1000. There are peaks and valleys, there are exceptions, but all in all we'll doomed to have neither positive nor negative luck. Which, for someone who doesn't think he has much luck (albeit facts prove otherwise), it's something lucky in itself.
Comment has been collapsed.
Yes, on average it should even out over time, but who knows how long it will take. It seems BreadMan above is an example of someone with many wins and yet strongly nonzero proportional luck. [Edit - that number had an error.]
Comment has been collapsed.
think that the basic luck statistic is interesting but it is even better to discuss it as a proportion of the expected wins.
I also think this is a way better method for calculating luck, as it removes the differences that comes from sheer number of wins. Somebody with 10 wins can't have a -28 win rate, which absolutely can happen if someone have multiple thousand wins.
Mine is -2 / 333 = -0.6%
Comment has been collapsed.
The steamgifts page I linked, https://www.steamgifts.com/stats/personal/community also reports your overall luck. In this paragraph on that page, the last sentence is your overall luck (the example text is what it says for me):
Luck Per Month
See if you've been lucky or unlucky throughout the past year. The graph estimates the number of giveaways you should have won each month based on the giveaways you entered, and displays the difference compared to the number of gifts you actually won. Overall, you've been unlucky, and won 4 fewer gifts than estimated since you first joined.
Comment has been collapsed.
Nice! That's the highest positive proportional luck reported so far in this thread.
Comment has been collapsed.
-10/44 = -0.227272..
so -23%.
Havnt won anything in the last 4 years so that seems about right.
Comment has been collapsed.
I doubt the stats mean much with such a tiny sample size, but so far for me: -1/2 = -50% bad luck :)
Comment has been collapsed.
11/53=20.75%
I'd say I've done alright here.
*edit: fixed a typo, but the percentage is still correct
Comment has been collapsed.
Almost! If you have 11 more wins than expected, that means your expected wins were 465-11 = 454, and your proportional luck is 11/454 = 2.4%. So you have 2.4% more wins than expected, I'd say that's a nice number over so many years.
Comment has been collapsed.
I suck at math - I've won 7 times, and the stats page says I've won 3 more than normal. So my luck should be sitting at around +0,75%, right?
Then again one of those wins was in a giveaway with 9000 entries (which if I'm not mistaken means around a 0.0001% chance of winning) that ended up being fake, so even good luck can be a real cruel mistress sometimes :|
Still, that's pretty amazing - very grateful to the steamgifts community, hopefully I'll be able to contribute soon
Comment has been collapsed.
It would then be 3/4 = 75%. Too bad about the one fake win, but it's a quite a high proportional luck nevertheless!
Comment has been collapsed.
16,295 Comments - Last post 8 minutes ago by Haplodh
25 Comments - Last post 10 minutes ago by lewriczin
1,519 Comments - Last post 12 minutes ago by Tristar
1,798 Comments - Last post 20 minutes ago by Cacciaguida
543 Comments - Last post 31 minutes ago by Aristofop
44 Comments - Last post 1 hour ago by Chris76de
517 Comments - Last post 5 hours ago by Marius11
66 Comments - Last post 3 minutes ago by DrTenma
79 Comments - Last post 3 minutes ago by CallMeKap
200 Comments - Last post 3 minutes ago by samwise84
81 Comments - Last post 5 minutes ago by philipdick
38 Comments - Last post 15 minutes ago by AlexForestry
16,779 Comments - Last post 23 minutes ago by MjrPITA
3,366 Comments - Last post 40 minutes ago by KPopPoyehavshiy
Hi all,
would you like to share what is your luck on Steamgifts?
The luck score means whether you won more or less than the expected amount of games based on the final odds of all the giveaways you joined. For example, if you joined 10 giveaways with 5 participants each, you have 1/5 chance to win per giveaway, so your expected amount of wins is 10*1/5 = 2. If you actually won 3 of those giveaways, your luck is +1, if you won 1, your luck is -1. Steamgifts reports your overall luck score at https://www.steamgifts.com/stats/personal/community , e.g. as "Overall, you've been unlucky, and won 4 fewer gifts than estimated since you first joined."
I think that the basic luck statistic is interesting but it is even better to discuss it as a proportion of the expected wins. So I propose that we could discuss "Proportional Luck", which is the proportion of wins you have lost due to bad luck or the proportion of extra wins you have gained due to good luck. Proportional luck computed as follows:
Proportional Luck = Luck / Expected Wins = Luck / (Current Wins - Luck)
So for example, I have currently won 38 gifts, and my luck is "4 fewer gifts" (Luck = -4). Therefore my expected wins were 38 + 4 = 42, and my Proportional Luck = -4 / 42 = -9.5 %. That is, I have lost nearly 10% of my expected wins due to bad luck.
Would you like to share your proportional luck score?
Comment has been collapsed.