You actually can cheese it. All you have to do is call a js function in the console after you make your edits, and it will regenerate the puzzle with the edited word list. I haven't tried multiple words in fields yet, that might break it. The only issue I had was it didn't cross off the changed word when I found it, but I figured that one out too. It just requires a bit more editing of the html.
I don't know how much changing words will help you cheat, since you still need to find the words. Maybe if you made every word the same, but I expect that to break it. In my case I just replaced Kneecap with Patella.
Edit: Multiple words with spaces in the fields equals solvable puzzle, but space squares render misaligned. Repeated words equals not solvable.
Edit 2:
Edit 3: And I've figured out how to adjust the grid size to make it even easier to find the one word. I promise not to actually cheat like this for future word searches. They aren't really that hard. Honestly, the effort required editing it like this probably isn't that much less than solving it legitimately.
You call it the stage 2 calculator, yet it appears to be for stage 1. It also took me several tries before I understood what the sequence stuff meant. Perhaps an example would've helped. Or had I known the calculator was for the first stage, I probably would've solved that a lot sooner.
I'm going to guess the text change to Monstrous Macaw is in case it gets a deathrattle buff to exclude it from triggering itself in that case. Would I be in the right ballpark?
Still need to do Illidan, but overall, pretty cool. I learned some things. After Book of Mercenaries is done, I think they should move on. We probably don't need a Book of Alternate Heroes or something. I guess there are a few we could learn more lore from, like Sylvanas.
Don't know about anyone else, but I'll pretty much definitely be converting my entire pack to dust. Except maybe if I get an epic or legendary.
I bought a couple bundles back in 2019/early 2020 myself. I've probably only spend around $100 ever on this game, and I haven't spend a cent in over a year. They still offered me the whale bundle.
It records at least everything necessary to make a replay. For example, any movement between deck, hand and board, so that should be useable to keep track of what's in your hand and what's in play. It's how the overlay works, and that can tell you things like what generated or drew a card in the other player's hand.
I recall there was an information leak. I think it was with C'Thun, the Shattered where the deck tracker would show you the other player had drawn a C'thun piece, so the deck tracker authors updated it to hide that information. I think there was some fear that Blizzard would ban people for using the tracker if it revealed stuff it shouldn't, which is technically against the rules, but I don't see how they could find out.
I believe detection of the mod could possibly be evaded. I know some anti-cheat mechanisms use signatures of known cheats, and Blizzard might be doing something more sophisticated, but that might not necessarily make evasion impossible. I had a look, it appears it uses DLL injection. I think botting would be more reliably detected based on behaviors like playing non-stop for over 24 hours with 0 breaks. Something I doubt any human would do.
There's also a logging function you can enable, which is how the deck tracker functions, and that records every move made, without having to access the game's memory space. I think actions are recorded as some kind of code, but most of the reverse engineering work has been done already, and the deck tracker is open source. If it isn't possible to write such a bot as a plugin for the deck tracker.
I think you might have misinterpreted me. I am less convinced this mod is necessary to make such bots, but I can see how being able to send keypresses would make programming a bot easier.
I guess you mean programming the bot to do its thing and ignoring anything the other player does. I didn't know any decks existed that could allow you to win often enough to climb all the way by playing that way. I suppose one could be written and debugged in 2 or 3 days, which is probably enough time before overpowered decks get nerfed.
As I sort of said earlier, I don't think Blizzard will be happy about third-party code being injected in their game. I'm not sure what they can do to stop the mod from working and prevent the author from just updating the mod. I guess they could detect it and ban users. Either way, if an account is playing for more than 24 hours straight with 0 breaks, it should be pretty obvious they are botting.
I haven't done much Windows development, could a bot not send mouse events to the game while it is in the background? Using the mod to send keyboard events would mean that it wouldn't have to know the precise position of UI elements.
It would also be able to get all the information it needs from the gameplay logs. Sure they default to disabled, but turning them on is pretty simple.
As for reaching Legend, even if the bot only knows how to play a single deck, how many situations would the bot need to know how to respond to in order to maintain a 50%+ winrate, especially after D5? I'm assuming your account didn't start with a 13 star bonus in this scenario.
Edit: I just remembered machine learning was a thing, like those AI that play SC2 at Grandmaster level. I don't know about reaching Legend in a couple of days given the amount of time that would probably be spent trying random shit before it learns what works, and all the RNG shit the game throws at you might be another layer of complexity.
Blizzard is probably going to kill this, and from a security perspective, it makes sense. Hopefully this will push them to add official support that doesn't rely on arbitrary code injection that could be easily exploited.
You actually can cheese it. All you have to do is call a js function in the console after you make your edits, and it will regenerate the puzzle with the edited word list. I haven't tried multiple words in fields yet, that might break it. The only issue I had was it didn't cross off the changed word when I found it, but I figured that one out too. It just requires a bit more editing of the html.
I don't know how much changing words will help you cheat, since you still need to find the words. Maybe if you made every word the same, but I expect that to break it. In my case I just replaced Kneecap with Patella.
