Are you lucky?
Based on the finished GAs you have entered, you would expect to have won approximately 209.8 of them. Won 198, but not su unhappy since at least my CV ratio looks better that way. And i've won some that are better than 30 cheap-bundled ones, so i've got that for me XD
Comment has been collapsed.
"Do you feel lucky punk? That's what I'm asking.
You can't be too tough; you got killed by an aspirin.
And your one-inch punch, same size as your pecker.
Leave the rapping to me; stick to Chinese checkers.
I'd beat you in round two, but that'd be unbelievable.
No one in your family ever lives to see a sequel."
Comment has been collapsed.
Could you describe what goes wrong? Do you get a "Do You Feel Lucky, Punk?" link added to any GAs entered page? If so, when you click on it, what happens? Do you have any other user scripts running that affect SG?
Simply comparing straight number of GAs entered won't give the same results - someone who entered only GAs with 100 or less entries would have a much higher expected wins than someone who only entered GAs with 1000+ entries.
Comment has been collapsed.
I just discovered something cool. If you go to your entered giveaways and search for a game, the script still comes up. When letting it do its thing, it seems to try to calculate the estimated amount of times you would have won that game in particular. That's an awesome feature!
However, there seems to be something wrong with the calculation. When I search for a game I know I have entered lots of giveaways for, and probably should have won by now (going only by statistics/probability), I get that I would have won approximately 1.5 copies of it. Seems plausible, since I've entered over 100 giveaways for it, and at least 30 of those giveaways are 30 or less entries.
Then I tried searching for a game I've only entered two giveaways for. Both 300+ entries. Estimated amount of wins: 0.4.
With some fine-tuning of how the calculation works, this script could be a very interesting tool for the curious :)
Comment has been collapsed.
I don't think the script is doing what you think it is, sadly. What I think is actually happening is it looks at the current page and gets the total number you've entered - and then based on that, requests the appropriate number of pages from your total GAs entered history (in the first case, it'll probably be looking at the first two or three pages and in the latter it'll be looking at just the first page).
It's a nice idea for a feature though - I will look into adding it - should be as simple as adding a new parameter to the requests it makes.
Edit: it was as simple as that - this is now a feature in version 1.1.0.
Comment has been collapsed.
I haven't tried it yet, but this is a really great idea. I did a similar calculation with a spreadsheet for a while, and have recommended the exercise to others when they complain. This, if it works right, makes that estimating a lot easier to do.
Edit: I'm unlucky. I'm about 5 games short according to the estimate (would be expected to win 15, have actually won 10)
Comment has been collapsed.
Does anyone remember when SG itself had calculated wins or something like that, and people still kept complaining? not sure if that's why it got removed, or if it was to avoid further confussion..
Now here you are with an optional script for those who want to know, want to get happy -feelin' lucky, punk?- or just depressed as in I'll-throw-myself-under-the-train kinda way... Anyway, the choice is yours and when/if a mod stops by here and says something
ps. I haven't tried it; I got the feeling I wont like the result, so...
Comment has been collapsed.
By the nature of expectation, you'd expect half of Steam Gifts users to be "unlucky". So I can see why they'd remove it from the default feature set - who wants to upset half their users? Don't see why they'd object to an optional script (beyond the fact it involves requesting every page of entered GAs).
Comment has been collapsed.
It's funny, I entered this cause believe it or not, I was watching Dirty Harry ('71) and even tho I have to continue, I left it soon after that famous line!
yeah me neither, as it's optional and not harmful to self or others nor offers and advantage, so it should be okay :)
I'm clarifying this just in case I somewhat managed to be cryptic and gave -again- the appearance of being annoyed or mad :P
It's cool! I'm more interested in a predictive script for better winnings rather than the amount: for me it's quality over quantity, let's say for example: how many bundle games or cheap games do the user has to "resign" aka not enter and/or how many does he/she has to create to be "prone" to win something nice?
