Fundamental Arena Flaw / Proposed Fix:

[2059015]

The fundamental flaw currently in arenas is that they feel like a job at both levels and depending on which end of the spectrum you are at, you either loved or hated one system over another. The heart of the love hate argument is essentially whether or not you agree that a more powerful deck (many times necessitated by spending) should net an advantage in arenas. The short answer is yes, it should. The reason is because in raiding certain cards vs aggregate deck power matter more. Thereby, I agree that fundamentally spending should net you savings of time via having great cards. That said, "skill" should also have some value in the system at large. Skill being my ability to "choose" a smart target vs simply grinding. As you improve your deck, the necessity of grinding should be reduced because of relative power alone. I.E. You create an incentive to not only be powerful, but to be intelligent. Thereby making choices "interesting".

As such, I propose the following system:

Currently, Kabam has a large player base divided into realms. For the purpose of arena, like realms should be matched with 4-5 other like realms to create a potential pool of decks with which to compete against. The reason for this is one of the issues currently is that given the span of power decks can have, there simply aren't enough individuals participating to provide a large enough non-repetitive sample to match each person against. I.E. I personally know who I can beat and who I can't right now. So I either have to skip / wait or grind. That's not interesting. There's no feeling of control when I see my guilds top 3 decks in my easy -> challenge slots randomly. Further, the game knows who can win a match-up, it's simply a matter of chance. What I propose is that the game selects the people you could potentially fight out of 3 categories:

Easy - 95%+ victory rate
Hard - 80%+ victory rate
Challenge - 60%+ victory rate

When I'm in arena I should NEVER face a fight where I have 0% chance of victory. My deck is 420/120, my "Challenge" shouldn't be 800/200. That's like giving me a pocket knife and telling me my "challenge" is a guy flying an Apache. There should always be a "can I win, can't I win" choice that isn't obvious.

That said, as I noted before, being more powerful should be an advantage. It shows some level of spend / time spent (usually a correlation). Money / Time (assuming both are done wisely) should reduce my time spent in the future. Essentially power is an annuity that pays out time savings. In the proposed system, this difference would be manifest in the # of points gleaned per kill. Someone with a 150/50 deck would get say 5,8,10 points where someone at the opposite end would get 10, 16, 20 points. There should be exponential gain to point improvement to a point, then level off and become more difficult. The rough idea would be a y^2=2x type graph (factors could be tweaked).



The overall goal of the system would be 2 fold:
1.) Make it interesting for those who seek something other than a mindless grind.
2.) Make it so having a better deck gives an inherent, but not unbeatable advantage.

The reason for the two above is that eventually arena feels like a job. Just a mindless grind you have to do to compete. That's bad for any game. So by making it engaging it starts to eliminate the "ugh...more arena" feeling. Then by making it so deck power isn't worthless, it gives incentives to play and gives an advantage to those who spend time and/or money (likely both). That said, just because you have the best deck shouldn't be an auto-win button, because it serves as a barrier to entry. If #10 spends 20% more time than #1, with all other factors constant, then #10 should beat #1. This keeps #1 from resting on their laurels lest #10 catch them. This would emulate just about everything in life and wouldn't be inherently unfair.

Just my two cents...

+1 to this. i like it but it would require people to actually fight everyone in order to get the data for the win-loss %. most people would still avoid fighting to save their streaks.

Easy - 95%+ victory rate
Hard - 80%+ victory rate
Challenge - 60%+ victory rate
When I'm in arena I should NEVER face a fight where I have 0% chance of victory. My deck is 420/120, my "Challenge" shouldn't be 800/200. That's like giving me a pocket knife and telling me my "challenge" is a guy flying an Apache. There should always be a "can I win, can't I win" choice that isn't obvious.

one of the ideas of having a top deck is also to block the path to rivals in the fight for the top, so you must have people better than you also

one of the ideas of having a top deck is also to block the path to rivals in the fight for the top, so you must have people better than you also

There's pros and cons to both. This isn't a 1 shot and you're out tournament, it's more about grinding it out by nature of the beast. So in the system described above players with top decks already get an inherent advantage with more points. It can't be a double whammy or you're creating too big a barrier to entry which is bad for the health of the game long term from both a sustainment and Kabam POV. Sustainment because you need new players coming in and feeling like they can compete to keep growing. Kabam POV because they want to keep you spending. If you become "the best" with an up-front investment that is permanently sustainable then why would you ever spend again? That's just bad business. This incentivizes people to keep spending while simultaneously ensuring it's not an overwhelming one. I.E. It's easier to maintain #1 than it is to get there, but still isn't an unbeatable advantage.

one of the ideas of having a top deck is also to block the path to rivals in the fight for the top, so you must have people better than you also

Truth and I agreed.

