Hi, Here is the log of the chatroom at http://farsides.com/chat/ (or irc://irc.freenode.net/#farsides) for the 2011-11-23.
2011-11-23 00:22:40+0000 blanchard (blanchard) is now online 2011-11-23 02:01:33+0000 mrjmad (mrjmad) is now online 2011-11-23 06:23:31+0000 dachary (dachary) is now online 2011-11-23 08:26:00+0000 dachary (dachary) is now online 2011-11-23 08:26:38+0000 <dachary> hello bonjour 2011-11-23 08:30:10+0000 <dachary> antoviaque: do you know how to interpret the "Visitor Age" page of owa ? 2011-11-23 09:30:11+0000 MCMic (MCMic) is now online 2011-11-23 09:34:58+0000 <dachary> blanchard: good bonjour sir ! 2011-11-23 10:00:20+0000 <antoviaque> hi dachary - sorry missed your ping : ) 2011-11-23 10:00:30+0000 <antoviaque> let me see for OWA 2011-11-23 10:03:44+0000 tartarugafeliz (tartarugafeliz) is now online 2011-11-23 10:04:06+0000 <antoviaque> dachary: for visitor age, I take the description "days since first visit" literally - ie when OWA identifies a recurring visitor, it looks at when was the first visit from him, and the difference gives the "age" 2011-11-23 10:06:25+0000 <dachary> hum 2011-11-23 10:07:52+0000 <dachary> antoviaque: so there are 184 visitors with age 1. Does this mean that 184 visitors came back the next day and then never returned ? 2011-11-23 10:08:39+0000 <antoviaque> dachary: I would assume it is the case yep 2011-11-23 10:09:24+0000 <dachary> And there are 23 visitors with age 13. This means that 23 visitors came back in the 13 days following their first visit and then never returned. 2011-11-23 10:10:04+0000 <dachary> antoviaque: do you find these figures interesting for your reasoning on the audience ? 2011-11-23 10:10:09+0000 <antoviaque> dachary: yep, ie that at the minimum they came once, and 13 days after that, and no more 2011-11-23 10:10:47+0000 <antoviaque> dachary: it's interesting, but it's less interesting than full-fledged cohorts 2011-11-23 10:11:25+0000 <antoviaque> for example, it doesn't say if they didn't came more than 13 days because the last time they came was today, or if they have dropped out already 2011-11-23 10:11:51+0000 <dachary> antoviaque: how is it interesting ? What reasoning can you do with these figures ? 2011-11-23 10:12:04+0000 <antoviaque> you can infer this kind of information by playing with the timeframes considered, but it's tricky 2011-11-23 10:13:53+0000 <antoviaque> dachary: well, at least you get a sense of the proportions on retention 2011-11-23 10:13:56+0000 <antoviaque> for example 2011-11-23 10:14:09+0000 * antoviaque doing some quick calculations 2011-11-23 10:16:59+0000 <antoviaque> for example over the last week, 115 people were returning visitors in total, vs 4300 visitors who just came one day 2011-11-23 10:17:00+0000 <antoviaque> you can remove 1K from yesterday, 3.3K visitors 2011-11-23 10:17:00+0000 <antoviaque> and remove the bounce 2011-11-23 10:17:25+0000 <antoviaque> it gives 583 people 2011-11-23 10:18:25+0000 <antoviaque> so you get a "feeling" of the proportions of people who are hooked vs new players 2011-11-23 10:18:31+0000 <antoviaque> it doesn't give your the retention rate 2011-11-23 10:18:50+0000 <antoviaque> (because a part of the returning visitors are from before the timeframe considered) 2011-11-23 10:19:12+0000 <antoviaque> but it's some guides 2011-11-23 10:19:52+0000 <antoviaque> the real good way to look at this is really cohorts - because then you can follow the evolution of a given group for a given timeframe over time, and you get numbers with real meaning 2011-11-23 10:20:02+0000 <antoviaque> does that help? : ) 2011-11-23 10:21:51+0000 * dachary thinking 2011-11-23 10:26:55+0000 <antoviaque> btw 2011-11-23 10:27:11+0000 <antoviaque> for cohorts, I think it would be tricky to have those from OWA 2011-11-23 10:27:40+0000 <dachary> antoviaque: how do you see these numbers you mention ? 2011-11-23 10:27:42+0000 <antoviaque> for this kind of things, rather than twisting it to do what we want, it could be good to just store more information in the DB 2011-11-23 10:27:56+0000 <antoviaque> and then use BI tools to build the cohorts 2011-11-23 10:28:30+0000 <antoviaque> dachary: the 115 people I just added the number of each visitors with age > 1 for last week's visitors 2011-11-23 10:28:50+0000 <antoviaque> 1K are visitors from yesterday 2011-11-23 10:28:55+0000 <antoviaque> and bounce is 83% 2011-11-23 10:29:12+0000 <dachary> ok 2011-11-23 10:31:03+0000 <dachary> what is your feeling with regard to the evolution of the user base ? For instance, what would be the retention rate ? 2011-11-23 10:31:21+0000 <antoviaque> I'm not sure 2011-11-23 10:31:29+0000 <antoviaque> that's what I'd like to find out : ) 2011-11-23 10:31:39+0000 <dachary> what does your gut tell you ? 2011-11-23 10:31:58+0000 <antoviaque> if I had to make a guess, from the numbers I showed you before, I'd say around 5% of visitors post-bounce 2011-11-23 10:32:38+0000 <antoviaque> (20%/4 since we've been open for 4 weeks) 2011-11-23 10:32:42+0000 <dachary> ok. So for 100 users there would be 5 users staying, 200 users => 10 users staying, 300 => 15 etc. 2011-11-23 10:33:06+0000 <antoviaque> yep - but it's a blind guess 2011-11-23 10:33:13+0000 <dachary> it helps anyway 2011-11-23 10:34:17+0000 <antoviaque> and actually post-bounce is not necessarily users - it's likely to be since there isn't much to do besides login in, but there could be a difference there too 2011-11-23 10:34:50+0000 <dachary> you mentionned there was a problem related to the fact that users come back but can't play because there are no players. How often do you figure this happens ? 