How many times do you grind a 1 pt node for placement?

Calnexin
Calnexin Posts: 1,078 Chairperson of the Boards
edited August 2016 in MPQ General Discussion
The new PvE system has most players adjusting. It's good for some, bad for others.

In the new system, have you ground a 1 pt node to achieve a placement reward? How many hits did you do, and what did you get out of it?
Failed to load the poll.

Comments

  • mpqr7
    mpqr7 Posts: 2,642 Chairperson of the Boards
    I can't always do four full clears per sub. I'm just trying hard this time because I really want that duck!! Although weekends are often tough to play, since I'm out with family instead.
  • Fightmastermpq
    Fightmastermpq Posts: 995 Critical Contributor
    You'd have to play a node hundreds of times at 1 pt to make up the difference in just waiting for it to refresh. If you finish your final grind and need to grind a 1 pt node in the final minutes for placement that just means you started your final grind too early and should have waited. I know it's hard to predict how long a final grind will take but regardless - grinding a 1 pt node should really never be a part of your initial strategy in an event.
  • Phumade
    Phumade Posts: 2,477 Chairperson of the Boards
    This will not be a popular post, but yes!

    1pt grinding is now very important to finishing Top 1/2

    This is what my observation has been.

    1. Reducing the number of clears also reduces the separation between players (specifically at the top 10 level)
    2. Locking the first 4 scores also reduced the separation. (Mathematically, its because gaining a 1 pt lead in the 1st pass, is effectively multiplied in the success passes. This situation was most apparent when we were doing 2 clears and a final grind.


    Those two changes were wonderful with respect to QoL and I'm not really interesting in moving backwards unless its for a special 1 off event.

    The net effect is that initial grind is much more important to establishing the "ceiling" of possible points for the sub. and the locked scores (first 4 passes) guarantees that all optimum players ending the opening grind with = pts but wildly different clock times.

    A first blush look would imply that crashing the nodes as close to end would be the ideal strategy. But, I think the node heal rates is actually lower than 5* DPS. Specifically, a 5* can harvest a node for 1` pt faster than that node can regenerate to 2 pts. Many of acknowledge and recognize this fact, and there is no doubt that this strategy won't be sustainable beyond 15 to 20 min bursts.

    But the possible scenario can and will occur where two 5* optimal scorers will be within 30pts of each other. In that scenario, can 1 pt grinding make a difference? I think yes, there are definitely ways to engineer a clear schedule that lets you get near optimal scoring, but still provide enough spare time to 1 pt grind to pip another player.

    Last point to bring up.

    Many people will say wait until the Very Very last second to collapse the final grind. This concept really asks "how many points do I gain by waiting an extra min to start my grind". But I would challenge back and ask. Can I generate more pts per min by doing something else vs waiting.

    Overall PVE has become much more balanced and competitive. Many players could effectively knock out 90% of the field in 2 subs. Now the margins are much closer and its much harder to obliterate someone in pve
  • broll
    broll Posts: 4,732 Chairperson of the Boards
    I'll never grind a 1pt node.
  • OneLastGambit
    OneLastGambit Posts: 1,963 Chairperson of the Boards
    As a consistent t10 player I've never done it. For t5 you Dont need to either but I imagine those ocd maniacs in t2 probably need it
  • wirius
    wirius Posts: 667
    Never because I have self-respect.
  • Fightmastermpq
    Fightmastermpq Posts: 995 Critical Contributor
    As a consistent t10 player I've never done it. For t5 you Dont need to either but I imagine those ocd maniacs in t2 probably need it
    I took 2nd in the Wasp release event without it. I even fell asleep for about an hour during my initial grind in the first sub and was able to stay in it by clearing as quickly as possible in later subs. Clearing quickly is much much much more important than grinding. By clearing just 20 minutes faster you can make up over 300 pts in an 8k sub. Even if you can get through that easiest match in 30 seconds, that's still 2.5 hours of grinding to get 300 pts. No thanks. Spend that 2.5 hours in lightning rounds, or shield sim and use the ISO you make there to fund some boosts that let you clear faster instead.
  • Phumade
    Phumade Posts: 2,477 Chairperson of the Boards
    As a consistent t10 player I've never done it. For t5 you Dont need to either but I imagine those ocd maniacs in t2 probably need it
    I took 2nd in the Wasp release event without it. I even fell asleep for about an hour during my initial grind in the first sub and was able to stay in it by clearing as quickly as possible in later subs. Clearing quickly is much much much more important than grinding. By clearing just 20 minutes faster you can make up over 300 pts in an 8k sub. Even if you can get through that easiest match in 30 seconds, that's still 2.5 hours of grinding to get 300 pts. No thanks. Spend that 2.5 hours in lightning rounds, or shield sim and use the ISO you make there to fund some boosts that let you clear faster instead.