Overall I feel that op is right.

BUT

We need to tweak somethings, first of all your win rate chances are bad.

Easy - From 95% to 85% win rate.
Medium - From 85% to 25% win rate.
Hard - From 25% to 5% winrate.

This. You should be able to beat that guy on challenge (but it aint gonna be easy), might be a guy wich seems impossible, but you can win, you may risk or you may wait. Its your call.
Challenge is suppose to be Challenge, not a enemy you can win 60% of the time, this would be bad.

Also, I really really think, AND THIS IS ONE OF THE MOST IMPORTANT THOUGHT ABOUT THE ARENA. That at some point of the arena, after the might rewards ends your deck power (status) will not matter for the system.
You'll be matched with other guys who are within the top ranks.

If you're top 12, trying to reach top 5 for example, you will have to fight your way with all players beetween 1-20 position for example.
If your deck will be able to take you there or not this is another matter, if top 20 is surrounded by 200k decks it would be easier for any player who is stronger than that to reach top 1, simple as that.

Truth and I agreed.

Also, I really really think, AND THIS IS ONE OF THE MOST IMPORTANT THOUGHT ABOUT THE ARENA. That at some point of the arena, after the might rewards ends your deck power (status) will not matter for the system.
You'll be matched with other guys who are within the top ranks.

If you're top 12, trying to reach top 5 for example, you will have to fight your way with all players beetween 1-20 position for example.
If your deck will be able to take you there or not this is another matter, if top 20 is surrounded by 200k decks it would be easier for any player who is stronger than that to reach top 1, simple as that.

although i 100% agree and like your idea,

but there would always be ppl saying "the strong will take all the prizes, and what makes the strong people entitle to all the rank prize" & "time spent should be rewarded too"

i think once a while, kabam should make "KAMIKAZE ARENA"
whoever spend the most time win, as simple as that lol

I'm not good at math and graphs make me wanna puke but I agree with the overall logic. We should have a choice who we fight in each of the 3 slots. Before the rollback the AI gave me my opponents which I knew could never beat me so instead of gaining the +10 might for the Challenge win I only gained +5. Which is wrong. If give the chance to fight someone harder at the chance to gain more of that might I would. But I didn't have a choice and was therefore punished for it. So a weaker deck with a smaller disparity between the one it beats for more points. That's where the major flaw was.

But when it comes to rank rewards my top deck should carry weight and be able to block weaker opponents and make it more difficult for them. Otherwise it just comes down to who has the most time to tap the screen.

Easy - 95%+ victory rate
Hard - 80%+ victory rate
Challenge - 60%+ victory rate

When I'm in arena I should NEVER face a fight where I have 0% chance of victory. My deck is 420/120, my "Challenge" shouldn't be 800/200. That's like giving me a pocket knife and telling me my "challenge" is a guy flying an Apache. There should always be a "can I win, can't I win" choice that isn't obvious.

This is what the new system, essentially, is supposed to do (not exclusively, but this is a core component of it). ELO in matchmaking is supposed to help determine probability of victory. It's hard to look at your card stats and say, "Oh, this is your % chance of winning" because the combos and skills are a huge factor in determining outcomes. And it's hard to create a straight comparison of combos and skills, because there is so much variety and so many other factors to consider (like stats).

So, ELO is supposed to help with this. Ideally, we'd want to use an actual simulation to assess your probability of victory. But alas, this would put huge load on the game, and it would case slow play and a lot of connection issues.

(note: we didn't set the same win% per difficulty as you, but the idea is the same.)