2011-11-23 10:35:17+0000 <antoviaque> I could very easily calculate those things for sure if we had a "registration date", a "last login date" fields 2011-11-23 10:36:12+0000 <antoviaque> hmm, I'm really not sure, I think would say bullshit - I'd say maybe in 25-40%, but I'll probably be wrong 2011-11-23 10:36:42+0000 <antoviaque> tartarugafeliz may have a more precise idea than me on this 2011-11-23 10:38:05+0000 <tartarugafeliz> this happens all the time 2011-11-23 10:38:22+0000 <tartarugafeliz> I would say we loose 80% of the people that are interested enough to register because of it 2011-11-23 10:38:22+0000 <dachary> ok. I'll try to provide a meaningfull report for clarifying these questions. 2011-11-23 10:39:19+0000 <tartarugafeliz> but I'm really bad with percentages... 2011-11-23 10:39:31+0000 <tartarugafeliz> I never did this before, so its more my feeling 2011-11-23 10:39:50+0000 <dachary> it's helpful. I know what to look for now. 2011-11-23 10:40:32+0000 <antoviaque> there is something else too, that may make it closer to tartarugafeliz' numbers - recurring visitors who don't find anyone one of the days they come back, and drop out because of that 2011-11-23 10:40:49+0000 <tartarugafeliz> yeap 2011-11-23 10:41:21+0000 <antoviaque> there isn't a lot of games created while we're away (about 5-10/day), but those are more likely to be from recurring visitors, and they are much more precious than "new" visitors freshly coming from SU 2011-11-23 10:41:32+0000 <tartarugafeliz> agreed 2011-11-23 11:01:32+0000 <antoviaque> I'm entering a few tickets about ideas I've had lately - tartarugafeliz: what do you think of http://tickets.farsides.com/issues/645 ? would you like that? would the other players like it? : ) 2011-11-23 11:01:44+0000 <tartarugafeliz> let me see 2011-11-23 11:02:27+0000 <tartarugafeliz> its really cool! I bet they would like that 2011-11-23 11:02:34+0000 <antoviaque> ok : ) 2011-11-23 11:02:44+0000 <tartarugafeliz> but we also have to think about what to do with this points later on, I think⦠no? 2011-11-23 11:08:49+0000 MCMic (MCMic) is now online 2011-11-23 11:10:55+0000 <dachary> antoviaque: the change in user names makes distorts the stats a little 2011-11-23 11:11:05+0000 <dachary> s/makes// 2011-11-23 11:11:12+0000 <dachary> email => user 2011-11-23 11:12:07+0000 <antoviaque> tartarugafeliz: what to do with them later on? 2011-11-23 11:12:21+0000 <antoviaque> dachary: ah, true... 2011-11-23 11:13:11+0000 <tartarugafeliz> I mean, why would I want points? 2011-11-23 11:13:53+0000 <tartarugafeliz> to get something on the game right? 2011-11-23 11:13:53+0000 <tartarugafeliz> not just to be on the rank, because personally I really don't care about rankings 2011-11-23 11:13:59+0000 <tartarugafeliz> what I care is if I'm going to have lets say⦠different types of hands I can use on the box with my names... 2011-11-23 11:14:10+0000 <tartarugafeliz> or cards that only I have... 2011-11-23 11:14:16+0000 <tartarugafeliz> this one is more complicated I guess 2011-11-23 11:14:24+0000 <tartarugafeliz> but I think we need to give rewards 2011-11-23 11:14:54+0000 <tartarugafeliz> what do you think? 2011-11-23 11:16:06+0000 <antoviaque> tartarugafeliz: ah yep of course - I have some ideas about that too (some of the other tickets are about this too : ) ) 2011-11-23 11:16:38+0000 <antoviaque> but in any case it will be good to do things gradually - the points have the advantage of being able to track the progression of a player 2011-11-23 11:16:42+0000 <dachary> I compute there are 919 new users registered since october 12th. Does this match your gut feeling ? 2011-11-23 11:17:48+0000 <antoviaque> getting it right allows to build things on a good ground 2011-11-23 11:17:49+0000 <antoviaque> so can be worth thinking about this well - I tried to make a repartition of points that rewards people for creating interesting games, but it will need some more thoughts 2011-11-23 11:18:32+0000 <antoviaque> dachary: yep, 923 user accounts 2011-11-23 11:18:37+0000 <antoviaque> http://cardstories.org/admin/auth/user/ 2011-11-23 11:19:06+0000 <dachary> ok, I'm on the right track ;-) 2011-11-23 11:19:27+0000 <dachary> 295 users played at least one game 2011-11-23 11:19:41+0000 <dachary> (either as an author or as a player ) 2011-11-23 11:19:42+0000 <antoviaque> dachary: gave you access to the django admin btw 2011-11-23 11:20:17+0000 <antoviaque> (with your cs.org account) 2011-11-23 11:20:46+0000 <tartarugafeliz> its a lot of people! 2011-11-23 11:21:08+0000 <antoviaque> that's a bad percentage, we're probably doing something wrong if they login but stop short of their first game in 70% of the cases 2011-11-23 11:21:44+0000 <tartarugafeliz> its probably the fact that nobody is there to play with them 2011-11-23 11:21:55+0000 <antoviaque> no, it means they don't even write a story 2011-11-23 11:22:00+0000 <tartarugafeliz> ohhhh 2011-11-23 11:22:55+0000 <antoviaque> tartarugafeliz: probably the issue with the input field has probably some responsibility in this 2011-11-23 11:23:05+0000 <tartarugafeliz> damn bubble 2011-11-23 11:23:08+0000 <antoviaque> http://tickets.farsides.com/issues/613 2011-11-23 11:23:25+0000 <antoviaque> I've seen it in a good % of the cases at la cite des sciences 2011-11-23 11:23:47+0000 <tartarugafeliz> Yeap, I think I have to redesign this 2011-11-23 11:24:33+0000 <antoviaque> tartarugafeliz: can you think of more or better ways to improve things on #613? 