PDA

View Full Version : Suggestions for a better PVP point structure



slauki
04-30-2017, 01:02 PM
Hey guys,

since this was buried under a lot of new posts, i wanted to bump it, and this discussion deserves it's own thread. We have seen quite bizarre
trends on the last week of PVP since we are punished the more we play, so some guys didn't play much and have excellent results (Hi Nicko :-) ).
The main reason is the bad point system which punishes us far more than we deserve along with thge current matchmaking.

Yesterday i was defending my top 10 in the arena and it was hard to gain points since for every 12 points i won, i lost 20. It was very annoying, i felt like someone was robbing my points, prolly
many of us feel the same.

We had a good discussion with sparton in the other thread, about the point system, and he gave us some parameters to think about.

He constructed a 5 archetypes of players and tried to look at the implication a changed point system will have. The main problem is that someone shouldn't be able to simply be number 1
even if his attack ratio and his defense ratio is bad. So it's a balancing act between gaining points and loosing if we want a fair and sane competition.

I took Spartons archetypes and added some more and tried to think about a better point system.

This are my results so far.

2875

Explanation:

My main goal was that in the end of the week every archetyp has a positive result just by participating. So everyone would have more fun since he gaines at least something. This is hard to do with standard amount for every player.

So i manipulated the defense losses and the attack losses in a way, that makes sense i belive.

If we have a crappy defense because we are new to the game or bad RNG this is not our fault. So we are not able to "build better defenses", because we don't have the tools. That's why we need some protection there. Once our defense reaches <15% holdrate i would suggest to cut the defense losses from -12 to -8,5 . This is at least some kind of protection for these guys. If someone intentianally put in a bad defense to profit from it this won't work out for him, because the points won from holds are very sweet, he won't be able to compare with the bigguns.

Same could apply to attacks. If our attack ratio reaches <85% i would eliminate attack losses so that the guys are not punished just by trying.

The hardest part was to balance hyper good-o, the good and the lazy great.
i thought that all of them should be in a close range. In this particular example you see that thy hyper good-o has a tiny edge over the good and a slight edge against the great but lazy player. I fell this is okay since we compare 1000 vs 500 vs 250 attacks. So i think this is reasonable.

The great player with 500 attacks outmaneuver them all by a large amount. Not sure if his advantage is too big, but OTOH it should be really big when he wins 98% of time in 500 fights and holds 30% in 500 defenses... But we can argue here for sure, about some pros and cons.
The free player have okay results, but they cannot compare with the people who played much more. This should be reasonable, because who plays more and better should get more points at all.


Please be aware, that the attack win is an avarage result. To prevent bumhunting we should give people more points when attacking highranks and less ponits when attacking lowranks, to keep the competition fair for everyone.

This is just one example of a possible way to go, i'm sure this can be done better, but i think it would be a huge progress to what we have now...
We need positive motivation and not negative motivation if this should succeed.


If you want to try to make it better use this excel spreadsheet i wrote. It calculats itself you only have to put in points.
I'm very sure that the calculations are correct, but if someone could verify my results this would be even better.

You can download it here:
https://1drv.ms/x/s!AmnD5WIsUqymh6xFO9VWBojaj8Ho2Q


Another suggestion: If the devs want to revive the arena, we need much more than a better pointsystem. More rewards are needed to motivate the people i belive.
Add more ironite/gold/souls to the arena drops. Add more weekly rewards and more stuff in the store that we can buy with coins. Consider to let us purchase stuff more often, maybe twice per week.
And so on. There are many ways to improve PVP, so let's start right now. I feel if PVP dies, the game will die too.

As always i appreciate constructive critic and stuff, let me know what you guys think. This is not finished yet, but we can discuss it, since the devs seem to listen.

Shaolin85london
04-30-2017, 01:47 PM
Yes, conducting every single archetype in a specific result, a rule system within a basic rule concept, d be the way to go imo. Is like a web or a rank inside the rank. I like that idea

Eddie_PL
04-30-2017, 01:53 PM
It supposed to be free to play mobile phone game, fun game. When I'm seeing excel spreadsheets with that kind of calculations that makes me sick. Some of you guys are taking this game too serious, take a break.