Currently, Kabam has a large player base divided into realms. For the purpose of arena, like realms should be matched with 4-5 other like realms to create a potential pool of decks with which to compete against. The reason for this is one of the issues currently is that given the span of power decks can have, there simply aren't enough individuals participating to provide a large enough non-repetitive sample to match each person against. I.E. I personally know who I can beat and who I can't right now. So I either have to skip / wait or grind. That's not interesting. There's no feeling of control when I see my guilds top 3 decks in my easy -> challenge slots randomly. Further, the game knows who can win a match-up, it's simply a matter of chance. What I propose is that the game selects the people you could potentially fight out of 3 categories:

This is something we are pushing for hard. I wanted it in Version 3.4, but it looks like it shall be included in the Version 4.0 update. This is also where the League/Ladder/Leaderboard system that I've previously mentioned comes into play. It ranks players across realms, rather than by individual arenas alone.

Glad to know you guys have a path to follow.
I feel excited just thinking on a realm arena

Easy - From 95% to 85% win rate.
Medium - From 85% to 25% win rate.
Hard - From 25% to 5% winrate.

This. You should be able to beat that guy on challenge (but it aint gonna be easy), might be a guy wich seems impossible, but you can win, you may risk or you may wait. Its your call.
Challenge is suppose to be Challenge, not a enemy you can win 60% of the time, this would be bad.

Also, I really really think, AND THIS IS ONE OF THE MOST IMPORTANT THOUGHT ABOUT THE ARENA. That at some point of the arena, after the might rewards ends your deck power (status) will not matter for the system.
You'll be matched with other guys who are within the top ranks.
.

Kinda disagree. Having a challenge be that low a % would necessitate spamming. I went for a number right around 50% to make it challenging, but not a stupid bet. I.E. If the top % was < = 25% for everyone then the only option would be to spam, because it'd NEVER be worth it to wait for the cool down. The number needs to be such that assuming a full win streak, it's better to be a 100% guesser on "Challenge" than it is to spam the button. That's the current fundamental flaw. It takes 0 skill to win arena, you sit, and you spam with a decent deck. The end. It should be a combination of 3 things, skill (in being able to tell who you can beat), time (you need to be able to invest it) and deck power. Those three components need to determine winning, currently only time does and maybe with reversion deck power. A monkey with a strong deck and a keyboard shouldn't beat a mid deck person with skill + slightly more time.

This is what the new system, essentially, is supposed to do (not exclusively, but this is a core component of it). ELO in matchmaking is supposed to help determine probability of victory. It's hard to look at your card stats and say, "Oh, this is your % chance of winning" because the combos and skills are a huge factor in determining outcomes. And it's hard to create a straight comparison of combos and skills, because there is so much variety and so many other factors to consider (like stats).

So, ELO is supposed to help with this. Ideally, we'd want to use an actual simulation to assess your probability of victory. But alas, this would put huge load on the game, and it would case slow play and a lot of connection issues.

(note: we didn't set the same win% per difficulty as you, but the idea is the same.)



This is something we are pushing for hard. I wanted it in Version 3.4, but it looks like it shall be included in the Version 4.0 update. This is also where the League/Ladder/Leaderboard system that I've previously mentioned comes into play. It ranks players across realms, rather than by individual arenas alone.

Good to know I was on the right track? That said, I reallllly like the cross realm idea vs cross arena. That's paramount (getting a bigger sample size) to evening this thing out.

Also, Droid players who were locked out, any word on the compensation for that? Ticket #02597258? has been open for 12 days...

When I'm in arena I should NEVER face a fight where I have 0% chance of victory. My deck is 420/120, my "Challenge" shouldn't be 800/200. That's like giving me a pocket knife and telling me my "challenge" is a guy flying an Apache. There should always be a "can I win, can't I win" choice that isn't obvious.

Give your pocket knife to Chuck Norris. I'd like to see how's the guy flying an Apache react.

Easy - 95%+ victory rate
Hard - 80%+ victory rate
Challenge - 60%+ victory rate

Your system has a fundamental flaw. The computation of the win percentage. It is a very complex algorithm which is essentially polynomial in complexity. 12 card activations per fight, a total of max fights = > Impossible to compute on so many players. Essentially impossible for a bunch of mobile developers. Heuristics may be implemented, but the win percentage could be so off you'd just get screwed over easily. Just fighting the decks let's say 10 times against each other and see who wins is not by far a good percentage calculator. Fighting them at least 100 for some meaningful percentage would lead to so much lag the game will be unplayable.

Edit:

There should be exponential gain to point improvement to a point, then level off and become more difficult

Do you even know what exponential means? Even a linear gain may be too much for this suggestion.

