Post by redonihunter in Adding actual analytics on Itch for our own releases

Viewing post in Adding actual analytics on Itch for our own releases

For indexing problems in general read here . But the rogue of nexus was last updated over 40 days ago. And none of the other games are indexed either. A game can get delisted temporarily after an update and waiting times can be over 40 days to get redindexed. But with several games I suspect other reasons. There is also no link on the publisher's homepage to Itch and they claim to be the developer on their homepage.

If you have split ways with the publisher you should look into the legal situaion if you still own your own game and how things are for publishing. While being a different magnitude, the developers of Pathfinder Kingmaker found out the hard way that they do not own their own game.

The Itch version appears #2 on internet search, so I assume it was indexed once or there are enough links to the games for search to find it. #4 is the youtube release video with 70 views.

A web version would not work on Steam of course. But it could be uploaded to other platforms or your own homepage. And you will see games that have in their settings check boxes for sending anonymous data. Also, I imagine you could collect certain data points about how the game is played without worrying about personal data protection laws. That's not gonna work for things where you connect the data to people.

Those kpi things you mentioned are for subscription based games. Especially the churn rate. You do not have a churn rate. Of course you could try to estimate how good your game is, by how long people play it. But you would need a lot of players to begin with to try extract why people like the game.

So my opinion about applying kpi for a game like rogue of nexus can be summarized by this short reddit from 2 years ago https://www.reddit.com/r/gamedev/comments/13y8tcy/what_would_be_potential_kpis_f...

I believe finding honest play testers for development and marketing the game to the target audience to be more helpful than looking for obscure data points that will only be usefull if you already have enough players to have statistics. It is the also the idea behind https://itch.io/board/255031/get-feedback , but there is not that many readers there. Finding unpaid honest testers is hard, so I understand the desire to replace that with data analysis.

The thing with the early death confuses me. I happen to play a roguelike here and there. It is part of the game mechanic. People of the target audience would not be deterred by the very game mechanic they liked and which made them try the game. They like difficulty. If there is a miscommunication and players think this to be a regular rpg and then they are confronted with the roguelike style restart, that's a recipe for disliking the game. Regular rpg do have bad ends and game over screens too, so it is not that obvious that dying here and there is the way the game is meant to be played.

Like Reply

Tenkarider56 days ago

Uhm, i see... reading about indexing, unless it actually is about a lack of link to itch in their site, all the rest seems to be regular. Also in the first days of their releases, if i was typing in the search bar of games the name of their released game then they would appear, after some time they stopped to appear. I let the publisher know, they asked to Itch about that one, we'll see if they can elaborate further on what happened.

I'd not call a separation, as a matter of fact they own the IP, i'm under contract and it will extend further after making the mobile version (not anytime soon), i also get royalties on it. Aside this current stagnant scenario, not that much has changed along the process.

So you mean like uploading the link to web version of itch game on your profile and possibly to other profile's platforms? Maybe this week end i'm gonna do a random test to know more about those data settings.

Ok maybe it could be the wrong growth engine i designated, need to make a double check but after all the point of what i'm doing with analytics is about learning and learning, figuring out the proper engine and the connented KPIs are among the information i have to learn, after all i'm just starting with that so there's a lot i have to understand.

Sales i think it's a vanity metric, despite clearly an high value vs low refund surely means there's some earnings. Anyway it's necessary to track also if they become stagnant over time; I guess play time and retention might be considered pretty much the same thing unless i'm missing an intrinsic difference; For the completion rate i think for the demo it could match with the Funnel i mentioned, for the full version instead, while for Steam probably achievements/Statistics are enough, i wonder how i can track that outside Steam; i guess i'm gonna keep an eye on that number of sessions as well.