Nicko
04-30-2017, 03:51 PM
It supposed to be free to play mobile phone game, fun game. When I'm seeing excel spreadsheets with that kind of calculations that makes me sick. Some of you guys are taking this game too serious, take a break.

Hmmm...judging others from your own limited perspective isn't always an effective way to view the world...especially when you then tell them what to do...

He's a teacher - for him this IS fun ;)

Silentknight
04-30-2017, 03:59 PM
Hmmm...judging others from your own limited perspective isn't always an effective way to view the world...especially when you then tell them what to do...

He's a teacher - for him this IS fun ;)
And greatly appreciated! Thx 4 time & effort slauki! Thx to Nicko & Sparton too!!!

Jofer16
04-30-2017, 04:08 PM
Hmmm...judging others from your own limited perspective isn't always an effective way to view the world...especially when you then tell them what to do...

He's a teacher - for him this IS fun ;)

Agreed, and nice to see the forums active today. Was a ghost town last night so I tried messing around haha. Everyone must have been burnt out from the day.

Nice post Slauki.

Nicko
04-30-2017, 04:30 PM
Agreed, and nice to see the forums active today. Was a ghost town last night so I tried messing around haha. Everyone must have been burnt out from the day.

Nice post Slauki.

Yeah I rushed on out right after reset - and with the monsoon we had here in Chicago kind of wished I wouldn't have!

Tritium
04-30-2017, 05:06 PM
I love this thread and your effort to help fix the point system's problems, slauki! That Eddie_PL guy just sounds like one of those insecure types who lashes out at others ─ don't listen to him or any of the other naysayers who might pop up here later on. Rather than complaining about the system, you're actively working to help make it better, which I find awesome. Your effort is greatly appreciated, slauki!

Ian
04-30-2017, 05:56 PM
A well thought out post Slauki and I hope the devs take on board what you are saying. I'm glad the matchmaking was fixed because nobody wants to have 400 defended when they've only attacked 200 times. The other thing which may harm participation is the added bugs - for players like me who can only find time for around 200-250 matches per week, the motivation disappears when the game doesn't even work: Server errors when you win, talisman or toon passive effects not working, etc etc. It's become more lottery than ever.
I don't want to take the thread off topic but we must also realise that these bugs that make the arena less enjoyable also coincided with the matchmaking fix.

Jofer16
04-30-2017, 10:59 PM
Yeah I rushed on out right after reset - and with the monsoon we had here in Chicago kind of wished I wouldn't have!

Been crazy and windy here lately too, think we're supposed to get snow tomorrow!

Silentknight
04-30-2017, 11:27 PM
Been crazy and windy here lately too, think we're supposed to get snow tomorrow!

No snow in forecast here, lots of rain tho! Clean everything up, gets rid of all the evidence!!! Lmao! Joy Ride reference.

Chaosego888
04-30-2017, 11:30 PM
Damn! I wish I lived in the North. This miserable humid hellhole sucks!

Silentknight
04-30-2017, 11:33 PM
Damn! I wish I lived in the North. This miserable humid hellhole sucks!

4c or 39f here

Tritium
04-30-2017, 11:49 PM
Man, you guys have it rough in terms of weather. :p (Don't look at my location). ;)

Chaosego888
04-30-2017, 11:55 PM
Man, you guys have it rough in terms of weather. :p (Don't look at my location). ;)

I have been there, a little more comfortable even when it's hot. I flew out there to see Maiden in June of 2008... 85 there is like 65 here

Sparton_LOTB
05-01-2017, 06:54 AM
Thanks for making a dedicated thread for this, slauki. Great explanation and chart for your approach, too.

