Version 0.599g

[Teddy]

Patch 0.599g

Summary: The race is on: official Steam Leardboards have arrived!

The Leaderboards are separated in four different categories, namely 1-4 Players. The top #3 single players, and the top teams will be displayed with portraits, while the rest will have to settle with their names! For those more interested in competing with their friends than the rest of the world, there's a Friends Only sorting option available.

There will likely be some cheating happening, and for now I want to see how the cheat detection is working. At some point I might implement an auto-ban for it but then the detection must be really robust in terms of false positives!

With the leaderboard addition we've made some changes to how scoring works. Before, you got "hard" grades that rounded down, so you needed double S to get a total rank of S. Now you get floating grades behind the scenes, meaning you can get a perfect S and a strong A and it will average out to an S. Also, we've added a fifth grading letter, "D" which awards no bonus points at all. Basically, the changes make it a bit easier to do well, but the scoring overall is actually more punishing.

If this patch seems to be reasonably bug free, I'll bring the Stable-players to the party in a few days/a week.

Additions

• The Steam leaderboards are live for Arcade Mode!
• In Arcade Mode, you can hold TAB after leveling up to access a quick level interface with your recently leveled skills
  
• More sounds! Smash, flamethrower and samurai spirit now got sounds, as well as some events in Arcade Mode
  
• Temple of Seasons has now gotten a proper, awesome Arcade Mode remix!

Changes

• A bunch of portraits have been updated
  
• Fixed a bug where Spirit Slash would stack bonus damage applied on each hit
  
• Winter's Winds + Ice Tremors combo has been tuned to have more consistent timings
  
• Floor scoring in Arcade Mode is now much more fluid - read the patch note forum thread for more info
  
• Fish now bites much faster in Arcade Mode, and the fishing event rooms have gotten decreased time
  
• You now get a silver skill point on all levels, even when you get gold points (works with old saves)
  
• Two instances of the same sound can no longer be played on the exact same frame, to prevent volume spikes

Bug Fixes

• Chickens in the Temple of Seasons will no longer escape their destiny to be herded (in Arcade Mode)
• The Phaseman arena should no longer have a weird magic barrier somewhere along the top
  
• The Sentry should now properly drop stuff when you defeat it in Arcade Mode
  
• The Pet Menu now pauses the game

Next Patch: For version 0.599h, I'll be adding Steam Achievements which mirrors the in-game achievements. We'll also be adding a bunch of new achievements to the game!

 

[KoBeWi]

I made a cheat detection test and no warning appeared (something is supposed to appear, doesn't it?). The cheat is obvious, but maybe I got too low score... (D-rank seems to work). I can still submit it if it helps somehow (testing purposes ofc)

Spoiler: Screen

This quick level menu is broken - you can't go right when selecting talents. Pressing right moves selection to left side.
Also, I'm not sure if this should be like that, but when you defeat Living Puzzle Block in Arcade, it remains on ground, but you can go through it.

And I love the new Season Temple music in Arcade

 

[Teddy]

I can still submit it if it helps somehow (testing purposes ofc)

Please do! I'm curious to see how that didn't trigger the warning message!

As for the QuickLevel bug... d'oh!

 

[KoBeWi]

Please do! I'm curious to see how that didn't trigger the warning message!

Done. I even got almost 7k damage on last floor XD

 

[Own]

Samurai Spirit? You mean Spirit Slash?

Samurai Spirit is a much cooler name.

Also, at seeing all the new skill points in Story Mode!

But then at not seeing missing Talent Orbs fixed too. Someday I'll be able to invest in +Potion Duriation, someday. ;_;

Edit: Berserker got sounds too.

 

[Teddy]

Done. I even got almost 7k damage on last floor XD

It's very obvious when looking at the metadata (your total damage was an impressive 19725), but somehow it didn't get auto flagged! I think I'll hotfix the quicklevel-thing and also throw in a little tweak for the cheat detection as well. Maybe you could try getting a score (doesn't even need to be high) using the same cheat again and see if it works

Samurai Spirit? You mean Spirit Slash?