That's far enticing for me that the maths, unfortunately it's also near impossible and/or inexact so I guess it belongs to dream land :P
Practical example you say? sure! Assume I want Rise of the Tomb Raider, this surreal script would suggest me to create... say 50 bundled games giveaways, and not entering more than 3 giveaways, excluding the REAL rise of TR -which you want to win-
don't take this last part seriously, it's more of an off-topic delusion :P
Comment has been collapsed.
Out of 14,635 entries, I should expect to have won 70.6 of them. I have actually won 71 so I guess I am nothing special :)
I have a thought for a future version that could be interesting, perhaps cache the parsed data into a text file. Save the URL of that entry in the file so that you can easily check against it and prevent double counting. This would allow the tool to run more quickly in the future with even less impact on the SG server.
Comment has been collapsed.
Caching the results is probably a good idea - I'd probably have to do it within the browser though (as I have no web server to do the caching on). Once I'm happy with the feature set (currently looking at adding a plot of how your expected wins has changed over time, and/or calculating expected points value of your wins), I will probably implement client-side caching.
Comment has been collapsed.
Nice tool! :) I'm a geek of statistics and I have always thought about how awesome a tool like this would be.
Indeed, I miss a couple of features that I think they will not be too hard to implement. I leave them here so if u want to include them =)
EV (expected value) calculation: well since you know about statistics I guess you know what this stat shows. The implementation should be easy I think, the tool just need to multiply de % of winning the GA by the number of points you need to enter it (e.g. 20p GA with 0,5% chance of winning: 20*0.005=0.01EV$)
Future statistics: since you are already reading the entered giveaways page you can also calculate all the statistics about already entered GAs. This data can be included in the plot. (Idea: plot the expected future winning in the same plot but in red for example, so you can see if you have a huge train or something like that) Also if you include EV calculation, future EV calculation would be awesome.
:) thanks again for the tool.
Comment has been collapsed.
Expected value was the top of my wishlist. Unfortunately, the points value of GAs isn't shown on the entered page and I'd have to load each of the individual listings to get their value - which would make my script even less performant and heavier on SG.
I did consider adding future statistics - but since they will be non-stable (i.e. more people will enter or leave before each GA finishes), I was reluctant to do it.
Comment has been collapsed.
What this script does is calculate how likely it would be for you to win GAs based on what you entered. If you entered a GA with say 500 other participants, you would expect to win 0.002 (1/500) of the GA. Obviously that doesn't really mean much as an individual value, but when you add them together, it should give you an idea of how many GAs you would have statistically expected to have won over time.
The plot is exactly that, but done daily - it sums all the expected win values for each day - so in your example, you can see a peak at the end of January - so you know you were more likely to have won a GA that ended then than the day before or after. The idea is the plot can give you an idea of whether you are getting better at entering high likelihood GAs or not.
Comment has been collapsed.
This script keeps getting better and better. I love the plot option that's now available.
Would it be possible to exchange the expected wins per day in the plot to total expected wins (adding all day values)? For me, that would be a more interesting plot. If possible, it would be totally awesome if you could plot both total expected wins and actual wins (collected from the wins page), so we can compare the expected wins to when the actual winning has (or hasn't) happened.
Comment has been collapsed.
Sorry, I'm not entirely sure what you're suggesting. Do you mean a cumulative plot? (I.e. each bar is the total expected wins up to that day)
Actual wins is a little bit of a pain as I'll have to query the won page as well, but it shouldn't be that hard.
Comment has been collapsed.
Yep, cumulative plot (couldn't remember what it's called in English when I wrote... third language and all).
I can imagine it would be a bit more work. I'm just taking a few ideas from the spreadsheet I have. Comparing actual wins to estimated wins is one of the most interesting parts (for me). When you first came with the script I didn't even think about possibilities for plots, but I'm very impressed with what you've done in just a few days.
Comment has been collapsed.
I've added the most requested feature of fetching your wins to compare it against your expected wins - as a percentage and as an extra line on the plot. Please let me know if it doesn't work properly if you have more than 50 wins - I can only predict what the interface will be (I assume it does pagination in the same way - but have no way of easily checking.