2011-11-23 11:25:56+0000 <tartarugafeliz> yes 2011-11-23 11:25:59+0000 <tartarugafeliz> I'll do that 2011-11-23 11:27:07+0000 <antoviaque> thx : ) 2011-11-23 11:27:32+0000 <dachary> 548 users created a game 2011-11-23 11:27:56+0000 <dachary> side node : I will consolidate these findings in a proper report, no worries ;-) 2011-11-23 11:27:59+0000 <antoviaque> put it in the comments, this way we can try to make sure we've thought about it exhaustively 2011-11-23 11:28:17+0000 <antoviaque> dachary: ahhh - "played a game" is completed a game? 2011-11-23 11:30:02+0000 <antoviaque> or is it played as player? 2011-11-23 11:30:18+0000 <dachary> antoviaque: I consider a game played (as opposed to created) if it makes it to the voting phase 2011-11-23 11:30:25+0000 <antoviaque> I see 2011-11-23 11:30:56+0000 <antoviaque> ok, so it's not that bad, but it still leaves a lot of room for improvements 2011-11-23 11:31:09+0000 <dachary> 1464 games were created and 958 were completed 2011-11-23 11:41:48+0000 <dachary> the 958 games were completed by 216 users 2011-11-23 11:42:06+0000 <dachary> the 958 games were completed by 216 authors 2011-11-23 11:45:17+0000 <dachary> I'll keep digging after lunch. 2011-11-23 11:46:53+0000 <dachary> I'll try to have a report ready around 5pm so that we can discuss / criticize it together. antoviaque will you be available ? I hope it won't take more than one hour of your time. It's easy to get to false conclusions when writing this alone and your input will make all the difference. 2011-11-23 11:47:27+0000 <antoviaque> dachary: np, I should be there : ) 2011-11-23 11:47:30+0000 <antoviaque> bon ap : ) 2011-11-23 11:50:53+0000 <dachary> danke 2011-11-23 11:51:27+0000 <dachary> antoviaque: how much was spent on advertisement since october 12th ? roughly 2011-11-23 11:51:34+0000 <antoviaque> let me see 2011-11-23 11:52:54+0000 <dachary> the average number of games played by player is 2. But the the median is 1. Primarily because 30% of the games were played by 10 users. This accounts for the fanatic presence of tartarugafeliz ;-) 2011-11-23 11:53:09+0000 <tartarugafeliz> hahahahaah 2011-11-23 11:53:15+0000 <tartarugafeliz> I played 1000 games 2011-11-23 11:53:18+0000 <antoviaque> dachary: $560 2011-11-23 11:53:56+0000 <dachary> antoviaque: ok 2011-11-23 11:54:57+0000 <antoviaque> ah plus adwords and pikaland, so actually a bit more than this (but pikaland is ongoing this week, and adwords had very very low conversion, only lasted 2-3 days) 2011-11-23 11:54:57+0000 <antoviaque> pikaland = $70 right tartarugafeliz ? 2011-11-23 11:55:20+0000 <tartarugafeliz> 80 actually :) 2011-11-23 11:55:31+0000 <dachary> that's roughly 60ct per user. 2011-11-23 11:56:23+0000 <dachary> bbl, lunch 2011-11-23 13:33:58+0000 <antoviaque> meeting time : ) 2011-11-23 13:34:02+0000 <antoviaque> eve__: start dailymeeting 2011-11-23 13:34:02+0000 <eve> antoviaque: Sorry, meeting in progress. 2011-11-23 13:34:05+0000 <antoviaque> raah 2011-11-23 13:34:10+0000 <antoviaque> eve__: end dailymeeting 2011-11-23 13:34:10+0000 <eve> Meeting Ended 2011-11-23 13:34:10+0000 <eve> Minutes available at http://paste.pocoo.org/show/511868 2011-11-23 13:34:15+0000 <antoviaque> eve__: start dailymeeting 2011-11-23 13:34:15+0000 * eve antoviaque arbrandes blanchard blanchard1 dachary mtyaka tartarugafeliz and al: daily meeting starts! Please say "eve_: yesterday|today|blockers <your report>" 2011-11-23 13:34:29+0000 <antoviaque> eve: yesterday Tested alpha13, pushed to live, removed one of the commits which was crashing the server, played with players (good affluence yesterday thanks to pikaland : ) ), discussed stats with dachary, worked on two stories for brainstorming future features (#645 Players earn points for each game they played, #646 Players join "tables" when they arrive in the game - last one still in progress) 2011-11-23 13:34:29+0000 <eve> Done: Tested alpha13, pushed to live, removed one of the commits which was crashing the server, played with players (good affluence yesterday thanks to pikaland : ) ), discussed stats with dachary, worked on two stories for brainstorming future features (#645 Players earn points for each game they played, #646 Players join "tables" when they arrive in the game - last one still in progress) 2011-11-23 13:34:29+0000 <antoviaque> eve: today Finish specifying #646, work on specifying couple more stories that I thought about lately, more discussions on stats with dachary, work on Plebia as time allows 2011-11-23 13:34:29+0000 <eve> Next: Finish specifying #646, work on specifying couple more stories that I thought about lately, more discussions on stats with dachary, work on Plebia as time allows 2011-11-23 13:34:29+0000 <antoviaque> eve: blockers None 2011-11-23 13:34:29+0000 <eve> Blockers: None 2011-11-23 13:34:30+0000 <tartarugafeliz> hellooo!!! 2011-11-23 13:34:31+0000 <blanchard> hello 2011-11-23 13:34:34+0000 <antoviaque> heya : ) 2011-11-23 13:34:43+0000 <tartarugafeliz> I loved alpha13!!! 2011-11-23 13:34:57+0000 <antoviaque> yep, very cool : ) 2011-11-23 13:35:53+0000 <tartarugafeliz> eve: yesterday I worked on Pikaland's campaign and played the whole day and night. (I'm kind of addicted, I think) 2011-11-23 13:35:53+0000 <eve> Done: I worked on Pikaland's campaign and played the whole day and night. (I'm kind of addicted, I think) 2011-11-23 13:35:53+0000 <tartarugafeliz> eve: today I'll think of more or better ways to improve things on #613 (Authors see a more obvious text area to enter their story) and I'll play CS. 2011-11-23 13:35:53+0000 <eve> Next: I'll think of more or better ways to improve things on #613 (Authors see a more obvious text area to enter their story) and I'll play CS 2011-11-23 13:35:53+0000 <tartarugafeliz> eve: blockers none 2011-11-23 13:35:54+0000 <eve> Blockers: none 2011-11-23 13:35:59+0000 <antoviaque> hehehe 2011-11-23 13:36:07+0000 <antoviaque> addiction is gooood :p 2011-11-23 13:36:11+0000 <tartarugafeliz> hhahaha 2011-11-23 13:36:13+0000 <blanchard> tartarugafeliz: :) 2011-11-23 13:36:17+0000 <tartarugafeliz> sometimes I dream that I'm playing 2011-11-23 13:36:29+0000 <antoviaque> ah, you cheat, you train during your sleep!! 