Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[DStu]

As noted elsewher the fact that the username of the one who hosts the game is encoded in the logfilename on goko lets you estimate how many people own which sets. Just download all logs, see which cards are used by which hashes and map them to the sets.

Now the hard part of this is to get all the logs, but I think I've complained enough on this elsewhere.  I only scanned the logs from 4 days, one in February because I had the logs from this day already, and 3 days last week.  I won't scan more because downloading the logs takes hours; also my script is too slow which could probably be fixed by using perl or something instead of bash for scripting, but there is no point in working for this if it takes me ages to get the logs in the first place.  Also, the logs from before April are gone afaik.

Anyway, numbers of 4 days probably don't say much, but now that I have them I just share. First, number of users that are seen with a card of a certain set for the first time:

date	Intrigue1	Intrigue2	Seaside1	Seaside2	Alchemy	Prosperity1	Prosperity2	Cornucopia	Hinterlands1	Hinterlands2	DarkAges1	DarkAges2	DarkAges3
2013-02-05	389	354	339	385	4	370	376	279	268	247	280	277	273
2013-04-06	839	781	745	819	445	818	800	656	645	599	652	628	603
2013-04-07	992	937	905	979	600	981	957	784	778	733	770	746	740
2013-04-08	1140	1075	1035	1138	739	1117	1120	910	910	856	898	861	864
2013-04-09	1186	1120	1089	1190	785	1169	1171	943	937	900	941	905	902
2013-04-10	1216	1146	1114	1215	805	1192	1195	960	959	922	963	930	928
2013-04-11	1233	1166	1143	1234	833	1214	1212	982	973	942	981	948	943
2013-04-12	1254	1176	1158	1257	852	1225	1227	996	985	952	999	959	960

Second, sum over the columns to get total number of users that are seen with a certain script up to the date.

date	Intrigue1	Intrigue2	Seaside1	Seaside2	Alchemy	Prosperity1	Prosperity2	Cornucopia	Hinterlands1	Hinterlands2	DarkAges1	DarkAges2	DarkAges3
2013-02-05	389	354	339	385	4	370	376	279	268	247	280	277	273
2013-04-06	450	427	406	434	441	448	424	377	377	352	372	351	330
2013-04-07	153	156	160	160	155	163	157	128	133	134	118	118	137
2013-04-08	148	138	130	159	139	136	163	126	132	123	128	115	124
2013-04-09	46	45	54	52	46	52	51	33	27	44	43	44	38
2013-04-10	30	26	25	25	20	23	24	17	22	22	22	25	26
2013-04-11	17	20	29	19	28	22	17	22	14	20	18	18	15
2013-04-12	21	10	15	23	19	11	15	14	12	10	18	11	17

[LastFootnote]

Did you remember to exclude Nobles, Haven, etc., since they're in the free sampler?

[Kirian]

So, $400k $40k or so in sales... likely closer to $600k $60k because, as you note, you only have a few days of data.

I guess that pays for salaries for a year or so, but man I wouldn't want the VCs to see those numbers.

Edited because I can no longer multiply in my head.

[DStu]

Did you remember to exclude Nobles, Haven, etc., since they're in the free sampler?

I took the content of the sets from goko's shop, so if they are also in the sets, they are included...

[SCSN]

So, $400k or so in sales... likely closer to $600k because, as you note, you only have a few days of data.

I guess that pays for salaries for a year or so, but man I wouldn't want the VCs to see those numbers.

How did you get those numbers? A reasonable upper bound for DStu's data is 1200 people who own all the sets and spent $50 each, for a total of $60k. As you noted it will be more because of the limited sample, but I doubt it'll add an order of magnitude.

[Kirian]

So, $400k or so in sales... likely closer to $600k because, as you note, you only have a few days of data.

I guess that pays for salaries for a year or so, but man I wouldn't want the VCs to see those numbers.