    Thats my point. Can someone engineer a schedule that puts them within 30 pts of the leader going into the final grind. and Can I create a clearing schedule that maximizes node values, AND also lets me grind?

    The devil is in the details of how you execute that schedule, but you won't need to grind 300 pts. I suspect that the best organized players can get that number down substantially, easily to under 20pts.
  • Fightmastermpq
    Fightmastermpq Posts: 995 Critical Contributor
    Phumade wrote:
    As a consistent t10 player I've never done it. For t5 you Dont need to either but I imagine those ocd maniacs in t2 probably need it
    I took 2nd in the Wasp release event without it. I even fell asleep for about an hour during my initial grind in the first sub and was able to stay in it by clearing as quickly as possible in later subs. Clearing quickly is much much much more important than grinding. By clearing just 20 minutes faster you can make up over 300 pts in an 8k sub. Even if you can get through that easiest match in 30 seconds, that's still 2.5 hours of grinding to get 300 pts. No thanks. Spend that 2.5 hours in lightning rounds, or shield sim and use the ISO you make there to fund some boosts that let you clear faster instead.


    Thats my point. Can someone engineer a schedule that puts them within 30 pts of the leader going into the final grind. and Can I create a clearing schedule that maximizes node values, AND also lets me grind?

    The devil is in the details of how you execute that schedule, but you won't need to grind 300 pts. I suspect that the best organized players can get that number down substantially, easily to under 20pts.
    I'm sorry but the math just doesn't agree with you. If the easiest node is worth 200 pts then you miss out on 200 pts by not letting it refresh. That means you have to make up those 200 pts......1 pt at a time. That's 200 matches.....just to break even. Then you are playing even more to get ahead.....again 1 pt at a time. I'll concede that if someone is unwilling to grind 1pt nodes their score is capped at a max of 7 clears and that max could be exceeded by someone willing to grind all day, but the time and effort required to do so is massive and highly unlikely to be the only difference between top spots.
  • Phumade
    Phumade Posts: 2,477 Chairperson of the Boards
    I'm sorry but the math just doesn't agree with you. If the easiest node is worth 200 pts then you miss out on 200 pts by not letting it refresh. That means you have to make up those 200 pts......1 pt at a time. That's 200 matches.....just to break even. Then you are playing even more to get ahead.....again 1 pt at a time. I'll concede that if someone is unwilling to grind 1pt nodes their score is capped at a max of 7 clears and that max could be exceeded by someone willing to grind all day, but the time and effort required to do so is massive and highly unlikely to be the only difference between top spots.


    icon_e_wink.gificon_e_biggrin.gif Totally disagree with you. But this is all about tactical execution and I'm not really interested in discussing the specifics of that. I think providing a broad strategy encourages players to experiment with the specifics to refine optimal for everyone.

