Would card prediction be accepted?

[anonymous555]

Just curious if this would be accepted or rejected before working on it.

1. For all the games archived in xml, construct counts of the following
   for each deck
   {
   for each turnnumber
   {
   for each enemy currentcard
   cardcounts[enemyid][turnnumber][priorcard][currentcard] ++
   }
   }

2. Store this away as xml

3. As you play, at the beginning of a turn
   Display the percentages and card names/graphics for the top n expected cards

cardcounts[enemyid][turnnumber][priorcard][predictedcard] / cardcounts[enemyid][turnnumber][priorcard].size

---

Or if you wanted even simpler, we could forget of the prior card, and just look at all cards typically played on a given turn.

[azeier]

This (or at least this direction) is something I would really like to add but honestly I'm not sure how accepted this would be by a lot of users. It does certainly tip the scales towards an "unfair advantage" quite a bit.

Also I think just using sheer numbers and not taking deck types / meta (changes, existing less in lower ranks, ...) into account will not yield any good results (in the long run).

[anonymous555]

Got this implemented

Nextcardpredict #274

[azeier]

http://www.elie.net/blog/hearthstone/i-am-a-legend-hacking-hearthstone-with-machine-learning-defcon-talk-wrap-up

Why we are not releasing our tool?
One thing you won’t see posted, however, is a software tool that we promised to release during our Defcon presentation. Following Defcon we had a series of conversations with the Hearthstone team about our research — apparently the email that I sent prior to Defcon didn’t reach the right person. They like our research on game/cards balance and are very enthusiastic and supportive about it.

On the other hand, they were very concerned that our real time dashboard that can predict your opponent’s deck will break the game balance by giving that person (that is, whoever has the tool) an unfair advantage. They also expressed concern that such a tool makes the game less fun by taking away some of the decision-making from the player. It was a difficult decision — I invested a lot of our time building our real-time dashboard tool with Celine — but we agree with the Hearthstone team and will not release the tool publicly.

So this might be taking it a step too far and blizzard could end up at least taking a stance on this.

[anonymous555]

Here's a text file with all the predictions. Judge for yourself.

https://github.com/anonymous555/Hearthstone-Deck-Tracker/blob/nextcardpredict/allpredictions.txt

[azeier]

I still think it's get a lot less useful after turn 1 without taking deck types / meta into account.

And the problem is, if we get it to a point where it does that it get's too good. So far one of the main points when reasoning why this tool is "okay" is that you can theoretically do it all on paper.
This.. not so much anymore.

Combining this with deck prediction (based on popular deck lists) would be probably extremely cool, but for another project :(.