How did you get those numbers? A reasonable upper bound for DStu's data is 1200 people who own all the sets and spent $50 each, for a total of $60k. As you noted it will be more because of the limited sample, but I doubt it'll add an order of magnitude.

That would be because I have apparently forgotten how to multiply in my head.  Yes, you're right, it's closer to $60k.

So that pays one person's salary for this year.  Ouch.

[Kirian]

I won't scan more because downloading the logs takes hours; also my script is too slow which could probably be fixed by using perl or something instead of bash for scripting, but there is no point in working for this if it takes me ages to get the logs in the first place.  Also, the logs from before April are gone afaik.

I'm able to get one day's logs in about 30 minutes... there must be some way to automate it to run each night while sleeping, right?

[DStu]

I won't scan more because downloading the logs takes hours; also my script is too slow which could probably be fixed by using perl or something instead of bash for scripting, but there is no point in working for this if it takes me ages to get the logs in the first place.  Also, the logs from before April are gone afaik.

I'm able to get one day's logs in about 30 minutes... there must be some way to automate it to run each night while sleeping, right?

My PC is in my bedroom, so this not really compatible with sleeping.  Also, I need longer than 30min, how do you catch them?

Edit: I think 1200 is to low for an upper bound, you get 150/day quite constantly, I would guess 2000 is safer, but anyway...

[Kirian]

I won't scan more because downloading the logs takes hours; also my script is too slow which could probably be fixed by using perl or something instead of bash for scripting, but there is no point in working for this if it takes me ages to get the logs in the first place.  Also, the logs from before April are gone afaik.

I'm able to get one day's logs in about 30 minutes... there must be some way to automate it to run each night while sleeping, right?

My PC is in my bedroom, so this not really compatible with sleeping.  Also, I need longer than 30min, how do you catch them?

Edit: I think 1200 is to low for an upper bound, you get 150/day quite constantly, I would guess 2000 is safer, but anyway...

I grabbed a DL manager extension for FF and ran it on one of the directories.  I suppose it's possible wget would be slower, though.

[LastFootnote]

Did you remember to exclude Nobles, Haven, etc., since they're in the free sampler?

I took the content of the sets from goko's shop, so if they are also in the sets, they are included...

Then if I understand you correctly, you're including false positives and overestimating the number of sets sold.

Let me make sure I understand what's going on here. Say your parser finds a game with Nobles. It knows that Nobles is part of Intrigue's "Peasants and Aristocrats" subset. It infers that the user hosting the game has bought "Peasants and Aristocrats". This is not necessarily true, as the user could have gotten Nobles from scanning in the promotional card they received when buying a physical copy of Dark Ages or the Base Cards.

[DStu]

That's correct, and that might be the reason why Seaside2 is always a bit larger than Seaside1.

[SCSN]

Edit: I think 1200 is to low for an upper bound, you get 150/day quite constantly, I would guess 2000 is safer, but anyway...

Oh I agree, I meant an upperbound for the sales that underlie your data. Seeing that no set is owned by >1200 people, your data represents sales strictly less than 1200 * (price of all the sets).

[DStu]

/push for new data

[Blueswan]

Judging by the fact that it actually takes quite a bit of work to even reach the top 1000 of the adventures ranking, I guess the total number of people owning all the sets (or most of them) is way higher than 1200.

[Mic Qsenoch]

Judging by the fact that it actually takes quite a bit of work to even reach the top 1000 of the adventures ranking, I guess the total number of people owning all the sets (or most of them) is way higher than 1200.

I only have base and am in the top 150 adventures (not as Mic Q, it's an alt) so I'm not sure what you're saying at all.

[Blueswan]

Judging by the fact that it actually takes quite a bit of work to even reach the top 1000 of the adventures ranking, I guess the total number of people owning all the sets (or most of them) is way higher than 1200.

I only have base and am in the top 150 adventures (not as Mic Q, it's an alt) so I'm not sure what you're saying at all.

OK, I was wrong then. I was under the impression that ou had to play a lot of adventures to get that high.