A couple things I want to point out:


The average attack win of 10 means lowering the floor from 12 to 10; probably still feels better overall if defense losses are usually not as severe as they are right now, but that's something to consider
This is a flaw with my original calculations as well, but the suggestion of average attack probably makes most sense for people ranked near the top; people who are ranked lower probably need to assume to have better average returns (it'll be rare that they decide to pick on just lower-ranked people, but that somewhat of a choice for them). This is not the worst thing in the world, but it makes the next point a bit more of a problem...
With the way the weekly point reset works, a lot of people getting to 1900+ by the end of the week will reset to around 1500 (and higher up people will reset even higher); this means almost every archetype listed will reach Warlord III points every week, which trivializes the climb through Acolyte and Duelist divisions unless we greatly raise the division VP requirements for Warlord (and/or add more divisions... although reward differentials between division is already discussed as pretty weak). Overall, I'd want the climb into Warlord's point requirement to be a bit more meaningful than that (while yes, it's a bit out of reach for the 250 people right now, this system would produce the opposite extreme). This could be done by not slashing the point losses for losing as much, or by further reducing point gains on successful attacks/defends, but I'm not sure if that just makes climbing feel slow in a different way (and would make the fight for the divisions near reset less turbulent)
Specifically forcing the loss points down based on your current win rate does complicate the point calculation more than I'd like, but it's overall a pretty smart idea to keep things in line based on win rate. I'd probably want it to gradually hit the breakpoints you suggest (such as -12 for defense loss normally, but -10 for 16 - 20%, then -9 for 15% or less or something), but that's a minor detail at that point, and going with just a single breakpoint is probably still fine.
Your approach requires three things our tech doesn't currently handle; handling half-points, adjusting point yields based on your respective win rate, and adjusting defense wins to be a particular value (instead of being based on what an attack win would be). I wouldn't really want to go the direction of fractional points, as that could be worked around without losing too much from what you've got there. Beyond that, the other two things aren't really a problem, it's just two things that need their own implementation and testing time, which if we do choose to go this route, it puts some time and resourcing barriers before we can have that implemented and then go live


In the next major update (not the upcoming hotfix update), we already have logic put in place to reduce point losses based on a fraction of the normal loss value, but if it doesn't turn out as well as we'd hope, I think what you're proposing here is probably along the lines of the next thing to try. Maybe not the exact same values you've proposed, but the concepts of a greater bonus for a defensive hold and overriding loss values based on win rate are sound ideas.

Jofer16
05-01-2017, 12:31 PM
Can’t wait for the updates fixing bugs and the point system, thanks for the info, hoping the bug fix one is today. Lord Schmeb’s latest post where he logged his activity in the bug reports forum is ridiculous lol. Must be getting much worse with more strike talismans. I had a bunch of icon disappear stalls in both PvP and PvE using them yesterday, sorry to my troopers that will lose those points :( Thinking it’s probably best to play as little arena as possible and not use those talismans in PvE until things are fixed.

Liebhild
05-01-2017, 01:28 PM
Thinking its probably best to play as little arena as possible and not use those talismans in PvE until things are fixed.

I still wonder, why those strike talismans are still in the game, they are causing so many trouble. And in my eyes they are also useless. There are toons that can counter, strike talismans make playing only more lottery-play even if they are able to fix them.

MrFreeze
05-01-2017, 02:55 PM
I haven't been playing enough recently to comment on the point system. I can comment on the climb though. The last time I gave pvp any effort was during the beta testing, so I was unranked coming back into it last week. Getting to warlord 3 wasn't hard at all. There didn't seem to be a lot of activity below the top 200 or so. It's possible some lower level players avoided me, but there were plenty of lv100 players down there too. It wasn't til I got to the top 100-150 that I started getting hit regularly.

Maybe if lower ranks got a rare soul or something there'd be more activity down there. There's not much incentive below top 25 for end game players. Mid game or low level players have to compete against maxed out players for a chance at top 250 and 1 rare soul. It's probably impossible considering how many end game players there are. Spreading out the rewards more would increase interest I think.

Sorry, a bit off topic, but that's what I saw climbing up from the bottom. Im almost tempted to set a crap defense and just farm iron coins down there til the last couple days.

Edit: the numbers, so you can see what I mean. I attacked 138 times, but was only attacked by other players 12 times. This week is a more even 74/73 right now.

JJJ428
05-01-2017, 03:19 PM
I be ok with the strike talismans being taken away just hope we are reimbursed any Iron coins and runes we spent on them.