2011-11-23 13:36:32+0000 <antoviaque> that's why 2011-11-23 13:36:33+0000 <tartarugafeliz> hahahaahahah 2011-11-23 13:36:37+0000 <blanchard> eve, yesterday : fight with corporate stuff (rapport d'AGE, cession d'actions, etc), schedule BA pitch on December 7th, prepared contract number 2 for Felix. 2011-11-23 13:36:37+0000 <eve> Done: : fight with corporate stuff (rapport d'AGE, cession d'actions, etc), schedule BA pitch on December 7th, prepared contract number 2 for Felix 2011-11-23 13:36:37+0000 <blanchard> eve, today : not much (freelance work), more corporate if time allows. Will play cardstories, it has been some time since i played 2011-11-23 13:36:38+0000 <eve> Next: : not much (freelance work), more corporate if time allows. Will play cardstories, it has been some time since i played 2011-11-23 13:36:38+0000 <blanchard> eve, blockers : none 2011-11-23 13:36:38+0000 <eve> Blockers: : none 2011-11-23 13:37:08+0000 <antoviaque> good luck with that blanchard : ) 2011-11-23 13:37:24+0000 <blanchard> thx :) 2011-11-23 13:39:14+0000 mtyaka (mtyaka) is now online 2011-11-23 13:39:34+0000 <antoviaque> all good with felix btw? 2011-11-23 13:39:34+0000 <antoviaque> blanchard: ah, also! if I remember correctly it's ok for sending a proposal to arbrandes right? 2011-11-23 13:39:43+0000 * antoviaque had lag again again again 2011-11-23 13:39:49+0000 <tartarugafeliz> blanchard: its cool with the 2 new cards! :D 2011-11-23 13:39:51+0000 * antoviaque looks up another IRC client 2011-11-23 13:39:58+0000 <tartarugafeliz> hi mtyaka!! XD 2011-11-23 13:40:09+0000 <blanchard> tartarugafeliz: i'm going to discover the new card :) 2011-11-23 13:40:14+0000 <mtyaka> hello :) 2011-11-23 13:40:25+0000 <tartarugafeliz> hahahaha 2011-11-23 13:40:29+0000 <blanchard> antoviaque: Felix, yep, it's on my side i need to send him the contract 2011-11-23 13:40:35+0000 <antoviaque> hey mtyaka! 2011-11-23 13:40:36+0000 <blanchard> hi mtyaka :) 2011-11-23 13:40:47+0000 <antoviaque> ok 2011-11-23 13:41:47+0000 <blanchard> antoviaque: re arbrandes : I had CNC on the phone this morning, we'll have official answer within 15 days. Let me check the forecast again, i don't remember what we said 2011-11-23 13:42:28+0000 <antoviaque> I think you said it was ok in both cases, but with CNC longer 2011-11-23 13:43:00+0000 <antoviaque> Since december is approaching, would be good to send him something, even minimal, now 2011-11-23 13:43:59+0000 <blanchard> antoviaque: you're correct 2011-11-23 13:44:25+0000 <antoviaque> What's the short option? full time january? 2011-11-23 13:44:49+0000 <blanchard> without cnc, jan is ok with arbrandes full time (but not after jan) 2011-11-23 13:44:54+0000 <antoviaque> ok 2011-11-23 13:44:57+0000 <antoviaque> and with CNC? 2011-11-23 13:45:24+0000 <blanchard> hmmm 2011-11-23 13:45:34+0000 <blanchard> wait i see a mistake in the forecast :( 2011-11-23 13:45:48+0000 <blanchard> i'll have a look at that and get back to you after the meeting 2011-11-23 13:45:53+0000 <blanchard> antoviaque: ^ 2011-11-23 13:45:55+0000 <antoviaque> ok 2011-11-23 13:46:33+0000 <antoviaque> I'll send him an email to explain a bit both options to him if ok with you, to get a feel from his side - we haven't talked for weeks, so it would be good imho 2011-11-23 13:46:50+0000 <antoviaque> mtyaka: next? : ) 2011-11-23 13:46:56+0000 <antoviaque> dachary: still eating? :p 2011-11-23 13:47:04+0000 <dachary> darn 2011-11-23 13:47:06+0000 <dachary> sorry 2011-11-23 13:47:08+0000 <antoviaque> ahaha 2011-11-23 13:47:10+0000 * dachary runs 2011-11-23 13:47:13+0000 <mtyaka> eve: yesterday Finished #6587 (unicodeencodeerror, second try), finished #554 (simple text change author -> game master), thought about #619 (handle players with same names). 2011-11-23 13:47:13+0000 <eve> Done: Finished #6587 (unicodeencodeerror, second try), finished #554 (simple text change author -> game master), thought about #619 (handle players with same names) 2011-11-23 13:47:15+0000 <mtyaka> eve: today If I have time, work on #619. 2011-11-23 13:47:15+0000 <eve> Next: If I have time, work on #619 2011-11-23 13:47:17+0000 <mtyaka> eve: blockers None 2011-11-23 13:47:17+0000 <eve> Blockers: None 2011-11-23 13:47:18+0000 <mtyaka> :) 2011-11-23 13:47:25+0000 <dachary> eve__: yesterday nothing 2011-11-23 13:47:25+0000 <eve> Done: nothing 2011-11-23 13:48:07+0000 <dachary> eve__: today analyze the game logs and owa to produce a report about the audience of the past month and define metrics. 2011-11-23 13:48:07+0000 <eve> Next: analyze the game logs and owa to produce a report about the audience of the past month and define metrics 2011-11-23 13:48:14+0000 <dachary> eve__: blockers none 2011-11-23 13:48:14+0000 <eve> Blockers: none 2011-11-23 13:48:46+0000 * dachary reading the backlog 2011-11-23 13:49:22+0000 <antoviaque> mtyaka: I saw kickly on the backlog from yesterday that it was tricky to solve #619 right? 2011-11-23 13:49:49+0000 <antoviaque> dachary: will be very cool to have those, the numbers you gave this morning are already making me think : ) 2011-11-23 13:50:01+0000 <mtyaka> I don't think it will be very tricky, but it will require some changes on the server side 2011-11-23 13:50:33+0000 <antoviaque> mtyaka: ok - thanks for fixing all those bugs btw : ) 2011-11-23 13:50:49+0000 <dachary> (still reading backlog) 2011-11-23 13:50:50+0000 <mtyaka> heh, well, I hope they are fixed for real this time 2011-11-23 13:51:01+0000 <antoviaque> ahahah - they eventually will :p 2011-11-23 13:52:53+0000 <dachary> done 2011-11-23 13:53:01+0000 <antoviaque> gg :) 2011-11-23 13:54:08+0000 <antoviaque> something else for today? 