    Over time we will see how strategies change. I posted my theory, because I think there is a fundamental change that can be exploited against prevailing strategy. But only more players experimenting with this strategy can ultimately prove or disprove it.
  • carrion_pigeons
    carrion_pigeons Posts: 942 Critical Contributor
    edited August 2016
    Phumade wrote:
    I'm sorry but the math just doesn't agree with you. If the easiest node is worth 200 pts then you miss out on 200 pts by not letting it refresh. That means you have to make up those 200 pts......1 pt at a time. That's 200 matches.....just to break even. Then you are playing even more to get ahead.....again 1 pt at a time. I'll concede that if someone is unwilling to grind 1pt nodes their score is capped at a max of 7 clears and that max could be exceeded by someone willing to grind all day, but the time and effort required to do so is massive and highly unlikely to be the only difference between top spots.


    icon_e_wink.gificon_e_biggrin.gif Totally disagree with you. But this is all about tactical execution and I'm not really interested in discussing the specifics of that. I think providing a broad strategy encourages players to experiment with the specifics to refine optimal for everyone.

    Over time we will see how strategies change. I posted my theory, because I think there is a fundamental change that can be exploited against prevailing strategy. But only more players experimenting with this strategy can ultimately prove or disprove it.

    Well, the math can be done. Obviously if you can complete a node fast enough, and are willing to grind long enough, 1pt nodes will eventually be worthwhile. The question is whether that threshold is actually reachable by players.

    On my current sub (Sim day 2), one full clear at max points (not counting the 1-time node, which I'll go ahead and ignore since everyone who is competing for a high score will get the same value from it) is worth 3375. So hypothetically speaking, a player who could clear everything at optimal times in zero seconds would, as I understand it, be able to earn 3375*(4+1+2/3+1/3)=20250 points without resorting to playing 1-point nodes. So a good way to judge the viability of grinding 1-point nodes is to think about how many times they would need to grind them in order to reach 20250 points.

    Grinding a sub down optimally isn't actually very different from a regular clear for most of it. The only difference is that there will be one node (presumably the most trivial node) that you never let regenerate. So the rest of the points will be calculated as above: 3225*(4+1+2/3+1/3)=19350. Add in the points for the first node before you grind it down (750), to get 20100. So you need to grind that one node 150 times in order to be able to beat the theoretical max-sans-grinding.

    Assuming a 5* team that only needs to make three matches to down the node with match damage, it presumably takes ~20 or so seconds to beat that node (7 clicks plus the time of the victory screen and rewards screen), so it'd take about 3000 extra seconds (48 minutes) of grinding the one point node in order to break even, and if you go the full 24 hours, it represents an extra ~4000 points. That sounds a bit mind-numbing to me, but if the goal isn't to actually max points but just beat the theoretical non-grinding optimum, you can let the node refresh for most of the time, and make an easy 99 points off the timer, and then only need to grind 52 times (~17 minutes) at the tail end in order to beat the break even level by a single point.

    That strategy is essentially the same as what's called a suicide burn in rocket science, and I'm going to go ahead and call it that here too, because I like the term, and because the reasons not to do it are basically the same as in rocket science, which is that the margin of error is ridiculously slim. You're basically gambling on your ability to make every click exactly perfectly for an extended period of time, with zero lag or computer issues or extended cascades or sudden heart attacks or anything.

    Regardless, the point remains that certain people certainly do have rosters where grinding the easiest node for a sufficient period of time will beat anyone who doesn't grind. The optimal time to do it is before your final clears of the other nodes, though, and since we don't live in a world where those can be done instantaneously (but instead will take, what, an hour?), that will make the total number of 1-point clears needed rise to something like 60. So if you actually want to suicide burn to beat the non-grinders by one point, and you have a roster that can beat the easiest node in 20 seconds, you'll need to set aside about 20 extra minutes before your final clears to grind out at least 60 extra points.

    If you need more time than that, then take 1440/(9.6-X) and subtract (90.4+X) from it, and multiply the result by X, (where X is the number of minutes your average clear would take) to get the number of minutes before your clears you'd need to set aside to break even with non-grinders. So for example, if it takes you 3 minutes from first click to last click to get in, beat the node, and get out, then it'll take you a little over 6 hours to do a suicide burn.

    Hope this was helpful.

    EDIT: (As a sidenote, this math only applies to the current sub, because the minimum node has a base number of 150 points. If the node you're grinding has a different number, then the 9.6 in the above math changes to whatever 1440/the base point level is, and the 90.4 changes to (2*the base point level squared-4320)/(3*the base point level).)
  • Phumade
    Phumade Posts: 2,477 Chairperson of the Boards
    I like carrier's analysis and I'll leave it to other to keep sharpening the details.