Edit: Multiple words with spaces in the fields equals solvable puzzle, but space squares render misaligned. Repeated words equals not solvable.
Edit 2:
Edit 3: And I've figured out how to adjust the grid size to make it even easier to find the one word. I promise not to actually cheat like this for future word searches. They aren't really that hard. Honestly, the effort required editing it like this probably isn't that much less than solving it legitimately.
They've always been editable text fields? I thought he was messing around with the inspector.
I had to go to Google for that first one because there's no way to search flavor text on the site, and it wasn't anything I thought it might be.
Mobile?
I'm not seeing that. I'm seeing 3 months, 2 weeks though. Perhaps you missed the 3 months part?
You call it the stage 2 calculator, yet it appears to be for stage 1. It also took me several tries before I understood what the sequence stuff meant. Perhaps an example would've helped. Or had I known the calculator was for the first stage, I probably would've solved that a lot sooner.
I'm going to guess the text change to Monstrous Macaw is in case it gets a deathrattle buff to exclude it from triggering itself in that case. Would I be in the right ballpark?
Heh, so he created a Reno that turns into a dragon before we knew he was canonically a dragon.
I wondered if that 12 win requirement had been lifted when I read this article. Not that it matters since 2 people have gone all the way anyway.
Still need to do Illidan, but overall, pretty cool. I learned some things. After Book of Mercenaries is done, I think they should move on. We probably don't need a Book of Alternate Heroes or something. I guess there are a few we could learn more lore from, like Sylvanas.
Don't know about anyone else, but I'll pretty much definitely be converting my entire pack to dust. Except maybe if I get an epic or legendary.
I see. Suggest != choose, so I figured you would retain veto powers.
So my original question remains then: if the best score is 11 wins for example, why doesn't that person get to suggest the next theme?
So ignore that bit. Got it.
My best run ever is 5 wins, so I think I'll be sitting this one out. Good luck to those who do enter.
From the rest of the article, I gather you meant as long as it's 12 wins. I want to ask: why?
I bought a couple bundles back in 2019/early 2020 myself. I've probably only spend around $100 ever on this game, and I haven't spend a cent in over a year. They still offered me the whale bundle.
It records at least everything necessary to make a replay. For example, any movement between deck, hand and board, so that should be useable to keep track of what's in your hand and what's in play. It's how the overlay works, and that can tell you things like what generated or drew a card in the other player's hand.
I recall there was an information leak. I think it was with C'Thun, the Shattered where the deck tracker would show you the other player had drawn a C'thun piece, so the deck tracker authors updated it to hide that information. I think there was some fear that Blizzard would ban people for using the tracker if it revealed stuff it shouldn't, which is technically against the rules, but I don't see how they could find out.
I believe detection of the mod could possibly be evaded. I know some anti-cheat mechanisms use signatures of known cheats, and Blizzard might be doing something more sophisticated, but that might not necessarily make evasion impossible. I had a look, it appears it uses DLL injection. I think botting would be more reliably detected based on behaviors like playing non-stop for over 24 hours with 0 breaks. Something I doubt any human would do.
There's also a logging function you can enable, which is how the deck tracker functions, and that records every move made, without having to access the game's memory space. I think actions are recorded as some kind of code, but most of the reverse engineering work has been done already, and the deck tracker is open source. If it isn't possible to write such a bot as a plugin for the deck tracker.
I think you might have misinterpreted me. I am less convinced this mod is necessary to make such bots, but I can see how being able to send keypresses would make programming a bot easier.
I guess you mean programming the bot to do its thing and ignoring anything the other player does. I didn't know any decks existed that could allow you to win often enough to climb all the way by playing that way. I suppose one could be written and debugged in 2 or 3 days, which is probably enough time before overpowered decks get nerfed.
As I sort of said earlier, I don't think Blizzard will be happy about third-party code being injected in their game. I'm not sure what they can do to stop the mod from working and prevent the author from just updating the mod. I guess they could detect it and ban users. Either way, if an account is playing for more than 24 hours straight with 0 breaks, it should be pretty obvious they are botting.
I haven't done much Windows development, could a bot not send mouse events to the game while it is in the background? Using the mod to send keyboard events would mean that it wouldn't have to know the precise position of UI elements.
It would also be able to get all the information it needs from the gameplay logs. Sure they default to disabled, but turning them on is pretty simple.
As for reaching Legend, even if the bot only knows how to play a single deck, how many situations would the bot need to know how to respond to in order to maintain a 50%+ winrate, especially after D5? I'm assuming your account didn't start with a 13 star bonus in this scenario.
Edit: I just remembered machine learning was a thing, like those AI that play SC2 at Grandmaster level. I don't know about reaching Legend in a couple of days given the amount of time that would probably be spent trying random shit before it learns what works, and all the RNG shit the game throws at you might be another layer of complexity.
Blizzard is probably going to kill this, and from a security perspective, it makes sense. Hopefully this will push them to add official support that doesn't rely on arbitrary code injection that could be easily exploited.