I see now that aaron already has addressed these issues and suggested that ELO could be used to compute the win percentage which he is actually right. However i'm not so sure the new system will guarantee one could win medium and easy fights. Certainly not if players are solely choosed by might and simply ordered by ELO.

Kinda disagree. Having a challenge be that low a % would necessitate spamming. I went for a number right around 50% to make it challenging, but not a stupid bet. I.E. If the top % was < = 25% for everyone then the only option would be to spam, because it'd NEVER be worth it to wait for the cool down. The number needs to be such that assuming a full win streak, it's better to be a 100% guesser on "Challenge" than it is to spam the button. That's the current fundamental flaw. It takes 0 skill to win arena, you sit, and you spam with a decent deck. The end. It should be a combination of 3 things, skill (in being able to tell who you can beat), time (you need to be able to invest it) and deck power. Those three components need to determine winning, currently only time does and maybe with reversion deck power. A monkey with a strong deck and a keyboard shouldn't beat a mid deck person with skill + slightly more time.

The rates I purposed was not taking in consideration combos / skills, its impossible to compute comparing 2 teams, taking all this in considerations, just a little math upon the base stats.
That being said, you would be able to mount specific teams to beat up all 3 oponents choosing the right cards / combos.
0
For example: I fight the tops of my server with a expecific team with Inta Kill combo, wich grants me a 70% victory rate upon them, we have "close" base stats, (900k / 250k VS 1.2k / 300k) but he have very superior combos, and still i have a better probability of victory choosing the right team.

The rates I purposed was not taking in consideration combos / skills, its impossible to compute comparing 2 teams, taking all this in considerations, just a little math upon the base stats.
That being said, you would be able to mount specific teams to beat up all 3 oponents choosing the right cards / combos.
0
For example: I fight the tops of my server with a expecific team with Inta Kill combo, wich grants me a 70% victory rate upon them, we have "close" base stats, (900k / 250k VS 1.2k / 300k) but he have very superior combos, and still i have a better probability of victory choosing the right team.

Insta Kill is quite underrated in my opinion. Specially against higher players.
It may not procc to often, but it can be a very nice one to have, if you are equal in strength with someone.

This is what the new system, essentially, is supposed to do (not exclusively, but this is a core component of it). ELO in matchmaking is supposed to help determine probability of victory. It's hard to look at your card stats and say, "Oh, this is your % chance of winning" because the combos and skills are a huge factor in determining outcomes. And it's hard to create a straight comparison of combos and skills, because there is so much variety and so many other factors to consider (like stats).

So, ELO is supposed to help with this. Ideally, we'd want to use an actual simulation to assess your probability of victory. But alas, this would put huge load on the game, and it would case slow play and a lot of connection issues.

(note: we didn't set the same win% per difficulty as you, but the idea is the same.)









This is something we are pushing for hard. I wanted it in Version 3.4, but it looks like it shall be included in the Version 4.0 update. This is also where the League/Ladder/Leaderboard system that I've previously mentioned comes into play. It ranks players across realms, rather than by individual arenas alone.

It's definitely doable to create this given the data you have, as well as creating a statistical model for it, especially when you have the exact proc rates and etc. Outsourcing a statistician could probably be helpful.

Although the previous system as per Aaron stated emulates the model the op proposed, the reward system for the strength of the deck was completely reversed and thus a failure in the eyes of the paying players. I haven't seen the top 5 players in the realm recently, the new system pretty much undermines all that they have spent.

The two system has it's flaws, both seem linear; either taking into account deck strength only or might range. A dynamic system could be better moving forward; with a cumulative percentage leaning towards the might system as it progresses.

Another way to implement this deck strength win is to start with 100 percent win probability on all 3. The main difference is that as a win is accumulated, these probability deteriorates. For example as u win hard it becomes 95, medium 97, easy 99 until this "bonus ez mode comes to 0".

On another thread a really smart observation was noted and that is the refresh and might bonus.

There is also another angle that could be considered which hasn't been discussed and that is lose ratio. I understand that people will try not to lose due to the might bonus/streak, but why not consider mining it on the back end and adding that as a matchmaking factor? For instance player a's matchmaking will be determined by how many players beat him when he is served as an opponent.

Ofcourse you may argue that people might lower their decks, and I agree but this is a double edged sword, it provides them an edge but it does so to the people who get him as a freebie. Might also be a good when I'm sleeping/working catch up system.

Please consider these even though I'm not a developer. Lawl