    But I'll rephrase the issue more clearly.

    I am not suggesting that someone should open the easy node to grind 1pts continously.

    I am saying that you can engineer a schedule that has you playing "traditional optimal" strategy and let you fit "pockets" of 1 pt grinding.

    ====
    As an example, a sub fight between two elite level 5* pve players is often decided on less than 50 pts. 5* rosters can easily farm 2 matches a min against the easiest nodes (the right cascade can litterally kill all 3 goons in 1 turn), so we are talking about less than 30 min of play.

    In a 90 min final grind situation, people's pace of clearing will change based on cascades and luck and many players will find themselves either ahead or behind schedule. This in turns means either speed up or SLOW down your pace of play.
  • carrion_pigeons
    carrion_pigeons Posts: 942 Critical Contributor
    Phumade wrote:
    I like carrier's analysis and I'll leave it to other to keep sharpening the details.

    But I'll rephrase the issue more clearly.

    I am not suggesting that someone should open the easy node to grind 1pts continously.

    I am saying that you can engineer a schedule that has you playing "traditional optimal" strategy and let you fit "pockets" of 1 pt grinding.

    ====
    As an example, a sub fight between two elite level 5* pve players is often decided on less than 50 pts. 5* rosters can easily farm 2 matches a min against the easiest nodes (the right cascade can litterally kill all 3 goons in 1 turn), so we are talking about less than 30 min of play.

    In a 90 min final grind situation, people's pace of clearing will change based on cascades and luck and many players will find themselves either ahead or behind schedule. This in turns means either speed up or SLOW down your pace of play.

    50 points is a lot of points when you're talking about squeezing blood from a stone. I would guess that unless you're willing to spend a very large amount of time grinding, this would never be worth it. (And if you are willing, why aren't you willing to do it optimally?)

    Every "pocket" of grinding 1-pt nodes you decide to do resets your timer. That means that every time you reset your timer without playing continuously to the end of the sub after that point is non-optimal, and represents a chance for other people to catch up to you. For example, if you have a pocket of grinding a 150-pt node halfway through a 1-day sub, you have to grind it more than 50 times in order to just catch up to where you would have been if you had left it alone. And that's just adding one pocket. Every pocket you add makes things harder.
  • Phumade
    Phumade Posts: 2,477 Chairperson of the Boards
    For example, if you have a pocket of grinding a 150-pt node halfway through a 1-day sub, you have to grind it more than 50 times in order to just catch up to where you would have been if you had left it alone. And that's just adding one pocket. Every pocket you add makes things harder.

    Does the pocket have to come at the 1/2 pt? Is there is a viable threshold even closer to sub end?

    At a fundamental level I would argue these pts.

    1. Is 15m of suicide burn viable for a normal person, is the threshold 10m? 5m? is 30pt boost "outcome determinative"?

    2. In the last 30 min of sub grind can I engineer a situation where I can convert slack time into sucide burn?

    3. We fundamentally assume that matches take 0 time and that we can open easy to hard in a linear fashion. In reality, nodes take time to open. so can we begin to take advantage of the fact that some nodes grind to 1 faster than others?

    are these little advantages enough to become "outcome determinative" in a sub race between 2 champed OML/PHX/ maxed punisher teams.

    Only time and actual real world battles will tell that tale. Go ask me on Line at "I snip you snap" I'll be more than happy to share my exact grind order. I just don't like discussing specific tactics on a d3 monitored forum.
  • Phumade
    Phumade Posts: 2,477 Chairperson of the Boards
    Some might wonder if this is an academic exercise that never occurs in the real world.