2011-11-23 13:55:10+0000 <dachary> nope 2011-11-23 13:55:14+0000 <mtyaka> nope 2011-11-23 13:55:27+0000 <antoviaque> okidok - thanks! 2011-11-23 13:55:31+0000 <antoviaque> eve__: end dailymeeting 2011-11-23 13:55:32+0000 <eve> Meeting Ended 2011-11-23 13:55:32+0000 <eve> Minutes available at http://paste.pocoo.org/show/511872 2011-11-23 13:59:53+0000 <antoviaque> blanchard: sorted out the issue with the financials? : ) 2011-11-23 14:00:26+0000 <blanchard> antoviaque: there was a dumb mistake in the forecast. My apologies. We're too short in Jan with our assumptions if no CNC (-4kâ¬). If we have CNC, we can offer Adolfo at least Jan and Feb. CNC is 99% sure but there is always a risk. 2011-11-23 14:00:44+0000 <blanchard> hmmm 2011-11-23 14:00:56+0000 <antoviaque> what could be the 1% risk? 2011-11-23 14:01:07+0000 <blanchard> i agree with the fact that it would be good to touch base with Adolfo though 2011-11-23 14:01:34+0000 <blanchard> 1% is le ministère de l'industrie who cofunds and hasn't signed yet. 2011-11-23 14:02:04+0000 <blanchard> But i don't see them refusing the CNC budget given that it was pre-agreed. But who knows. 2011-11-23 14:02:50+0000 <blanchard> do you think you could touch base with Adolfo and confirm in 15 days from now on the duration ? 2011-11-23 14:03:09+0000 <blanchard> Or do you prefer to wait 2011-11-23 14:03:17+0000 <blanchard> antoviaque: ^ 2011-11-23 14:03:19+0000 <antoviaque> I could say something like: "We've got the funding to garantee 2 months of your full time in Jan & Feb - we're still waiting for the written confirmation so I'll confirm in two weeks, but wanted to check with you" 2011-11-23 14:03:30+0000 <blanchard> seems good to me 2011-11-23 14:03:37+0000 <antoviaque> yep, it's pretty much what you said too - ok : ) 2011-11-23 14:03:46+0000 <blanchard> I'm sorry for the mistake on the forecast antoviaque 2011-11-23 14:04:01+0000 <blanchard> I'll upload the correct version 2011-11-23 14:04:03+0000 <antoviaque> np - what's the visibility then with the mistake corrected? 2011-11-23 14:04:08+0000 <antoviaque> (with cnc) 2011-11-23 14:04:13+0000 <antoviaque> March? 2011-11-23 14:04:14+0000 <blanchard> march. 2011-11-23 14:04:17+0000 <antoviaque> ok 2011-11-23 14:04:31+0000 <blanchard> Not sure we can have Matjaz/Loic/Adolfo in march though 2011-11-23 14:04:42+0000 <blanchard> but until Feb no problem 2011-11-23 14:04:48+0000 <blanchard> antoviaque: ^ 2011-11-23 14:04:51+0000 <antoviaque> yep, it was beginning of April, it becomes beginning of March, understood 2011-11-23 14:05:26+0000 <antoviaque> ok, sending an email (I'll copy bar@) 2011-11-23 14:53:03+0000 <dachary> antoviaque: Considering the progression of the user base on a daily basis, I find that the returning users do not progress. 2011-11-23 14:53:15+0000 <dachary> I wonder if I'm not missing something. 2011-11-23 14:53:45+0000 <antoviaque> you mean there are as many new recurring players as there are churning players? 2011-11-23 14:53:50+0000 <dachary> http://dachary.org/wp-uploads/2011/11/users-output.txt 2011-11-23 14:54:30+0000 <dachary> around october 27 there are more recurring / returning players (la villette) 2011-11-23 14:54:58+0000 <dachary> but the rest of the time it does not vary in proportion of the user base 2011-11-23 14:55:32+0000 <antoviaque> could be that yep, that we don't keep players long enough for the number of active players to grow 2011-11-23 14:56:16+0000 <dachary> A reccuring user is a user who comes to the game (even if he does not play) after the first day (he is counted new the first day he arrives) 2011-11-23 14:56:58+0000 <dachary> antoviaque: does this definition make sense to you ? 2011-11-23 14:57:05+0000 <antoviaque> yep 2011-11-23 14:57:48+0000 <dachary> OWA sucks at re-injecting data. Because it so much depends on javascript and is not log based. 2011-11-23 14:58:10+0000 <antoviaque> one way to confirm would be to try to see, week arrival week day of arrival, what's the average length of last connection - first connection 2011-11-23 14:58:39+0000 <dachary> could you expand on this ? 2011-11-23 14:58:42+0000 <antoviaque> *week of arrival by week of arrival 2011-11-23 15:00:15+0000 <antoviaque> ie try to see for how long people keep coming back to the game 2011-11-23 15:02:22+0000 <antoviaque> if there is only very few players who stay more than 1-2 weeks, it would explain why it does not progress 2011-11-23 15:09:16+0000 <dachary> I'm on it 2011-11-23 15:54:40+0000 <dachary> I'll be a few minutes late, done in 20 minutes 2011-11-23 16:06:14+0000 <dachary> antoviaque: http://dachary.org/wp-uploads/2011/11/users-spread.txt 2011-11-23 16:06:25+0000 * antoviaque looking 2011-11-23 16:06:57+0000 <dachary> http://dachary.org/?p=1041 2011-11-23 16:07:13+0000 <dachary> antoviaque: I'm ready to discuss the first draft 2011-11-23 16:07:57+0000 <dachary> antoviaque: SPAN is the difference in days between the first visit and the last 2011-11-23 16:08:13+0000 <dachary> DAY is the number of days the user came to the game 2011-11-23 16:08:47+0000 <antoviaque> dachary: what's the definition of span & days? 2011-11-23 16:09:14+0000 * dachary needs to obfuscate the mails 2011-11-23 16:09:27+0000 <antoviaque> yup 2011-11-23 16:09:43+0000 <antoviaque> ok 2011-11-23 16:11:34+0000 <dachary> obfuscated 2011-11-23 16:12:31+0000 <dachary> I'm surprised the span increases after a player returns more than 5 days 2011-11-23 16:12:48+0000 <dachary> does that make sense to you ? 