Samurai Spirit is a much cooler name.

Yes! And yes!

But then at not seeing missing Talent Orbs fixed too. Someday I'll be able to invest in +Potion Duriation, someday. ;_;

Someday

 

[KoBeWi]

Maybe you could try getting a score (doesn't even need to be high) using the same cheat again and see if it works

OMG, such cheater XD

Spoiler: Screen

But this one looks hilarious though (have you ever seen so much Essence?):

Spoiler: Screen2

EDIT: The warning of course appeared also for second screen, but I took it after dismissing it.

 

[Own]

What sort of criteria does the anti-cheat system even look for, out of curiousity? I can think of quite a few things it could silently record over the course of the run to raise a flag.

- Highest damage > #, where # is something literally impossible to hit even with max crits on a maxed skill. You can't hit 5k naturally, right?

- Score exceeding # on early floors, where that score is impossible to get even with all S-Ranks.

- Gold exceeding # on early floors, again an impossible amount.

- All stat values individually exceeding an impossible barrier given the limitations of Arcade. ATK, MATK, HP, EP, Shield HP, Defense, etc. I wonder if, under the current limitations, someone could sneak under the radar by just giving themselves 999 or even 250 defense and leaving everything else alone?

Also, strange to see someone already having beaten Arcade Mode. I wonder what their build was? If someone is that good I'd expect them to post here.

Edit: Oh, they've already been deleted. Welp~

 

[Teddy]

The auto detection only looks for memory editing on the most useful stats, currently. I think I know why Kobewi managed to "beat" the detection, but we'll see when the hotfix comes live and he tries it again!

The additional metadata contains a load of stuff, such as total damage taken, stats, enemies killed, etc. The highest hit getting recorded sounds like a sound addition, though!

 

[Teddy]

OMG, such cheater XD

I think the hotfix is live on Steam now! Jump in and do the same HP cheat again and see if it triggers!

 

[KoBeWi]

It triggers now!

 

[Roemer]

Yeey, I'm on the 2p and 3p highscore board Nice update! One more bug I just saw: In the winter challenges where frost comes from different sides, only the host could see the frost and the other players not.

 

[Own]

The auto detection only looks for memory editing on the most useful stats, currently.

DEF is one of the most useful stats, right? You don't need high damage if literally everything does 1 damage to you.

Also, does it check if stats are ever above some threshold, or just at the end of a run? Someone couldn't just give themselves 999 attack for floors 1-12, then edit it back down to 100 before submitting I assume?

 

[Teddy]

DEF is one of the most useful stats, right? You don't need high damage if literally everything does 1 damage to you.

Also, does it check if stats are ever above some threshold, or just at the end of a run? Someone couldn't just give themselves 999 attack for floors 1-12, then edit it back down to 100 before submitting I assume?

DEF is one of those stats, and once a flag is set it can't be unset.

That said, I'll pre-emptively admit that the detection is a work in progress, and it probably won't ever cover all the creativeness a dedicated person could muster. Hopefully the autodetection + metadata is enough to see most, though!

 

[Own]

A few more questions about the cheat/ multiplayer system:

1. Two people are playing in Arcade Mode, one drops out mid-way through. Does the game still treat this as a 2-player run when it's over, or just a 1-player run?

2. Two people are playing in Arcade Mode. One plays legit, the other cheats their butt off. Does the non-cheater get the warning message as well when they both finish the run, or does the game only check for cheating on a per-player basis?

3. Two people are playing in Arcade Mode. The host is legit, P2 cheats their butt off. P2 quits / ALT+F4's before the end of the run. Does each player's metadata contain the metadata of everyone else who was in that run? Or, with the cheater no longer in the game when it's over, does it just contain the legit player's info which appears clean?

 

[GarlicJelly]

A few more questions about the cheat/ multiplayer system:

1. Two people are playing in Arcade Mode, one drops out mid-way through. Does the game still treat this as a 2-player run when it's over, or just a 1-player run?