Anyway, despite it's required that to be new users in order to be genuine datas, actually for A/B testing about 10 people is enough. While it's called A/B testing, it's not implying that the target of such experiments will be actual testers, they are just players. Unity for example had a A/B testing function which basically would target a designed % of random players among the total, and you could collect the differences among analytics... without such help unfortunately what i can do instead is simply swapping with an updated build and grasping the data differences among new users... obviously this workaround is much worse, that's why it's a shame that restriction on third parties. (it's a drammatically slower process, compared to that)

It's not like i'm trying to replace testers or actual people feedback: what i'm saying is that both should be integrated in the learning process. Also, any source of information serves a different purpose and allows to learn different information among them: for example meeting people on a fair event allows to listen both their feedbacks but also their unexpressed reactions (you need to notice them); analytics datas are more generic and KPI driven but i can grasp those datas from distance and potentially as a stable flow over time, without spending more time/resources than the required one to analyze such datas; maybe it's a tad late for me to talk about testers, but i think what i should search for is for the so called early adopters (players which stick with the game at an early stage of developement, despite being WIP and with flaws), i should have started several time ago, but perhaps what i'm missing is something like a dedicated Discord server for my game, in order to start making grow a community around it.

Once i finally manage to setup and planning for the analytics i'm gonna release an update and start to add more presence of my game here and there, i noticed that thread... anyway i'll seize any other opportunity, at the proper time.

Roguelite to be more precise, in my case. That scripted early death doesn't exist anymore, you lose only when you lose now. I think as i mentioned before that the problem was that such run end was called too much early: it takes more time to show extra interesting content of the game, but i must also consider to make that happen sooner, i'll evaluate that, and this is where analytics can help me in knowing if i'm doing it right. Said that, it 50% resembles a classic RPG, so i cannot exclude the chance there's a miscommunication of the genre, if there's a way to figure that out, i'll try to track that info as well

Like Reply

redonihunter56 days ago

So you mean like uploading the link to web version of itch game on your profile and possibly to other profile's platforms? Maybe this week end i'm gonna do a random test to know more about those data settings.

That game just looks like it would be playable in a web version. And a web version is a good tool to spread a demo version of the game. If you do that on the typical web game hosters (including Itch), you might get comments on those pages and gather some feedback. But it boils down to the old problem. You need to gather traffic and potential players, and that does not happen just by existing.

If you have enough traffic you can worry about collecting statistical data and I assume that is easier on a web version, since people typically need to be online to play anyway. You still should make sure that the data you collect is either agreed upon by the player or not relevant under personal data protection laws.

For that kpi and growth engine, better double check if that theory is even applicable. That theory stuff is for a different kind of games. You do not have retention. People might play the game for 10 hours and be done with it - in a good way. Retention is something for games that aim to have regular plays for months and sell their services. Like advertisements and microtransactions. Only for that you need that thing called retention. Sure, for some games a high game play time indicates that some people like playing the game a lot.

Roguelite to be more precise, in my case.

I do not really differenciate between those and both have nothing to do with the game Rogue anyway for a long time.

If you consider your game a roguelite, than actually, the early deaths and many deaths would be an essential game mechanic. The key difference between a -like and a -lite is the advancement by death. It is not unlike how you advance a clicker game incrementially by restarting and getting a bonus for that. In a rogue-lite you do the same, only the restart is a perma death of your old character. And in the meta-game, even a rogue-like has advancement since the player has more knowledge.

Communicating what your game is, is important. But unfortunately, you cannot rely on terms like rogue-like and rogue-lite for that. Most people do not know the difference or care. You better spell it out in the promotional material.

Like Reply

Tenkarider55 days ago

When you say gather traffic i guess you mean always the same thing required for any store page, right? Anyway what you said about the player agreeing ringed a bell, and in fact, after a double check, i found out that even in this case (web version) it falls under the GDPR, PIPL and CCPA laws so again from that perspective we're back to square one, with Itch. In any case it's worth to consider that option for gathering more feedbacks on itch.

Yeah, i'll check it out for KPI and stuff: afterall all that learning thing is about making assumptions and then prove them right or wrong with datas, metrics i think are more about knowing if you're doing it wrong, rather than the opposite... an anomaly in datas can also mean that something wrong is happening in a specific point of the game (or a specific point of the loop).

Mh, is that the case for the genre? On reddit there are 2 different groups about roguelites and roguelikes, the second one in particular won't allow to talk about roguelites, i guess the niches will care more about that... In any case i'll try to be more specific, thanks

Like Reply