2011-11-23 16:13:57+0000 <dachary> I could explain it as : "if someone comes back 5 days, it is more likely to come back again after a few days. he is a recurring player. As opposed to people who play a few days (at most 5) in a row and never return." 2011-11-23 16:14:06+0000 <dachary> but it's a wild guess on my part ;-) 2011-11-23 16:14:47+0000 <dachary> antoviaque: are you reading the draft ? If you do I'll get a coffee and expect your questions & wtf ? ;-) 2011-11-23 16:15:34+0000 * antoviaque reading 2011-11-23 16:15:43+0000 * dachary coffee 2011-11-23 16:15:44+0000 <antoviaque> the draft? 2011-11-23 16:15:50+0000 <dachary> antoviaque: http://dachary.org/?p=1041 2011-11-23 16:15:52+0000 <antoviaque> ok 2011-11-23 16:15:59+0000 <antoviaque> go get you coffee yep ;p 2011-11-23 16:16:10+0000 <dachary> and a delicious macaron 2011-11-23 16:16:22+0000 <dachary> there are only a few left 2011-11-23 16:17:30+0000 <antoviaque> dachary: "958 were completed ( grep -c action=complete )" => beware, there is a timer for a little bit more than a week, that can get games to complete even without action from the GM 2011-11-23 16:18:19+0000 <dachary> that won't make a log entry though 2011-11-23 16:18:33+0000 <dachary> ok 2011-11-23 16:18:46+0000 <dachary> do you think it has a significant impact ? 2011-11-23 16:21:30+0000 <antoviaque> Probably not 2011-11-23 16:22:13+0000 <antoviaque> it probably affects a bit the statistics, but since it's only for a week and usually indicates a player who has left, it's not worth redoing imho 2011-11-23 16:22:15+0000 <dachary> The approach was to try to understand the dynamic of the users in the past 5 weeks. This was driven by questions and each time they popped a small script was made to compute it. 2011-11-23 16:22:50+0000 <dachary> Reusing owa for this would have been a self defeating strategy. 2011-11-23 16:23:13+0000 <dachary> And if the current report can be made to be sensible, it will define how to configure owa. 2011-11-23 16:23:51+0000 <dachary> What is actually needed for you to understand what happens. 2011-11-23 16:24:03+0000 <dachary> Which may or may not be what owa provides. 2011-11-23 16:24:36+0000 <antoviaque> "No player actually start completing a game as an author. Their first game action is always pick (indicating a player) and not vote (indicating a game author moving to the vote phase)." => not sure what you mean 2011-11-23 16:24:36+0000 <antoviaque> that would be very surprising 2011-11-23 16:24:36+0000 <antoviaque> You mean that the first action they do in the game is always to finish a game as a player, never to finish a game as an author? 2011-11-23 16:24:37+0000 * antoviaque reading a first time quickly to get a feel, then I'll re-read more thoroughly 2011-11-23 16:31:08+0000 <antoviaque> Yep, the global view matches the observations in game - ie a small proportion of users keep coming back regularly 2011-11-23 16:32:01+0000 <dachary> You mean that the first action they do in the game is always to finish a game as a player, never to finish a game as an author? : yes 2011-11-23 16:32:45+0000 <dachary> If you find it surprising, I will double check. 2011-11-23 16:32:50+0000 <antoviaque> This would be very very surprising, it contradicts what we see in game - players start as authors 2011-11-23 16:33:04+0000 <antoviaque> and then with a little help, they start joining other's games 2011-11-23 16:33:20+0000 <dachary> Probably a mistake somewhere then. 2011-11-23 16:34:09+0000 * antoviaque going back to the report to try to get a better view 2011-11-23 16:34:39+0000 <dachary> antoviaque: after a quick read, do you miss something important ? first impression ;-) 2011-11-23 16:35:07+0000 <antoviaque> hmm, yes, but I'm not sure what yet 2011-11-23 16:38:12+0000 * antoviaque opening a spreasdheet to put all the numbers in there : ) 2011-11-23 16:43:10+0000 <antoviaque> "The games were created by 548 different players but only 216 different players completed games" => 216 is the number of authors who created a game that was completed, or the number of players who played someone else's game to the end? 2011-11-23 16:43:27+0000 <dachary> number of authors 2011-11-23 16:43:30+0000 <antoviaque> ok 2011-11-23 16:43:32+0000 <dachary> just authors there 2011-11-23 16:44:21+0000 <dachary> antoviaque: I'm sure you'll find inconsistencies in the figures. This gets confusing after a while. 2011-11-23 16:44:57+0000 <antoviaque> yep, stats are full of little details that change everything : ) 2011-11-23 16:45:01+0000 <dachary> But it shows a few fundamental things that will hopefully help understand the current status. 2011-11-23 16:46:48+0000 <tartarugafeliz> Grace: can you add people? - this player just asked this question 2011-11-23 16:47:51+0000 <dachary> add people ? 2011-11-23 16:48:45+0000 <tartarugafeliz> wait give me just 5 mint to finish this game and I'll explain 2011-11-23 16:49:21+0000 <antoviaque> "The total number of players or lurkers is 919" => this includes authors right? (ie those who never played as player, but only as author) 2011-11-23 16:50:17+0000 <dachary> antoviaque: yes. This includes all users who displayed the page of the game (i.e. issued a action=state call to the game server). This is the most inclusive figure of all. 2011-11-23 16:51:58+0000 <antoviaque> ok, should be roughly equal to the number of users, which fits 2011-11-23 16:53:08+0000 <dachary> yes. It would be wrong if it was *more* 2011-11-23 16:53:10+0000 <dachary> ;-) 2011-11-23 16:53:38+0000 <tartarugafeliz> so!!! sorry! the situation here was: 3 players loved to play between them 2011-11-23 16:53:49+0000 <tartarugafeliz> but they don't know each other, so they wanted to add each other on the game 2011-11-23 16:53:54+0000 <antoviaque> dachary: yep, it's less ;p 2011-11-23 16:54:03+0000 <tartarugafeliz> so they could come and play between them in the future. 