2. Two people are playing in Arcade Mode. One plays legit, the other cheats their butt off. Does the non-cheater get the warning message as well when they both finish the run, or does the game only check for cheating on a per-player basis?

3. Two people are playing in Arcade Mode. The host is legit, P2 cheats their butt off. P2 quits / ALT+F4's before the end of the run. Does each player's metadata contain the metadata of everyone else who was in that run? Or, with the cheater no longer in the game when it's over, does it just contain the legit player's info which appears clean?

I think that the anti cheating system flags the game as a "cheated" run once a cheat is detected during the run so even if the player cheating drops out the game would still detect the run as a cheated one.

 

[Teddy]

1. Two people are playing in Arcade Mode, one drops out mid-way through. Does the game still treat this as a 2-player run when it's over, or just a 1-player run?

To avoid invasions and possibly unforseen score interactions, the game will think the amount of players is the same as when the run started. In other words, if someone drops out, the remaining player(s) simply has a tougher run ahead of them. It's not a perfect system, but it's less imperfect than giving players the ability to transform a run into a different leaderboard, so to speak.

The weird part with this comes with the fact that teams sometimes can't submit as a group. If a three player run gets the three player high score, and one of them can't submit for one reason or another, the third player will be shown as "anonymous".

At some point I'll probably add a reconnect option, where connections from the same IP/SteamID is accepted mid-run (in cleared rooms), and the player returns with the same HP as when he/she left. There's a lot involved in accurately copying an ongoing arcade run, though, so while technically trivial it's still pretty cumbersome and bug-prone work.

2. Two people are playing in Arcade Mode. One plays legit, the other cheats their butt off. Does the non-cheater get the warning message as well when they both finish the run, or does the game only check for cheating on a per-player basis?

3. Two people are playing in Arcade Mode. The host is legit, P2 cheats their butt off. P2 quits / ALT+F4's before the end of the run. Does each player's metadata contain the metadata of everyone else who was in that run? Or, with the cheater no longer in the game when it's over, does it just contain the legit player's info which appears clean?

Whenever it makes sense, the metadata is shared (such as damage taken, enemies killed, etc.). If the game's own detection gets triggered on one machine, it propagates to every other player as well, so the entire run gets flagged as a cheating run. Everyone gets the cheater warning.

 

[GarlicJelly]

At some point I'll probably add a reconnect option, where connections from the same IP/SteamID is accepted mid-run (in cleared rooms), and the player returns with the same HP as when he/she left. There's a lot involved in accurately copying an ongoing arcade run, though, so while technically trivial it's still pretty cumbersome and bug-prone work.

I haven't experienced the difference between 3 players and 4 players but it sounds like this could be exploited pretty easily. It wouldn't surprise me if someone found a way to use this to their favor.

 

[Teddy]

I haven't experienced the difference between 3 players and 4 players but it sounds like this could be exploited pretty easily. It wouldn't surprise me if someone found a way to use this to their favor.

The game will still be scaled for four players! Disconnecting willingly and leaving your team mates fighting tougher battles while you miss out on the rewards seems entirely bad for everyone. The only way I see this used (unless someone finds an exploitable bug) is when someone crashes or their internet dies during a fight, and they want to reconnect ASAP.

 

[GarlicJelly]

The game will still be scaled for four players! Disconnecting willingly and leaving your team mates fighting tougher battles while you miss out on the rewards seems entirely bad for everyone. The only way I see this used (unless someone finds an exploitable bug) is when someone crashes or their internet dies during a fight, and they want to reconnect ASAP.

Yeah but how big is the gap in difficulty from 3 to 4 players? And how does the EXP work if the enemies are scaled for 4 players but there are only 3 present to receive EXP? I mean if the difference isn't that big and it's manageable to play as 3 players in a 4 players difficulty then couldn't there be a way for players to exploit this? And if you get more EXP as 3 players in a 4 players session then couldn't that also be exploited? I wish I could try these things out myself right now.