Comment has been collapsed.
'Apps,extensions,and user scripts cannot be added from this website'
How to fix this?
Comment has been collapsed.
It's what shows up in Google Chrome. The browser downloads the script instead - opened in notepad, ctrl+c/v to console and the code works... But I cannot plot the result, after clicking this "plot it!" link I get this http://i65.tinypic.com/jaxwdu.jpg
Does it happen because I don't use Tamper/Greasemonkey?
Comment has been collapsed.
Yeah - you'll have to install the Chrome extension "Tampermonkey" before you can use my script I'm afraid. It fetches an additional Javascript library for plotting - which is why you get that error.
Comment has been collapsed.
1,770 Comments - Last post 41 minutes ago by D3STRUCTION
2 Comments - Last post 1 hour ago by maruten
51 Comments - Last post 2 hours ago by RCSWE
541 Comments - Last post 2 hours ago by nalf2001
13 Comments - Last post 3 hours ago by osztihun
21 Comments - Last post 6 hours ago by Seibitsu
3 Comments - Last post 9 hours ago by lostsoul67
11 Comments - Last post 2 minutes ago by erintesden
20 Comments - Last post 13 minutes ago by Prosac
112 Comments - Last post 53 minutes ago by achilles335
32 Comments - Last post 58 minutes ago by Vincer
125 Comments - Last post 1 hour ago by achilles335
713 Comments - Last post 1 hour ago by Mikurden
6 Comments - Last post 1 hour ago by JMM72
Well, do ya?
I've noticed quite a few people complaining that they aren't winning as many GAs as they'd like. Well, I have a little stats knowledge and plenty of programming knowledge, so I thought I'd make a user script that looks over your GA entries and see how many times you'd actually expect to have won, which you can compare to your actual wins and either feel awesome or have something else to complain about!
It's a standard Greasemonkey/Tampermonkey userscript - please let me know if you have any problems using it. It will add a link to the top of your Entered GAs page (http://www.steamgifts.com/giveaways/entered) and will look back at all entered GAs when you click on the link. Don't expect results immediately - the process could take a while (mostly to avoid overloading SG).
https://raw.githubusercontent.com/kelnage/sg-lucky-punk/master/Do%20You%20Feel%20Lucky%2C%20Punk.user.js
Happy with the results? Or do you want to console yourself in your unluckiness? This thread is the best place to talk about it - thanks to super-user tso184 for creating it.
Edit: thanks for all your feedback. If you got a weird result (significantly lower or higher than you'd expect), please make sure you've got the latest version (by following the above link) and I think I have ironed out most of the issues people initially saw.
Edit (2016/03/22): new functionality! You can now see your daily expected wins on a bar plot (thanks to plot.ly for their fantastic JS plotting library). The userscript will now take a little longer to initially load (due to caching of the remote script) but once that's done it should be as fast as usual.
Edit (2016/03/26): now you can plot your actual wins against expected wins. See when you've been lucky or otherwise. Also includes a percentage "luckiness" (or unluckiness if less than 100%).
Edit (2016/03/28): fixed incompatibility with Extended Steamgifts userscript.
Edit (2016/03/29): changed phrasing of percentage (un)luckiness since I didn't make what it meant clear. Also excluding GAs between June and October 2014 due to a database corruption at that time.
Edit (2016/03/31): released version 1.5.1, caching results, so after your first time, the results will appear much quicker. Also changed behaviour of graph to show a constant line when no changes have happened (i.e. between wins or no GAs entered).
Edit (2016/04/14): thanks to the sharp eyed luizfcsilveira, I've found a bug in my script that meant all GAs giving away 1000+ copies of games were being ignored. The new version (1.5.4) fixes this. It will automatically clear your cached results when you first run it sadly, but it will also make your expected wins more correct!
Edit (2016/04/19): released version 1.6.0, which now shows future expected wins values (based on the number of entries when the script is run). Also works when searching for a specific game.
Comment has been collapsed.