2011-11-23 16:54:05+0000 <tartarugafeliz> that's it. 2011-11-23 16:54:22+0000 <antoviaque> tartarugafeliz: I've started specifying http://tickets.farsides.com/issues/646 to solve this : ) 2011-11-23 16:54:26+0000 <tartarugafeliz> social network right? I think its becoming something that people expect everywhere 2011-11-23 16:54:56+0000 <antoviaque> see if you think it would help 2011-11-23 16:54:57+0000 <tartarugafeliz> ahh that's really smart 2011-11-23 16:58:32+0000 <antoviaque> tartarugafeliz: the devil is in the details - it's tricky to get right. If you have time to review it & comment, I'd be happy to hear about it 2011-11-23 16:59:00+0000 <tartarugafeliz> I'm going to read it with atention later on but from what I've seen its really cool 2011-11-23 16:59:05+0000 <tartarugafeliz> I think they will like it 2011-11-23 17:00:52+0000 <antoviaque> dachary: still putting together my spreadsheet - it get much clearer as I do this : ) 2011-11-23 17:01:48+0000 <dachary> antoviaque: it is a very good method to manipulate the data to get familiar with it ;-) 2011-11-23 17:05:26+0000 <antoviaque> dachary: "The total number of players who played at least a game, either as an author or as a player is 295" => is it people who *completed* a game, either as an author or as a player, right? 2011-11-23 17:06:31+0000 <dachary> no 2011-11-23 17:06:37+0000 <dachary> let me check 2011-11-23 17:08:59+0000 <dachary> dam 2011-11-23 17:09:02+0000 <dachary> this number is wrong 2011-11-23 17:09:04+0000 <dachary> dam 2011-11-23 17:13:03+0000 <dachary> antoviaque: nice catch, fixing 2011-11-23 17:13:17+0000 <dachary> this will improve the figures ;-) 2011-11-23 17:13:43+0000 <antoviaque> :) 2011-11-23 17:14:57+0000 <dachary> you can change 295 by 363 in your spreadsheet 2011-11-23 17:15:05+0000 <dachary> i'm updating the post 2011-11-23 17:18:09+0000 <dachary> the error came from a typo : vote instead of voting... 2011-11-23 17:18:34+0000 <dachary> antoviaque: 2011-11-23 17:19:28+0000 <antoviaque> ok, thanks : ) 2011-11-23 17:22:18+0000 <antoviaque> so it means that 147 players completed a game as player, but never as an author? (40% of all the people who completed a game) 2011-11-23 17:23:37+0000 <antoviaque> pretty high number, interesting 2011-11-23 17:27:39+0000 <antoviaque> ah, I know what I was feeling I was missing - how old the player is in the recurring players list 2011-11-23 17:28:19+0000 <antoviaque> the span gives an indication, but there is no way to really differentiate between players who arrived 2 days ago from the ones who arrived one month ago 2011-11-23 17:29:23+0000 <antoviaque> ie if the player arrived one month ago and has a span of 2 days, we know it has churned 2011-11-23 17:29:30+0000 <antoviaque> if he arrived 2 days ago, he's active 2011-11-23 17:29:51+0000 <antoviaque> dachary: ^ 2011-11-23 17:31:09+0000 <antoviaque> fyi, it gives: 2011-11-23 17:31:44+0000 <antoviaque> http://paste.debian.net/146752/ 2011-11-23 17:33:36+0000 <dachary> if a player has not connected in a month you consider it lost ? 2011-11-23 17:36:35+0000 <antoviaque> yep 2011-11-23 17:37:08+0000 <antoviaque> some will come back later on though, and then they can be removed from the churn % 2011-11-23 17:37:23+0000 <dachary> http://paste.debian.net/146752/ I don't understand why it does not show higher numbers in span 2011-11-23 17:37:49+0000 <antoviaque> actually http://paste.debian.net/146754/ 2011-11-23 17:38:00+0000 <antoviaque> higher numbers? 2011-11-23 17:38:00+0000 <dachary> ah :-) 2011-11-23 17:38:22+0000 <dachary> (http://paste.debian.net/146752/ stops a span 13) 2011-11-23 17:38:56+0000 <antoviaque> yep, omited some on that first one : ) 2011-11-23 17:39:19+0000 <dachary> antoviaque: what is the third column ? 2011-11-23 17:39:27+0000 <dachary> ah 2011-11-23 17:39:29+0000 <dachary> I see 2011-11-23 17:40:32+0000 <dachary> I figured the period was not large enough to consider churn 2011-11-23 17:40:54+0000 <antoviaque> yes, it's a little small for that 2011-11-23 17:41:02+0000 <antoviaque> but it gives some good hints already 2011-11-23 17:41:30+0000 <dachary> and in any case, since the recurring players do not grow in proportion of the user base, there is no way to extrapolate. Something is not working. 2011-11-23 17:43:03+0000 <dachary> All other numbers show trends that can be improved (such as the conversion rate from "account created" to "played or authored a game" which is probably where there is a lot of room for improvement) 2011-11-23 17:44:18+0000 <dachary> But this is not working at all and is, I think, what the most important discovery. Unless proven wrong. Do you see anything that would prove this wrong ? 2011-11-23 17:44:37+0000 <dachary> It has such great consequences that it deserves triple check ;-) 2011-11-23 17:44:58+0000 <antoviaque> Hmm, I'm not so sure - it's tricky to get conclusions on the small number of recurring players left 2011-11-23 17:45:00+0000 <dachary> "this is not working at all" => players do not come back 2011-11-23 17:45:35+0000 <dachary> antoviaque: or the conclusion is that a very very small number of players come back 2011-11-23 17:45:48+0000 <dachary> so small that it does not show on 1000 users 2011-11-23 17:46:26+0000 * dachary takes another look at the figures 2011-11-23 17:46:43+0000 <dachary> antoviaque: would you agree to say that a players coming back more than 5 times seems to be recurring ? 