Ringe666-7406
05-01-2017, 03:43 PM
In addition to Slauki's proposal for higher rewards, it'd be nice to get some bonuses for finishing battle with a fury (at least minimal, for the sake of it). Great way to crown a successful battle.

mjmxiii
05-01-2017, 05:08 PM
Thanks for the great posts Slauki. Agree the points system needs to change as PvP is a great idea but has always seemed flawed to me due to it's point system. Every attempt to adjust it only seems to magnify the intitial flaws. Would welcome a change to it but a few things that would motivate me more than a better points system follow...

I would like to see better items in store as that would have motivated me to play more from the start. For example, the arena store should not only have SoW, current souls and skill shards, it should also have rare GA souls, Eternal souls (as in souls that only drop Eddie chars, not just Eternal Ed), and then the soul fragments that they speak of could be a separate item that would over time, give you specific chars (an example being a Prisoner Tier after reaching a certain amount of them, if he drops through other means, you look to the next char tier you want most) and all should be unlimited purchase.

While I dont question the value or really understand why they would release items like these, they should have never released talismans like Strike and Immunity through the arena store. All those items seem like they should be drops from secret locations or BNW and would have given us a bit more incentive or reward for the daily trooper grind.

A minor thing that makes me SMH, why not just have an arena tab in the game store?

slauki
05-03-2017, 10:17 PM
Thanks for making a dedicated thread for this, slauki. Great explanation and chart for your approach, too.

A couple things I want to point out:


The average attack win of 10 means lowering the floor from 12 to 10; probably still feels better overall if defense losses are usually not as severe as they are right now, but that's something to consider
This is a flaw with my original calculations as well, but the suggestion of average attack probably makes most sense for people ranked near the top; people who are ranked lower probably need to assume to have better average returns (it'll be rare that they decide to pick on just lower-ranked people, but that somewhat of a choice for them). This is not the worst thing in the world, but it makes the next point a bit more of a problem...
With the way the weekly point reset works, a lot of people getting to 1900+ by the end of the week will reset to around 1500 (and higher up people will reset even higher); this means almost every archetype listed will reach Warlord III points every week, which trivializes the climb through Acolyte and Duelist divisions unless we greatly raise the division VP requirements for Warlord (and/or add more divisions... although reward differentials between division is already discussed as pretty weak). Overall, I'd want the climb into Warlord's point requirement to be a bit more meaningful than that (while yes, it's a bit out of reach for the 250 people right now, this system would produce the opposite extreme). This could be done by not slashing the point losses for losing as much, or by further reducing point gains on successful attacks/defends, but I'm not sure if that just makes climbing feel slow in a different way (and would make the fight for the divisions near reset less turbulent)
Specifically forcing the loss points down based on your current win rate does complicate the point calculation more than I'd like, but it's overall a pretty smart idea to keep things in line based on win rate. I'd probably want it to gradually hit the breakpoints you suggest (such as -12 for defense loss normally, but -10 for 16 - 20%, then -9 for 15% or less or something), but that's a minor detail at that point, and going with just a single breakpoint is probably still fine.
Your approach requires three things our tech doesn't currently handle; handling half-points, adjusting point yields based on your respective win rate, and adjusting defense wins to be a particular value (instead of being based on what an attack win would be). I wouldn't really want to go the direction of fractional points, as that could be worked around without losing too much from what you've got there. Beyond that, the other two things aren't really a problem, it's just two things that need their own implementation and testing time, which if we do choose to go this route, it puts some time and resourcing barriers before we can have that implemented and then go live


In the next major update (not the upcoming hotfix update), we already have logic put in place to reduce point losses based on a fraction of the normal loss value, but if it doesn't turn out as well as we'd hope, I think what you're proposing here is probably along the lines of the next thing to try. Maybe not the exact same values you've proposed, but the concepts of a greater bonus for a defensive hold and overriding loss values based on win rate are sound ideas.



thank you for your detailed reply. Was very busy the last days so i had no time to answer. first i have to say, that's great that you are thinking about/planning changes and that you listen to suggestions from the community.

yeah i wasn't sure how easy it is to implement some of my suggestions, but i found that this approach would balance the numbers you gave us and it would be very good from a psychological perspective.
Especially the reduced losses for specific ratios it's a really good way to archive more fairplay i feel, because people with non exclusive roosters would have better chances to reach something and they deserve some protection too. a more gradual approach like you mentioned would be even better, i just wanted to try to keep things simple.
i'm sure you guys will find some smart solution for the technical problems just wanted give some players input, since a fresh perspective can help from time to time :).