    Here is my current sub leaderboard.
    I've hit main mode 2x on schedule, its around 1hr 25 min after sub 2 start, and all clcoks are running. (everyone has the same sub 2 score atm)

    Overall leaderboard
    1. 28515
    2. 28484
    3. 28461
    4. 28348
    5. 28341
    6. 27965

    a 23 pt difference is around 10 min of 1pt grinding. When the margins are that thin every point advantage can mean the difference between award tiers
  • Fightmastermpq
    Fightmastermpq Posts: 995 Critical Contributor
    Phumade wrote:
    Some might wonder if this is an academic exercise that never occurs in the real world.


    Here is my current sub leaderboard.
    I've hit main mode 2x on schedule, its around 1hr 25 min after sub 2 start, and all clcoks are running. (everyone has the same sub 2 score atm)

    Overall leaderboard
    1. 28515
    2. 28484
    3. 28461
    4. 28348
    5. 28341
    6. 27965

    a 23 pt difference is around 10 min of 1pt grinding. When the margins are that thin every point advantage can mean the difference between award tiers
    So I was busy when Sim opened and couldn't play the first sub so I decided to bracket snipe and got in on a fresh S3 at #23 about an hour after the final sub opened. I finish my first 4 clears in 1h 17m. Was 5th to finish. All 5 of us at 19060. I checked a little later and the entire T10 was filled with guys at 19060. I noticed throughout the night and the next day that we were steadily dropping lower and lower, but none of the 5 of us changed score. Others were grinding 1 pt nodes (actually saw one guy get to 19250 and start going up 1 pt at a time for a while), some where over-clearing, but we all stayed. With 1h 30m left in the event I checked my leaderboard. Those same 4 guys were still right in front of me at the same score in 33rd or so. I waited. 1h 20m....no change...1h 15m I see #4 is moving. I wait, but I can't stand it any longer - I'm worried about not having enough time to finish myself. I start moving with 1h 10m, hit everything 3 times, and finish in 1st with 8m left. In the last 8 minutes I dropped to 4th. The only difference in our T5 was me and #4 switched places.....because he started his final grind sooner.

    Speed matters more than anything else.
  • Phumade
    Phumade Posts: 2,477 Chairperson of the Boards

    Speed matters more than anything else.


    Of course, no body denies that or suggest anything different. The question is can you develop a schedule that creates the most flexibility balancing speed and consistency and scoring opportunity.

    Let me describe the problem in a different light to demonstrate where and why and how we can look at the opportunity.

    Where:
    Specifically we should look at the last 90 min of the sub. Ultimately, this is really what we are talking about. How to play the last pass to maximize pts.

    Why:
    Intuitively that means play the highest points nodes last and as close to zero as possible. But this concept fails to account for the risks for failure. Many players are quite comfortable with timing out 1 node, but what about 3, what about 27 nodes. Can you predict and execute a consistent pace of play over that grind? Are the consequces of finishing early as great as finishing late? Would a "rational" player build in a buffer time vs an "optimal" player who times to finish within 1 min of sub end. In addition, once we chain additional nodes does this affect the buffer time. These are not trivial problems that are easily solved symbolically on the back of napkins at lunch.

    How:
    Once we accept that virtuallly all players prefer to build in a small time buffer (because the conseques of leaving points on the table is higher than finishing 1 min ahead of close." and that no player can simultaneious crash each node. We must then accept that node 1 (easy node) has to be ground before node 9 (hardest). The creates a time gap that can be mined for points faster than if they just let the node heal. The only real question is if I have to clear node 1 before node 9, and I have to account for Good/bad cascades, can I tactically recognize opportunities that might allow me to make up points against a "traditional-optimal" using "Risk mitigated-optimal"

    Method of Investigation:
    Perhaps an alternative analysis could be to assign a probablity on the success of beating a node in X amount of time. Repeating this analysis (i.e. monte carlo simulation) might tell us that there should be a modest buffer to increase the chances of clearing all nodes while accounting for bad and good cascades.

    But really this is an engineering challenge this is trying to amplify the difference between the heal rate and the DPS rates.
  • mckauhu
    mckauhu Posts: 740 Critical Contributor
    Never say never. Maybe in that case if positioning makes any difference to rewards and I'm just 1 point behind someone else and have already played every node to 1pt (which is extremely rare for me. Nearly not going to happen)