2011-11-23 17:46:57+0000 <antoviaque> If you look at it, you have 39% of registered who finish a game, and 9.57% of registered who come back on another day 2011-11-23 17:46:58+0000 <dachary> in the sense that its span is largare ? 2011-11-23 17:47:13+0000 <dachary> larger 2011-11-23 17:47:20+0000 <antoviaque> so 25% of people who complete a game come back another day 2011-11-23 17:48:51+0000 <dachary> but the RETURNING column here does not vary http://dachary.org/wp-uploads/2011/11/age-output.txt 2011-11-23 17:49:26+0000 * dachary has an idea 2011-11-23 17:49:44+0000 <antoviaque> yes, but that's where stats get nasty with devil in details 2011-11-23 17:50:24+0000 <dachary> considering the players coming back more than 5 days 2011-11-23 17:50:27+0000 <dachary> http://pastebin.com/ELeBeexa 2011-11-23 17:50:37+0000 <antoviaque> What it says for me is that players usually don't come back for more than a few days 2011-11-23 17:51:10+0000 <antoviaque> but having this "few days" is already quite big - they could not come back at all 2011-11-23 17:51:55+0000 <antoviaque> given that it's more of a minigame right now (ie no progression or retention features) 2011-11-23 17:54:12+0000 <dachary> antoviaque: out of this list http://pastebin.com/xNFA5v1E could you please remove those you know ? the idea is to keep only the users noone knows, those who were actually recruited 2011-11-23 17:55:06+0000 <antoviaque> http://pastebin.com/7nLVrEVu 2011-11-23 17:55:33+0000 <antoviaque> those are our current hardcore players : ) 2011-11-23 17:56:16+0000 <dachary> ;-) 2011-11-23 17:56:29+0000 <dachary> i'll be softer and call them recurring players 2011-11-23 17:56:45+0000 <antoviaque> hehe 2011-11-23 17:57:34+0000 <dachary> 10 out of 1000 is 0.1% . 2011-11-23 17:58:17+0000 <dachary> You have two populations 2011-11-23 17:58:22+0000 <dachary> 3 2011-11-23 17:58:50+0000 <dachary> a) players who play one day and leave, b) players who come back for a few days in a row, max 5, c) players who stay 2011-11-23 17:59:22+0000 <antoviaque> yep 2011-11-23 17:59:22+0000 <dachary> that's what I read from the stats, does this seem logical to you ? 2011-11-23 17:59:27+0000 <antoviaque> completely 2011-11-23 18:01:03+0000 <antoviaque> It's good to have this split, it allows to isolate one of the most important challenges, ie if we can convince people in b) to stay longer 2011-11-23 18:01:10+0000 <dachary> If I think in terms of revenue stream (I know it's a bit early for that ;-) I have weeks to seduce the players who stay into buying something. I have a few days to seduce the b) and there is no chance I'll get anything out of the first ;-) 2011-11-23 18:02:14+0000 <antoviaque> yup - but we probably wouldn't be able to monetize a game in the current state, retention is way too low 2011-11-23 18:04:22+0000 <antoviaque> or are you trying to say that retention can't be fixed? : ) 2011-11-23 18:30:45+0000 <dachary> I was saying that but I figured it's there, just too small to show on stats. 2011-11-23 18:31:09+0000 <antoviaque> dachary: still there? 2011-11-23 18:36:35+0000 <antoviaque> yep - personally I'm really happy to see that we have already some recurring players (ie people in category c) 2011-11-23 18:37:00+0000 <antoviaque> the difficulty now will be to see if we can make this group % bigger, but I think we have a few good cards to play (without bad jokes :p) 2011-11-23 18:37:33+0000 <dachary> yes, going over the data again 2011-11-23 18:37:43+0000 <antoviaque> ok 2011-11-23 18:37:50+0000 <dachary> it's the first time I face a real situation ;-) 2011-11-23 18:38:09+0000 <dachary> well, not the first time but I was not involved in analytics ;-) 2011-11-23 18:38:49+0000 <dachary> a user currently costs 60cts 2011-11-23 18:39:05+0000 <dachary> the recuring users are 1/1000 2011-11-23 18:39:22+0000 <dachary> they can *surely* be monetized 2011-11-23 18:39:33+0000 <dachary> and the income they generate must cover all the costs 2011-11-23 18:40:23+0000 <dachary> these players currently cost 60usd each 2011-11-23 18:40:27+0000 <dachary> roughly 2011-11-23 18:40:52+0000 <dachary> 1/100 recuring user not 1/1000 ;-) 2011-11-23 18:41:34+0000 <dachary> antoviaque: in the business model how much is the expected revenue per user ? 2011-11-23 18:42:48+0000 <antoviaque> hmm 2011-11-23 18:42:55+0000 <dachary> IIRC you can't expect more than $30 per user over its lifetime. 2011-11-23 18:43:07+0000 <antoviaque> I don't remember how much blanchard put - maybe he's around? 2011-11-23 18:43:25+0000 <antoviaque> dachary: yep, that's why I was saying that it's too low now 2011-11-23 18:43:51+0000 <dachary> how much is the expected conversion rate between a recurring user and a paying user ? 2011-11-23 18:44:48+0000 <antoviaque> around 5% 2011-11-23 18:44:58+0000 <antoviaque> but it changes over time of course 2011-11-23 18:52:38+0000 <dachary> antoviaque: is http://farsides.com/groups/business/ appropriate to post todays report ? 2011-11-23 18:53:37+0000 <antoviaque> dachary: yep, perfect one : ) 2011-11-23 19:01:03+0000 <dachary> Although today's work was not technical, I'm happy that it gave me a good visibility on how cardstories currently works. And I hope it will help you define priorities. 2011-11-23 19:47:19+0000 dachary (dachary) is now online 2011-11-23 19:52:08+0000 MCMic (MCMic) is now online 2011-11-23 19:54:07+0000 tartarugafeliz (tartarugafeliz) is now online 2011-11-23 21:06:12+0000 J-Mad (J-Mad) is now online 2011-11-23 21:58:42+0000 mtyaka_ (mtyaka_) is now online
_______________________________________________ Farsides mailing list - [email protected] Wiki: http://farsides.com/ List: http://farsides.com/ml/ Forum: http://farsides.com/forum/ Ideas: http://farsides.com/ideas/ Chat: http://farsides.com/chat/