one big problem came to my mind again:
if the defense wins will give us a huge points boost, this would open the doors for cheaters. i could even outpace jofer, when i go with my 2nd account vs. my main account and loose intentionally in 30 seconds per round (20 secs of them loading time). i could earn 4500points in 3 hours (assuming 1 fight per minute and 25 wins per defense) or 6750 points (assuming 1,5 fights per minute and 25 points) which would be great for the big saturday rush ;) while it would be very obvious, sneaky players could use it in a more incognito way. i wrote a thread some months ago about this possibility to exploit the system without an attack limit, i think you saw that:

http://forum.ironmaidenlegacy.com/showthread.php?4938-How-to-exploit-PVP-vol-II&highlight=cheat

so some kind of attack limits would be necessary to prevent people from cheating, especially if defense would matter even more. i think, that this has been overlooked by now.

anyway, i'm curious to see the next changes hoping for some awesome improvments in the points and rewarding structure. :-)

cheers

Sparton_LOTB
05-05-2017, 11:17 PM
yeah i wasn't sure how easy it is to implement some of my suggestions, but i found that this approach would balance the numbers you gave us and it would be very good from a psychological perspective.
Especially the reduced losses for specific ratios it's a really good way to archive more fairplay i feel, because people with non exclusive roosters would have better chances to reach something and they deserve some protection too. a more gradual approach like you mentioned would be even better, i just wanted to try to keep things simple.
i'm sure you guys will find some smart solution for the technical problems just wanted give some players input, since a fresh perspective can help from time to time :).

Yeah, like I mentioned above, the idea of overriding gains specifically for defensive holds or cutting losses when at specific win rates is definitely a good idea, it'll just be a matter of the where and when to prioritize tech time for those kinds of changes.


one big problem came to my mind again:
if the defense wins will give us a huge points boost, this would open the doors for cheaters. i could even outpace jofer, when i go with my 2nd account vs. my main account and loose intentionally in 30 seconds per round (20 secs of them loading time). i could earn 4500points in 3 hours (assuming 1 fight per minute and 25 wins per defense) or 6750 points (assuming 1,5 fights per minute and 25 points) which would be great for the big saturday rush ;) while it would be very obvious, sneaky players could use it in a more incognito way. i wrote a thread some months ago about this possibility to exploit the system without an attack limit, i think you saw that:

http://forum.ironmaidenlegacy.com/showthread.php?4938-How-to-exploit-PVP-vol-II&highlight=cheat

so some kind of attack limits would be necessary to prevent people from cheating, especially if defense would matter even more. i think, that this has been overlooked by now.

Definitely been something we're aware of, we're just hoping that our current ability to monitor and adjust based on people exploiting the system would be sufficient. Obviously that's been kicked up now that someone actually did do that, but it's still something we don't condone and will not hesitate to ban users/adjust scores accordingly if people do engage in such activities.

Nonetheless, we are investigating the cost of implementing a measure which should stop the exploit from being possible instead of relying on manually correcting instances that occur.

Nicko
05-05-2017, 11:23 PM
Nonetheless, we are investigating the cost of implementing a measure which should stop the exploit from being possible instead of relying on manually correcting instances that occur.

Absolutely needed. You guys have your hands full as it is - to try to monitor thousands of accounts for cheating? Seems impossible.

Glad you're looking into this!

Sparton_LOTB
05-05-2017, 11:30 PM
Absolutely needed. You guys have your hands full as it is - to try to monitor thousands of accounts for cheating? Seems impossible.

Glad you're looking into this!

A combination of automated checks, random sampling of accounts, and player reports can sound surprisingly reasonable when an engineer says something could take days to do (and then who knows how much more for QAing and further fixes), but yeah, at some point you've burned more time than the cost to implement the preemptive measure in the first place.

Ultimately, we don't want to sit on an issue that actually affects live players if we can help it (such as someone being artificially inflated with points). Systematic issues (people being able to inflate others, or people sometimes still being in a state where they're not matchmade against frequently) are another matter, as those take time and usually cannot be pushed to live players on a whim.